INVESTMENT AND TRADING

An Interactive Qualifying Project Report

Submitted to the Faculty of
WORCESTER POLYTECHNIC INSTITUTE

In partial fulfillment of the requirements for the

Degree of Bachelor of Science
by:
Adrian C. Delphia
Brendan M. Hamm
Srinivas Vasudevan
Samuel T. Veilleux
Submitted Tuesday March 13th, 2012

Advised by Professor Hossein Hakim

Table of Contents
List of Figures ................................................................................................................................................ v
List of Tables .............................................................................................................................................. viii
Abstract ......................................................................................................................................................... 1
Chapter 1: Introduction ................................................................................................................................ 2
1.1 Intro to Markets .................................................................................................................................. 2
1.2 Project Description.............................................................................................................................. 5
Chapter 2: Background Research .................................................................................................................. 7
2.1 Intro to Forex ...................................................................................................................................... 7
2.2 Fundamental Analysis (FA).................................................................................................................. 8
2.2.1 Introduction to Fundamental Analysis......................................................................................... 8
2.2.2 Macroeconomic Models .............................................................................................................. 9
2.2.3 Central Banks............................................................................................................................ 11
2.2.4 Interest Rates ............................................................................................................................. 11
2.2.5 GDP and Debt ............................................................................................................................. 12
2.2.6 Inflation ...................................................................................................................................... 13
2.2.7 Unemployment .......................................................................................................................... 14
2.2.8 Consumer Price Index (CPI) ........................................................................................................ 14
2.2.9 Using Fundamental Analysis ...................................................................................................... 15
2.3 Technical Analysis ............................................................................................................................. 16
2.3.1 Introduction to Technical Analysis ............................................................................................. 16
2.3.2 Candlesticks................................................................................................................................ 16
2.3.3 Support and Resistance Lines .................................................................................................... 25
2.3.4 Moving Averages (Simple and Exponential) .............................................................................. 26
2.3.5 Moving Average Convergence/Divergence (MACD) .................................................................. 29
2.3.6 Bollinger Bands .......................................................................................................................... 30
2.3.7 Stochastic Indicator.................................................................................................................... 32
2.3.8 Commodity Channel Index (CCI) ................................................................................................ 34
2.3.9 Relative Strength Index (RSI)...................................................................................................... 36
2.3.10 Fibonacci Retracements ........................................................................................................... 37
2.3.11 Average True Range(ATR) ........................................................................................................ 39
ii

2.3.12 Directional Movement Indicator.............................................................................................. 40

2.3.13 Average Directional Movement Indicator (ADX) ..................................................................... 42
2.3.14 Parabolic Stop and Reverse (SAR) ............................................................................................ 43
2.3.15 Ichimoku Kinko Hyo ................................................................................................................. 45
2.3.16 Forces Index-FI ......................................................................................................................... 47
Chapter 3: Methodology ............................................................................................................................. 49
3.1 Trade Methodology .......................................................................................................................... 49
3.1.1 Personal Trade Strategy Adrian Delphia ................................................................................. 49
3.1.2 Personal Trade Strategy Brendan Hamm ................................................................................ 56
3.1.3 Trade Strategy for Srinivas Vasudevan ...................................................................................... 63
3.1.4 Trade Strategy for Samuel Veilleux ............................................................................................ 68
3.2 Business Plan..................................................................................................................................... 76
3.2.1 Profit Allocation and Customer Relations .................................................................................. 76
3.2.2 Trade Logistics and Money/Risk Management for our Group................................................... 76
Chapter 4: Trade Portfolio .......................................................................................................................... 78
4.1 Overall Performance Summary ......................................................................................................... 78
4.2 The Group Account ........................................................................................................................... 79
4.3 Adrian Delphias Account .................................................................................................................. 80
4.4 Brendan Hamms Account ................................................................................................................ 82
4.5 Srinivas Vasudevans Account ........................................................................................................... 83
4.6 Samuel Veilleuxs Account ................................................................................................................ 84
Chapter 5: Parameters for Launching Our Company .................................................................................. 86
5.1 Company Structure ........................................................................................................................... 86
5.2 Findings Regarding Effective Marketing ........................................................................................... 87
Works Cited ................................................................................................................................................. 89
Appendix A: Recent Macroeconomic News ................................................................................................ A1
A.1 The Pegging of the Swiss Franc ......................................................................................................... A1
A.2 The Largest Ponzi Scheme Ever Played............................................................................................. A3
A.3 United States Federal Bank Twisting the Yield Curve ....................................................................... A4
A.4 The Grecian Default The Euro and PIIGS ....................................................................................... A6
A.5 Selling your Debt Italy and Spain Auctions .................................................................................... A8
iii

A.6 Oil Pipeline ........................................................................................................................................ A9

Appendix B: Programming in the Forex Market ......................................................................................... B1
B.1 Abstract ............................................................................................................................................. B1
B.2 Programming for MQL4 .................................................................................................................... B1
B.2.1 MATLAB Interface ...................................................................................................................... B1
B.2.3 Trailing Stop For Live trade ........................................................................................................ B4
B.2.4 Draw Entry Exit Program ............................................................................................................ B7
B.2.5 Double Tops ............................................................................................................................. B14
B.2.6 Support and Resistance Indicator: Identifying Critical Levels .................................................. B23
B.2.7 Pivots........................................................................................................................................ B33
B.2.8 Bars .......................................................................................................................................... B38
B.2.9 Order History Write ................................................................................................................. B46
B.2.10 Currency Meter ...................................................................................................................... B50

iv

List of Figures
Figure 1: Dark Cloud Cover bearish reversal pattern on EURUSD 30 minute chart ................................... 19
Figure 2: Engulfing Bearish Reversal on a 5 minute EURUSD chart ............................................................ 20
Figure 3: Hanging Man bearish reversal pattern in a 5 minute EURUSD chart .......................................... 20
Figure 4: Hammer and Inverted Hammer on EURUSD 1 minute chart...................................................... 21
Figure 5: Shooting Star Bullish Reversal pattern on a 5 minute EURUSD chart ......................................... 22
Figure 6: Shooting Star bearish reversal on a 5 minute EURUSD chart ...................................................... 22
Figure 7: A HeikenAshi candlestick chart overlying a regular candlestick chart ......................................... 25
Figure 8: The figure on the left shows two different simple moving averages and the figure on the right
shows a simple moving average plotted with an exponential moving average. [51] ................................ 28
Figure 9: A detailed picture of the MACD indicator. In the top portion of the figure the two EMA are
plotted against the candlesticks and the bottom portion shows the actual indicator. The height of the
histogram bars are the difference between the two EMAs. In this case the major EMA is the 26-period
EMA and the minor is the 12-period EMA shown above. .......................................................................... 30
Figure 10: A graph of the Bollinger bands plotted. At any point each outside band should be equidistant
from the moving average middle band. The outer bands can be used for levels of support and resistance
respectively as shown in the figure.[51] ..................................................................................................... 32
Figure 11: Two versions of the Stochastic plotted. The top one is the fast version and the bottom is the
slow. One can see the general pattern of the currency pair with the indicator, as when they remain
above the 80 line the market seems to be overbought, and oversold when below the 20 line. ........... 34
Figure 12: An example where the Commodity Channel Index predicted a bull market successfully; noted
by the first 3 upward sloping arrows. The currency pair increased in value as the CCI passed the 100 level
each time. [52] ............................................................................................................................................ 35
Figure 13: An RSI graph plotted with a currency pair. This shows how the RSI can chart the current
strength or weakness of a security.[51]...................................................................................................... 37
Figure 14: An example of how the market obeys the common Fibonacci levels as support and resistance
levels. Moreover, it shows how trading with this alongside the previously discussed MACD indicator can
show entry positions into the market. [52] ................................................................................................ 38
Figure 15: From this figure one can see how increasing ranges support the idea that traders will keep
selling or buying a stock throughout the day, or let it consolidate if the range is decreasing. [52] .......... 41
Figure 16: The ADX Indicator with DI+ in Blue, DI- in Orange, and ADX in lime green ................................ 43
Figure 17: Parabolic SAR ............................................................................................................................. 44
Figure 18: Ichimoku Kinko Hyo with Tenkan-sen = Red, Kijun-sen = Blue, Chinkou Span = Lawn Green,
Senkou Span A = Orange, and Senkou Span B = Purple. The hatched areas are the Kumo. An orange
Kumo is a bullish signal and a purple Kumo is a bearish signal. Notice the Kumo is shifted forward and
thus is an indicator of future sentiments.................................................................................................... 46
Figure 19: Forces Index - Notice how long periods of small movement has little impact on..................... 47
Figure 20: A sample graph showing when it is allowed and not allowed to trade according to my system.
[53] .............................................................................................................................................................. 51
Figure 21: EURUSD Scalping. ....................................................................................................................... 54
Figure 22: Summary of three previous trades made. ................................................................................. 55
v

Figure 23: A Sample trade for the above trading strategy ......................................................................... 59
Figure 24: An example of a Double Top pattern with a minor nested Double Top pattern and two minor
nested Double Bottom patterns. ................................................................................................................ 60
Figure 25: A sample trade from my final strategy. In this trade a sell stop was placed with sufficient
room to an un-shown resistance line indicated by the top yellow line drawn by our program
pivots_v1.07. Our trailing stop program then takes over the trade, ensuring a good exit point. ............. 61
Figure 26: Srinivass Second Trade.............................................................................................................. 67
Figure 27: Support and Resistance Lines [54] ............................................................................................. 68
Figure 28: Setup for a Trade Based on One Resistance Line ...................................................................... 72
Figure 29: Execution of a Trade Based on One Resistance Line ................................................................. 73
Figure 30: Choosing stop-levels and exit strategy for market moves based on support and resistance
levels ........................................................................................................................................................... 74
Figure 31: Setup for a Trade Based on a One Support and One Resistance Line ....................................... 75
Figure 32: Execution of a Trade Based on a One Support and One Resistance Line .................................. 75
Figure 33: Performance Chart for All Accounts .......................................................................................... 79
Figure 34: Performance Chart for the Group Account ............................................................................... 80
Figure 35: Performance Chart for Adrian Delphia ...................................................................................... 81
Figure 36: Brendan Hamm's Trades ............................................................................................................ 83
Figure 37: Srinivas Vasudevan's Trades ...................................................................................................... 84
Figure 38: Samuel Veilleux's Trades............................................................................................................ 85
Figure 39: This is the CHFEUR value during the pegging. Notice the massive drop instantaneously and
how it isnt really a peg since the SNB agreed to buy it at any lower price, and it isnt following the Euros
value precisely. ............................................................................................................................................. 2
Figure 40: A Screenshot of the Draw entry exit Program. Two short position trades are shown in this
chart. (Green and yellow lines are from a separate indicator). .................................................................... 8
Figure 41: Double Top Pattern contains 3 trend lines denoted A, B, and C ............................................... 15
Figure 42:Results of the Double Top/Bottom program "DoubleTopsv3" ................................................... 16
Figure 43: Sliding Window .......................................................................................................................... 23
Figure 44: Identified Extrema ..................................................................................................................... 24
Figure 45: Extrema Sorted .......................................................................................................................... 24
Figure 46: Applying Buckets ........................................................................................................................ 25
Figure 47: Support Output .......................................................................................................................... 25
Figure 48: The support and resistance lines drawn automatically using our expert advisor. .................... 26
Figure 49: Pivots indicator set to find 10 pivot points (yellow lines). The extrema are shown as bolded
green lines. In this instance the lowest point is the only point of interest for the bottom bucket, thus
only 9 yellow lines appear. ......................................................................................................................... 34
Figure 50: The Bars Program is displayed in a separate window from the chart. The blue histogram
displays the body size for each individual bar, while the red line displays the total bar size. The black and
purple lines are for a 5-period moving average of the body and total size respectively. Similarly, the
green and orange lines are for a 12-period moving average of the body and total size respectively. ...... 39
Figure 51: Zoom of the Bars Indicator using the same values as the previous figure. ............................... 39
vi

Figure 52: Currency Meter Running in Table Mode ................................................................................. B53

Figure 53: Currency Meter Running in Graphical View ............................................................................ B54

vii

List of Tables
Table 1: Overall Trade Summary for Group 5 ............................................................................................. 78
Table 2: Group Account Trades................................................................................................................... 80
Table 3: Adrian Delphia's Trades ................................................................................................................ 81
Table 4: Brendan Hamm's Trades ............................................................................................................... 82
Table 5:Srinivas Vasudevan's Trades .......................................................................................................... 83
Table 6: Samuel Veilleux's Trades ............................................................................................................... 84

viii

This page intentionally left blank

ix

Abstract
The goal of this project was to create a successful trading strategy for use in the forex market
and create a positive track record which could be used to launch a money management company.
Several different trading strategies were considered, and were subsequently tested both through live
trading and through programming automated trading robots. Additionally, other programs were
created along the way, both to aid in manual trading and in data gathering and processing. With our
performance history we then sought out possibilities for launching a money management company.
Such strategies and examinations of the markets illuminate both the benefits and detriments of several
potential trading philosophies, and provide a solid background for the beginning trader.

Chapter 1: Introduction
1.1 Intro to Markets
In the world of finance, a capital market is any place real or virtual in which financial instruments
are traded between different parties. There exist many different kinds of markets in existence today,
but they all operate by way of financial transactions of securities for currency. A variety of financial
instruments are traded on a regular basis, including stocks, bonds, commodities, and even currency
itself. All financial markets operate on the basic principles of supply and demand and thus the basis of
investing is to create a portfolio which is expected to rise in value over the course of time (i.e. is
expected to increase in demand). [43, 44, 45]
The most well-known of all the markets is the stock market, where partitions of companies are
bought and sold. There are many different stock markets around the world, some of them physical
trading rooms like the New York Stock Exchange (NYSE) while some only provide electronic transactions
like the NASDAQ exchange. [45]
Stock exchanges trade partitions of companies in exchange for currency. When a company
wishes to raise capital for various operating expenses or expansion projects they often seek to become
publicly traded companies. In the U.S. to become a publicly traded company an institution must first file
with the governing board, the Securities and Exchange Commission. Once this has been done the
company must now be approved to be listed on an exchange, which often have criteria the company
must meet to remain a listed stock. Once it has done this an Initial Public Offering (IPO) is announced
within that exchange. [44]
The IPO is basically a one-time sale of a partition of the company (referred to as a stock) at a
price set by that company. The company then gets the capital raised from this sale; however the
purchaser of this stock now owns a portion of the company. Sometimes these shares are voting shares,
which entitles the holder to vote upon business decisions made by the company, while other times they
2

do not give the holder of the stock any voice in the operation of the company (non-voting shares). For
this reason any entity controlling more than 50% of all voting shares of a company, holds the controlling
share as they can win any vote held by the company.
Regardless of the type of share held by an entity these shares can now be resold within the
market which they are listed under. These shares retain value based upon the principles of supply and
demand. If a company is performing well, the stock of that company will rise in value as more investors
seek shares of that company, whereas if it does not perform well the price falls as investors try and
unload their shares on the market. [45]
An offshoot of the stock market is the mutual fund market. A mutual fund is a conglomeration
of several different stocks managed by a mutual fund manager. By creating these funds the manager
allows investors to purchase shares of an entire sector or a diversified fund. Since stocks are only sold in
integer values, many high value stocks may not be feasible for some investors, but mutual funds allow
for investors to hold partial shares of these same stocks. In addition mutual funds allow investors
securities which are less volatile than stocks as the performance of an individual company is secondary
to the performance of the fund, which in some cases is the entire sector. [46, 48]
The bond market is one in which currency is exchanged for a bond, a security which offers a
specific interest rate of return over a given period of time. When this time expires (known as date of
maturity) the bond can then be exchanged back to the issuer in exchange for the purchase price plus the
interest accumulated over that time, very similar to a loan. The most common types of bonds in the U.S.
are issued by the Treasury Department to provide liquidity to the U.S. economy. Many other nations
also issue these bonds to raise capital for their economies. These types of investments are generally
seen to be extremely low risk investments, but generally offer low rates of returns. Bonds are not
always low-risk as some nations have bad track records for defaulting on loans (i.e. Greece), but the
bonds of these riskier nations often interest investors by offering higher interest payments. Additionally
3

in the U.S. many states, towns, and cities offer municipal bonds. These bonds are obviously riskier than
Treasury Bonds, as they do not have the large economical backing provided by the taxpayers of the
nation, but once again offer higher interest rates. [47, 48]
The commodities market is one in which physical goods are traded such as gold, oil, and
agricultural products. Commodities can be traded either directly at the time of the transaction, but are
often traded as futures. Futures are essentially transactions that take place today, with the promise of a
deliverable product at a specified date. They provide price stability to large companies which rely upon
receiving these goods at a future date for a price set today. In this manner airlines can ensure the price
of a plane ticket several months into the future by ensuring that they will receive their jet fuel at a
specific price set today. [50]
The foreign exchange market (forex) is the largest and most volatile of all the markets today. In
the forex market a trader can exchange one form of currency for another. Such markets provide
liquidity to entities which require foreign capital to conduct transactions abroad. Unlike other markets,
the majority of transactions do not result in any deliverable product. In fact an entity must be registered
to actually physically exchange these currencies. Despite this these non-receivable transactions still
provide liquidity to the market, and help it operate more efficiently. [49]

1.2 Project Description

This report investigates basic ideas and information related to investments and trading.
A variety of topics and ideas are discussed, ranging from a daily publication from a renowned
investor about trading, to different fundamental and technical indicators used to predict
security movement. Trading stations and trading systems are discussed and developed as well
as current political and economic issues affecting global economies such as the Grecian default
crisis going on now or Operation Twist occurring in the United States presently.
There is a heavy focus on indicators used to trade forex profitably, and the difference
for each one is lengthily discussed. Many different common and more advanced methods for
trading using the technical indicators are described, and the equation for each indicator is
explained, showing where it came from and how it is plotted on the charts. Fundamental
factors are discussed and whether investing based off of fundamental or technical indicators is
generally a safer trade. Finally, programming in MQL4, a language developed to create
automated trades on the Meta Trader 4 platform is reviewed and a relatively simple program is
demonstrated from scratch to buy and sell under certain conditions based off of technical
indicators.
This report is an introductory melting pot of many aspects of investing, with a primary
focus and relation to foreign currency exchange. Many indicators are used to trade multiple
securities such as stocks or commodities, but the described applications and methods used
cater specifically to currency trading. This report dives into a variety of related investment
topics, and how the worlds current economic, political and social health interlaces with the
currency market; Everything affects Everything! However, all topics discussed here only scratch
5

the surface of each idea compared to the level of intricacy each topic can be studied in depth
by professional investors.
Another heavy portion of this report is focused on individual and group trading
portfolios. Each individual had an account valued at $100,000 leveraged 50:1 and trades were
made according to their plan so that they were like our real capital being invested. With our
portfolios, the possibility of starting a company was researched lightly.
Finally, programming became a major focus of this project, creating many useful tools
for the trader. Programming is a useful way to make life simpler, faster, and more efficient in
everyday tasks. When applied to the foreign currency exchange market it can be used to create
powerful tools that can lead to ridiculous profits in some cases. Whether it is the newest
indicator that nobody else is using, or the latest automated trading robot that makes 20% a
week programming is effective in the market. Moreover it can be used to present the data in a
different way which can be in turn used to see trends and make custom indicators.

Chapter 2: Background Research

2.1 Intro to Forex
The foreign exchange market (forex) is a market in which different currencies are exchanged
between parties. Historically, the forex was instituted in order to provide banks and companies liquidity
in various currencies rather than using gold or other precious metals for foreign transactions. [1]
However, the use of computers today has reshaped the face of trading markets, allowing many smaller
players to enter the market.
In today's forex market the majority of currency trades do not result in an actual physical
transaction, but rather a position in the market to be resold to a third party at a later time. In fact, many
of the online brokers which are used by traders never deliver any currencies. However, these players
who act as middlemen provide liquidity to the market, which in turn allows it to operate for the
purposes of its induction.
The forex market is the largest market with an average of $4 trillion USD (U.S. dollar) traded
every day, based on 2010 information. [2] In 1998 the daily turnover was $1.5 trillion every day. [2] It is
not hard to see that the growth rate in this market is astounding and continues to grow. Year after year
the USD remains the highest traded currency; more than 40% of all forex deals include the USD as one
of the currencies. [2]
In the forex market currencies are exchanged not sold, so if one wishes to purchase EUR (euro)
they must buy it with another currency i.e. USD. In a spot transaction a buy (long position) order would
be placed with a broker on EUR/USD at a specified price and when there is another party who places an
order to sell EUR/USD, the broker completes the transaction. The latter party would be said to be selling
(short position) the EUR/USD. The broker then makes profit based on the spread, which is the difference

between the buy and sell price. The spread is typically between 1 and 5 pips of the base (first) currency,
for the major pairs.
One of the features that make forex trading so popular today is the leverage. In the United
States the forex market is limited to offering a 50:1 leverage, meaning for every 50 units of that currency
a speculator only need to invest 1 unit of that currency to hold a position, while in other countries
regulation may allow for up to a 500:1 leverage. An interesting consequence of this leverage system is
that it becomes possible for an investor to lose more than the original amount invested. Thus, forex
brokers hold the right to exercise a margin call. Margin calls occur when an investor no longer has
enough free margin in their account to maintain a current position, in which case the transaction is
closed immediately.
While spot transactions make up a large portion (37%) of the total volume of forex transactions
that take place, foreign exchange swaps make up 44% of all forex trades. [2] A swap is a transaction in
which one currency is exchanged for another for a specified amount of time, the time of maturity. At the
time of maturity the currencies are then exchanged back to the original parties. This is a common type
of investment for large corporations who wish to hedge different positions with differing maturity
terms, to mitigate risk in the market.

2.2 Fundamental Analysis (FA)

2.2.1 Introduction to Fundamental Analysis
Fundamental Analysis in the world of securities is a methodology of determining the true value
of a security, based on macro-economic issues, past data, and long-term trends. The basis of such a
system is that in such a large market the short-term prices often differ from the "true" price of a security
and that over a longer timeframe the "true" price will be approached. [32] In the Forex market there are
several main contributors to the base price of a currency pair. Since currencies are always traded in pairs
the fundamental analysis essentially comes down to a comparison of the current and expected future
8

economic conditions of the countries in question (or region for EUR). Thus, FA is largely influenced by
the central banks, their policies, and their leadership, as well as domestic factors such as GDP, interest
rates, unemployment, and manufacturing. [32] Reports on those topics are released at predefined
intervals governing bodies credited to determine them.
Given the importance of these numbers, speculations are made for their values before
the official release. If the actual data released is greatly different than widely accepted speculative
estimates, great volatility can be seen in the market. Economic reports, such as unemployment or GDP,
are means by which a country's economic health is directly measured. These reports are released to the
public at predictable, scheduled intervals. It has been proven again and again that if these reports
indicate the presence of even a small deviation from the status-quo, it can cause huge changes in the
market. These reports, along with speeches from important officials (such as Chairman of the Federal
Reserve Bank of USA, Secretary of the Treasury, President of the Federal Reserve Bank of San Francisco
and so on7) are watched carefully by forex traders.
Speeches by the aforementioned weight-carrying officials are watched closely enough that there
are websites dedicated solely on notifying users about when speeches and reports are happening. This
is used by some as an indicator that the market could become highly unpredictable during that time
period, and dissuades them from trading. Other traders study how these reports change the market
and, by utilize a real-time feed to the report or speech, trade the news (anticipating market movement
given what has been reported). That practice can be highly lucrative, but is viewed by some to be too
risky to trust.
2.2.2 Macroeconomic Models

While no individual model is consistent in predicting the future of the forex market, there are
several models in use today which attempt to quantize the market. Several of the models used in our
9

current economic climate do a decent job at predicting and shaping exchange rates, but all of them have
their own assumptions and flaws. "The Purchasing Power Parity Approach (PPP) holds that in the long
run, exchange rates will adjust to equalize the relative purchasing power of currencies." [32] The
essential basis of such a model is that goods are worth the same value in different countries and the
difference in currency value stems from the overall purchasing power of a nation. Thus, PPP is largely
based on tradable goods and services while neglecting many other issues which influence exchange
rates. This model is decent at predicting long term prices and excels when modeling currencies
experiencing high inflation, however is limited in its inability to make short to medium term predictions.
[32]
John Williamson has put forth the concept of "fundamental equilibrium exchange rate" (FEER),
in which a country is to develop a macroeconomic strategy to achieve an internal balance in terms of
unemployment and inflation. [33] The problem with this concept is that there is little consensus on the
proper way to measure the equilibrium. Using different models, different attempts to define this
equilibrium price can have very different results. "John Williamson has suggested that FEER calculations
could not realistically justify exchange rate bands narrower than plus or minus 10 percent." [32] The
Monetary Approach to exchange rates is based upon the total supply and demand of a currency. Since
central banks have monopolies on the supply of currency available, this system really boils down to the
demand and central banking policies. This approach is limited in that it does not account for bonds or
the supply and demand in the goods and services sectors. However, this model does account for the
much greater short-term influence of policy while the prices of goods and services tend to lag behind.
[32]
The Portfolio Balance Approach (PBA) is a theoretical improvement on the monetary approach
in that it does not assume bonds and other financial assets to be perfectly substituted whether in
foreign or domestic regions. However, there is no universal agreement as to a proper way to value all
10

such instruments simultaneously. The PBA approach is often a much more complex model than other
approaches and while using one system of relations may produce accurate results in one circumstance,
it is unlikely to predict results in others. Due to this, many financial institutions prefer to use simpler
models which have a more universal agreement. [32]
2.2.3 Central Banks

The Central Banks of a nation (or region) provide the basis for most any fundamental analysis, as
they directly control the rate at which currency is produced. It is the duty of a central bank to attempt to
keep their currency within a band or target zone around a FEER. [34] By increasing the production of a
currency the central bank induces inflation and thus decreases its true value. In the United States the
Treasury, in consultation with the Federal Reserve System, has the responsibility for setting U.S.
exchange rate policy, while the Federal Reserve Bank New York is responsible for executing FX
intervention. [35] The Federal Reserve is controlled by the Federal Open Market Committee(FOMC),
which is a revolving board of the members of the reserve banks and treasury. The Federal Reserve is a
system of 12 different regional banks controlled by one governing body. [34] The Federal Reserve differs
from other central banks in that in addition to shaping the monetary policy of the nation, it is also
tasked with unemployment policy. "Economic developments in this country [United States] have a major
influence on production, employment, and prices beyond our border; at the same time developments
abroad significantly affect our economy." [23]
2.2.4 Interest Rates

Interest rates, both long and short term, are largely influenced by the central banks as they set
the rates at which private sector financial institutions borrow money. In addition, the influx of money
(through the purchase of bonds) results in a decline in interest rates of both parties, as is common with
11

the current U.S. China financial relationship. [36] One of the best predictors of economic climate is the
yield curve, which is the spread between the interest rates of treasury notes with differing maturity time
frames. The yield-curve method for forecasting markets has outperformed most other methods for longterm predictions. This is important due to the fact that policy decisions usually have a long-term impact.
[37]
2.2.5 GDP and Debt

Historically, inflation has been the main concern of developed nations when it comes to
monetary policy. However, in the majority of developed world economies, successful strategies for
controlling inflation rates have been so successful that financial instability has become the major issue.
[24] In recent years, with many nations facing financial crisis, the debt toGDP ratio has increasingly
become the most important factor to currency evaluation.[22]
Gross Domestic Product (GDP) is the total value of all the goods and services produced internally
and externally by a nation. GDP and inflation have a high correlation as rapid growth can induce inflation
devaluing currency, raising prices on goods and services, and decreasing the purchasing power of a
nation. [39] Despite this, GDP remains the best method for determining the financial stability of a
nation.
A nation's debt can belong to either of two categories: domestic debt and foreign debt. Foreign
debt occurs when one nation provides loans to another nation, while domestic debt is owned by private
entities within the nation. The most typical form of such debts is bonds. While there are many different
types of government bonds available domestic debt does not affect the value of currency nearly as
much as foreign debt does.
Foreign debt arises largely from the unwillingness of the borrower to allow direct foreign
investments into the infrastructure and private sector alike. Historically, Treasury bond yields have been
12

below the growth rate of the economy, which is in part due to the security of the U.S. dollar. The
national debt in the U.S. has been consistently growing faster than GDP since 2000 from approximately
38% to 53% in 2009. [40]
2.2.6 Inflation

According to Mishkin, Inflation policy involves a medium-term strategy for monetary policy
which incorporates five main elements: "1) the public announcement of medium-term numerical
targets; 2) an institutional commitment to price stability as the primary monetary policy; 3) an
information inclusive strategy in which many variables, and not just monetary aggregates or exchange
rate, are used for deciding the setting of policy instruments; 4) increased transparency of the monetary
policy strategy through communication with the public and markets about the decisions of the
monetary authority; and 5) an increased accountability of the central bank for attaining objectives." [36]
Inflation directly impacts the exports and imports which a nation experiences. When inflation
occurs, the relative value of the currency drops making goods and services from that nation cheaper to
foreign investors. When this happens, exports typically increase while imports decrease. "Over time, the
depreciation in the home currency would lead to the determination of exchange rates growth in the
nations exports and a decline in its imports, and thus, to an improved trade balance and reversal of part
of the original depreciation." [32]
While moderate inflation may be good in developed economies, it can have serious
repercussions for smaller economies. Many of these economies rely much more heavily on imports and
hyper-inflation will cause the prices of these imports to balloon. Such a circumstance can wreak havoc
on the economies of smaller nations, as has been seen throughout much of the Asian market towards
the end of the 20th century. [41]

13

2.2.7 Unemployment

Unemployment is greatly affected by such factors as exports, imports, and interest rates. It is
this correlation that makes it an important indicator of the current financial stability of a nation.
Unemployment can be directly linked to financial climates, as can readily be seen by the high levels of
unemployment experienced in underdeveloped nations undergoing hyperinflation. In addition,
unemployment can be the result of deflation in developed countries, as the strength of currency directly
reduces the nation's ability to export goods.
Unemployment also acts beyond the broader scope of macroeconomics and begins to affect the
psyche of a population. If a nation begins to experience higher levels of unemployment, the population
risks going into an economic stasis by which they become much more conservative in their financial
dealings. In such a case the liquidity in markets begins to dry up, as consumer spending drops and
creates a so-called "snowball effect", which leads to recession, and perhaps depression. This has been
readily seen in the U.S. since the collapse of the stock market in 2008.
2.2.8 Consumer Price Index (CPI)

The Consumer Price Index (CPI) gives a households approximate cost for a basket of goods and
services, in comparison to the cost of that same basket in 1982-1984. [42] The data is released
monthly, in the middle of the month. Every year before the release of Januarys numbers, all the data
gathered previously is considered and new indices are calculated. The new indicies are adjusted for
seasonal factors, and replace the previous years CPI data, along with that of the last 5 years. The CPI is
used by the federal government to make adjustments to cash flow mechanisms such as pensions,
Medicare, and adjustments to insurance policies. Resultantly, the CPI tends to have some effects on
every investor in some way. Investors with a fixed-income should of course assure that their profits land
them ahead of inflation, for having a yield which is not ahead of inflation leads to a loss of real wealth
14

(despite a numerically apparent monetary gain!). The U.S. index is based on the prices of goods and
services, including food and beverages, housing, apparel, transportation, medical care, recreation,
education, and communication, that people buy for day-to-day living in urban areas across the country.
[42]
Because the CPI can be considered a relatively direct measure of inflation, it is of great value to
the forex trader. Because it marks the buying-power of the dollar, it can cause great movement in the
stock market, as well as the Forex market. A CPI which indicates that the value of the dollar is sharply
decreasing can be expected to drive investors to change their preferences on the economy to invest.
[42] Resultantly, the value of the dollar will fall even farther, because that preferential change will tend
toward an increased selling of the dollar. That in turn drives away more investors, and further decreases
the value of the currency.
The government also keeps close watch on the CPI, so attempts can be made to keep the
currency stable. Accordingly, the states monetary policy is made with these statistics in mind [42]. If
indices point towards an inflating dollar, the monetary authorities are likely to increase interest rates in
an effort to drive the dollar value back up. Similarly, if deflation is indicated, lowered interest rates are
assumed with the goal of restoring the dollars former value. Additionally, values such as Social Security
beneficiary payouts and food-stamp amounts are derived from the CPI so payouts can stay at a relatively
constant economic value, even if the dollar amount has changed due to inflation.
2.2.9 Using Fundamental Analysis
Many of the preceding indicators are affected heavily by each other. For instance, Federal
banks use Consumer Price Indices to make decisions based on their monetary policy. If a CPI points
towards an inflating dollar, the monetary authorities are likely to increase interest rates in an effort to
drive the dollars value back up. Thus, any single fundamental indicator should not be used blindly to

15

forecast the markets movement. Instead, several fundamental indicators should be considered
together, to determine if a common trend can be discerned.
Using several different fundamental indicators, a picture of the economic conditions driving the
forex market can be painted. However, a completely different picture might be painted by the data in
the recent price charts. There are several reasons for this, but what is important is in understanding
how to use both fundamental indicators AND the technical indicators discussed in the next section.
Since bothtechnical analysis and fundamental analysis each provide a unique view of the same picture,
their use as complimentary methods can be quite successful.

2.3 Technical Analysis

2.3.1 Introduction to Technical Analysis

The following comprises a list of indicators that help one decide when and if they should
enter a trade position on the foreign exchange market. They are different from fundamental
indicators in that they treat all currencies the same. They are a pure mathematical formula
applied to past data of a currency to help predict current and future trends. The theory behind
technical analysis lies in the assumption that major economic influences tend to have little
effect on short term fluctuations and thus are negligible. They rely on past values that the
currency has traded at, and since they are purely statistical they are considered more reliable
than trading the news to most investors.
2.3.2 Candlesticks

Candlestick charts are used widely in various markets today to graph the price
fluctuations of currencies, stocks, and commodities. They were first introduced in the
16

18thcentury to chart the prices of rice in Japan, but have since been applied to everything which
is traded in an open market. [3]
A candlestick chart uses a symbol commonly referred to as a bar, which conveys the
price of the open, close, low, and high for the given time increment of a chart. Typically
candlesticks will have one of two colors depending on whether the price has increased or
decreased in that time interval. Typically if the close price is higher than the open price, a
candlestick will be green and if the close price is lower than the open price it will be red.
However, most programs contain the ability to change the colors of these charts, so one must
be familiar with the program they use.
The candlesticks consist of two major parts, the wick and the body. The body is the wide
part of a candlestick and contains the open and close information for that time period. The wick
extends off of the main body and shows the high and low prices for that period. The main
reason that candlestick charts are so widely used is that they only incorporate the exact prices
which a currency pair has traded at, without any calculations. After all, price is the most
important piece of information an investor can have.
Additionally, since the market is in constant oscillation often the market overreacts and
a specific trend will eventually reverse itself. Candlesticks can be particularly useful in
determining when this reversal will occur and there are several patterns which are indicative of
such a case.

17

Classes of Candlesticks

Candlesticks take on a variety of different sizes and shapes (respective to the tail/body
size and ratios of both). If a candlestick bar has no tail then we term it a Marubozu. If it has a
small body and the tails are of equal length as the body, it is dubbed a spinning top. In both of
the prior cases the candlestick by itself is usually pretty irrelevant when we try to determine a
trend reversal. A similar candlestick is one which has a short body and no tails called a 4-price,
which usually only happens during suspended trading periods. This is a pattern one would
never wish to trade, as it is indicative of low volume.
A Doji is a candlestick which has very long wicks and a small body. A Doji is usually
indicative of market-indecision and makes for a poor entry position as the price may continue in
either direction. Similar to the Doji are the dragonfly and gravestone candlesticks. Both
incorporate a small body, but the dragonfly has 1 long tail on the bottom and the gravestone
has its long tail on the top. Both of these are single bar reversal patterns; the dragonfly is a
bullish reversal whereas the gravestone is a bearish reversal. Strength of the dragonfly is
increased by a previous bearish period, while the reverse is true for the gravestone. [3]
Candlestick Trend-Reversal Patterns
Dark Cloud Cover and the Piercing Line

Two of the most important reversal trends are the Dark Cloud Cover and the Piercing
Line patterns. In a Dark Cloud Cover pattern a bullish trend must have occurred for four
consecutive bars followed by a bar which has a close lower than the opening. In addition, this
last bar must have opened above the high of the previous bar and closed below the midpoint of
18

the previous bars' opening and close mark. The Dark Cloud Cover is indicative of a bearish
reversal pattern. The Piercing Line is the opposite, a bullish reversal pattern such that there is a
bearish trend for four bars followed by a positive bar whose close is above the midpoint of the
prior bar. In both cases it is usually safe to wait for a "confirmation bar" in the form of a
candlestick continuing in the direction of the reversal trend.

Figure 1: Dark Cloud Cover bearish reversal pattern on EURUSD 30 minute chart

Engulfing Reversals

There is also a class of pattern referred to as Engulfing Reversals. A Bearish Engulfing

reversal takes place when there is a bullish trend followed by a bar whose body extends long
enough to "engulf" the previous bar. Similarly a Bullish Engulfing reversal follows a bearish
trend and "engulfs" the prior bar. Similar, but not as strong as these signals, is the Harami
reversal pattern. The Harami bullish/bearish reversals are the same as those in their engulfing
counterparts, however rather than the last bar engulfing the prior one, the prior bar engulfs the
last bar.
19

Figure 2: Engulfing Bearish Reversal on a 5 minute EURUSD chart

Figure 3: Hanging Man bearish reversal pattern in a 5 minute EURUSD chart

20

Hanging Man and Hammers

The hanging man is another important bearish reversal pattern. It is characterized by a

bullish trend ending with a bar with a small body and a wick on the bottom which is longer than
the body. Whether this last bar's close is higher than its open or not is irrelevant, only the wick
to the bottom plays importance here. Likewise the hammer pattern is a bullish reversal pattern,
characterized by a bearish trend followed by the same last bar as the hanging man. In addition,
there is also a bullish reversal pattern called Inverted Hammer which is a very strong signal. An
Inverted Hammer occurs after a bearish trend ending in a long bodied downward trending bar
which is followed by an inverted (upside-down) hammer with a long wick up.

Figure 4: Hammer and Inverted Hammer on EURUSD 1 minute chart.

Shooting Stars

Another class of reversal patterns is the Shooting Star class. Following a bearish trend if
the last downward bar has a large body and is followed by a spinning top whose body is
21

completely below the body of the prior bar then we classify it a shooting star bullish reversal.
Similarly, if there is a bullish trend, whose last upward bar has a long body and is followed by a
spinning top, whose body is completely above the prior bar, we define it a shooting star bearish
reversal.

Figure 5: Shooting Star Bullish Reversal pattern on a 5 minute EURUSD chart

The opposite of the bullish shooting star is the bearish shooting star. It is shown in the
figure below.

Figure 6: Shooting Star bearish reversal on a 5 minute EURUSD chart

22

Raindrop

The Raindrop/Star reversal pattern is one which is usually only seen in longer n-period
candlestick charts. A bearish reversal trend begins with 3 upward trending bars have the
pattern long body, then short body (usually a spinning top) above the large body, then long
body below the short. After this last long bar if another spinning top follows with its body
removed that of the previous bar we classify this as a star pattern. The same is true if the first
three bars are bearish and the spinning tops are below the other bodies where the last has an
upward trend. In the latter case this would be a raindrop reversal. However, these patterns
themselves are not very strong or common and so it is wise to wait for a confirmation bar
before trading. In fact they are so rare in the forex market, after hours of searching one could
not be found for this paper.
Morning Star and Evening Star

Lastly, are the Morning Star and Evening Star patterns, which are two of the strongest
pattern signals one can see. The Morning Star is a bullish reversal in which a longer than normal
downward (usually 4 or 5 long bars down) trend has occurred, followed by a spinning top or
doji with an upward trend whose body is completely removed from the previous bars' body.
Finally, a last upward trending bar is needed to confirm the pattern; as such a large downswing
warrants some caution. The Evening Star is the bearish reversal counterpart to the Morning
Star and differs only in a bullish trend followed by the spinning top/doji above and removed
from the last bar. Once again a final confirmation bar must follow before trading. They are very
similar to the shooting star patterns and thus images have been omitted.
23

HeikenAshi

HeikenAshi is an offshoot of traditional candlestick chart which uses formulae for the
open, close, high and low, rather than using the actual values for any timeframe. They are
calculated as follows:

The HeikenAshi candlesticks are often used in trading because they make clear what the
current direction of a trend is. This method mainly alters the size of the wicks and it can readily
be seen that a bar with no upper wick is in a downward trend and a wick with no lower wick is
generally in an upward trend.

24

Figure 7: A HeikenAshi candlestick chart overlying a regular candlestick chart

2.3.3 Support and Resistance Lines

Support and resistance lines are the next most popular indicator used by traders. A
support line is a line drawn horizontally on a chart at a specific price that has not been crossed
below in a period of time. Likewise, a resistance line is one in which the price has not crossed
above in a time period. Usually several different support and resistance lines are plotted in
conjunction.
The basic premise of these lines is that they indicate prices, that when reached,
experience high levels of long or short positions respectively. Support and Resistance lines are
used in the market as indicators of entry and exit positions. When the price approaches either,
25

one can be more confident in their ability to "bounce" back as past pressures around these
levels have been experienced. This "bounce" is usually caused by larger institutions placing
orders at these levels. If support or resistance lines are broken then it is a sign that previous
sentiments around those prices has been broken and thus is indicative of a trend continuing in
the respective direction.
2.3.4 Moving Averages (Simple and Exponential)

Moving averages are a great indicator for identifying trend direction and strength and
some of the most commonly known and used technical indicators. When the slope is positive it
indicates an uptrend, and when its negative a downtrend. The steepness of the slope is
proportional to that trends strength [4]. It is possible to plot more than one MA at a time
which can combine to indicate buy and sell points. One common method is plotting a fast
moving average against a slower one. For example, the faster one follows the trend better so
the period is smaller, say 50 and the slower one would have a period of more than 50, say 100.
If the faster one cuts under the slower one it is a bearish crossover and it indicates a sell
point. If the fast one cuts above the slower one it is a bullish crossover and one might
consider buying. There are two common types of moving averages, exponential and simple
which are explained below.
Simple Moving Average (SMA)

The simplest form of the moving average plots the average value of the past x candle
sticks, where x can be chosen by the user. Say you set x to be 10 in MT4. For any given point in
time on the SMA plot, that value was determined by the average of the last 10. Depending on
26

your trading station what it plots can be customized such as plotting the past ten opening
values instead of the past ten closing values or whatever the user prefers.

In the above equation, xn represents the nth most recent data point of the user selected
value. For instance, if the user wanted a simple moving average of the past 25 days using
opening prices, x1 would be yesterdays closing price and x25 would be the closing price 25 days
ago. All of these are summed together and then divided by the number of points (in this case
25) to plot one data point.
Exponential Moving Average

Very similar to the SMA but it places more weight on recent data. So for an EMA with a
period of 10 minutes, data points from the most recent minute would weigh more than the 2 nd
most recent minute and so on. Traders tend to use exponential averages more because more
recent prices are considered more relevant than older ones, which the simple moving average
does not consider. This makes logical sense over moving averages with huge periods, since data
200 days ago could be out of date to predict todays price if there were other global factors
affecting data back then [4].

In the above equation, EMAcurrent is the most recent data point that is being calculated,
EMAmost_recentis the most recent data point collected. So for instance, if you were doing a 10 day
27

moving average, EMAmostrecentwould be yesterdays calculated EMAcurrent. Pricecurrentisexactly

what it sounds like, the current trading price and is generally expressed as a factor of the
period, shown in the following equation.

N can be days, or minutes, or any time frame chosen by the user. Alpha isnt always
calculated this way for all exponential moving averages, but this is the most common way it is
found. Alpha is just how much less each subsequent EMA data point will be worth. It is always
between 0 and 1 where a higher alpha means that older data is discounted faster.

Figure 8: The figure on the left shows two different simple moving averages and the figure on the right shows a simple
moving average plotted with an exponential moving average. [51]

One can notice how the moving average with a smaller period of 50 follows the prices
much more closely than the one with a period of 50. That makes sense because when the
period is 15 only the past 15 data points are used which were all closer to the current price
whereas the past 50 data points can contain values very far from the current price affecting the
28

whole line. These figures also show the crossover lines between a fast and a slow moving
average indicating buy and sell points respectively. It is hard to tell from the figure on the right
but the exponential moving average relies more heavily on recent data to plot the indicator
versus the simple moving average which treats every data point the same [4]. This is why the
two lines have the same period but different shape.
2.3.5 Moving Average Convergence/Divergence (MACD)

This is a very common indicator that can be utilized for trend and divergence
recognition. It essentially plots two different exponential moving averages. When the MACD
cuts through its horizontal line in the positive direction it generally indicates a bull market, and
vice versa a bear market when it cuts down through the trigger line. The vertical bars are the
difference between two customizable exponential moving averages, so when one EMA crosses
the other, the bars switch from one side to the other on the horizontal axis [5].

In the above equation, EMAminoris the exponential moving average with the smaller
period and the other term EMAmajor is the moving average with the larger period.

29

Figure 9: A detailed picture of the MACD indicator. In the top portion of the figure the two EMA are plotted against the
candlesticks and the bottom portion shows the actual indicator. The height of the histogram bars are the difference between
the two EMAs. In this case the major EMA is the 26-period EMA and the minor is the 12-period EMA shown above.

When the MACD is indicating a trend one way and the currency pair is going the
opposite way negative divergence is indicated. The author at forexindicator.org states that this
commonly leads to a downside movement, whereas during positive divergence an increasing
movement in price is usually noted by this author [5]. There are several different trading
strategies which use the MACD indicator as the basis.
The MACD crossover initiates a buy signal when the MACD rises above the signal line
and initiates a sell order when it falls below. It is also popular to buy/sell when the MACD
crosses zero. The MACD indicator is very insightful and can be even more reliable when
combined with other technical indicators described further into this section. [18]
2.3.6 Bollinger Bands

Bollinger bands, when first added to a currency pair seem intimidating due to the fact
three lines appear across your screen, but mathematically speaking, it is a simple indicator. The
30

lowest plot on your screen will be a simple moving average minus 2 standard deviations [6]. The
middle one is just the moving average and the upper on is the moving average plus 2 standard
deviations. The period of the moving average and number of moving averages can all be
customized in MT4. The general equation for Bollinger bands actually consist of three simple
equations, one for each line. The middle line is a simple moving average and the equation has
been shown previously. The other two equations making up the outside two lines are shown
below.

Where SMA is a moving average with a chosen period N. Note this can also be done with
an exponential moving average as well. K is a user chosen constant, usually around 2, and is
the standard deviation of the moving average. Essentially K is how many standard deviations
from the moving average you want each band to be [7]. While the equation used for the
standard deviation is very similar to its probabilistic definition, it should be noted that such
stochastic processes, as exist in the market, have no "regular" distribution. For this reason
standard deviation in this sense is merely a representation of such a concept in a continuous
environment.

31

Figure 10: A graph of the Bollinger bands plotted. At any point each outside band should be equidistant from the moving
average middle band. The outer bands can be used for levels of support and resistance respectively as shown in the
figure.[51]

Bollinger bands can be used in a variety of ways. The lower and upper bands often
indicate support and resistance respectively. Moreover, the width between the upper and
lower bands indicates market volatility. When they are very narrow the market is in a period of
consolidation with not much movement happening and when they spread wide there is a large
price movement. They are also commonly used to predict trend reversals, which is when the
trends slope direction switches polarities [8]. When the price dips below the lower band for
some time it could be an indicator to go long. This is great when the bands are very narrow and
they suddenly swing out wide; if you go with the trend you have a good chance to make some
quick pips.
2.3.7 Stochastic Indicator

This indicator is an oscillator that is generally used to measure momentum. It consists of

32

Two horizontal lines generally set at the 80 and 20 levels, and two signal lines, %K and %D.
There are two main flavors of the stochastic, namely fast and slow [8]. The fast one is more
sensitive to the market which causes more false market indications, whereas the slow one is
less sensitive but indicates a position after the market has already moved.

The terms Low and High are the lowest and highest price respectively over the past n
periods, chosen by the user. Priceclose is the most recent closing price. The %K line is plotted
alongside of the %D line which is simply an exponential moving average of %K with N=3.
When the signals move above the 80 line, the market is considered to be overbought,
and oversold when they move below the 20 line. If they stay above or below these limits this
can be an indicator of a strong trend in that direction. Additionally, this oscillator is a great tool
for entry and exit levels. If the %K line crosses the %D line and moves out of the 20 or 80 level
range it is time to go long or short depending on which crossed what [8] To go long would be to
wait for the %K to go above the %D and cross the 20 mark.

33

Figure 11: Two versions of the Stochastic plotted. The top one is the fast version and the bottom is the slow. One can see the
general pattern of the currency pair with the indicator, as when they remain above the 80 line the market seems to be
overbought, and oversold when below the 20 line.

2.3.8 Commodity Channel Index (CCI)

The CCI index is generally used as an entry and exit signal. It comes equipped with two
levels, 100 and 200 and if the index passes the +100 it signals a good time to enter a long
position, and exit once it moves back to or below the +100. One can infer the sell conditions
easily from this, using the -100 and -200 levels instead. Essentially, this indicator shows a
currency pairs variation from its mean value [9].

The formula looks complicated at first, but it just contains many functions. The outside
scalar of 66.67 is just to make the final result a neater number to display on a graph and is not
always used. The typical price is the average of the sum of the high, closing, and low prices. The

34

CCI is found by subtracting the SMA of the typical price from the typical price, and dividing that
by the standard deviation of the typical price.
Another use of the CCI is using it as a reversal indicator. If you draw a trend line with the
signal, using either the peaks or the valleys, and that line is majorly broken that could indicate a
reversal. It also has positive and negative divergence to show trend reversal, whether the pair is
going down and the signal is doing the opposite or vice versa. It usually signifies the market is
changing, but may take some time [9]. Finally, if the index passes the +200 mark and eventually
makes its way back down to the +100 level, go short on the trade, and vice versa for the
negative levels. Like all indicators, this is not a 100% guarantee; otherwise everyone would be
doing it!

Figure 12: An example where the Commodity Channel Index predicted a bull market successfully; noted by the first 3 upward
sloping arrows. The currency pair increased in value as the CCI passed the 100 level each time. [52]

35

This is one use of the CCI where it should be bought after passing the +100 level and
closed once it comes back down and breaks through the +100 level going in the opposite
direction. This is a very diverse indicator good for many trade predictions.
2.3.9 Relative Strength Index (RSI)

Much like the stochastic indicator, the RSI is another great way to find out if a market is
overbought or oversold. The RSI can range from 0 to 100 and oversold and overbought
conditions are marked at the magnitudes of 30 and 70 respectively.

The RSI is found by converting the relative strength of a signal (defined above as RS) to
an index between 0 and 100 as shown above. The relative strength is found by taking the ratio
of the average U and D values (represented by an EMA) which are the upward and downward
changes respectively. U is characterized by the difference between the current closing price and
the previous closing price. The difference between the previous closing price and the current
closing price is D [13].
If the price approaches a resistance level and the RSI goes above the 70 line,this
indicates a bearish market, and vice versa for a bullish market. The RSI can also be a trend
confirmer simply by comparing it to the 50 line. If its above the 50 mark its an uptrend and a
downtrend if below the 50 mark. Moreover, this indicator can be used as a divergence spotter,
noting when the market increases while the indicator shows a decreasing downtrend. This
alludes to the likelihood of the market switching from up to down.

36

Figure 13: An RSI graph plotted with a currency pair. This shows how the RSI can chart the current strength or weakness of a
security.[51]

As seen by the figure from investopedia.com, the market is overbought when the 70
level is breached and oversold when the 30 level is crossed. However this also can have positive
and negative divergence so one must be careful to watch for this!
2.3.10 Fibonacci Retracements

Based off of ratios created by the famous Fibonacci sequence, this indicator is excellent
for predicting support and resistance levels. The most common ratios are .382 and .618,
derived from the famous sequence. If a trend is bearish and reaches the .382 line it is likely to
bounce back up according to the Fibonacci retracement [14].
(

The previous equation shows one specific retracement level. There is no real formula for
this indicator; rather there is a group of ratios found using sequential Fibonacci terms. For
37

example, dividing any nth term of the sequence with the nth+1 term will yield a value close to
.6180 for all n. The larger the n, the more approximate it will be. That is just to find one level.
To find another, say the 38.2% level, dividing any nth term in the series with the nth+2 term will
approach .3819 for all n.

Figure 14: An example of how the market obeys the common Fibonacci levels as support and resistance levels. Moreover, it
shows how trading with this alongside the previously discussed MACD indicator can show entry positions into the market.
[52]

In this figure, the trader is using the MACD with the stochastic oscillator coupled with
the Fibonacci levels to spot support and resistance levels. Notice how they form at the ratios
from the Fibonacci series. The way the forex market and people trade is truly mysterious.
The basic idea behind these retracement lines is the tendency for a security which has
undergone a large price swing to undergo a period of correction thereafter. In such a
38

circumstance often these support and resistance levels are indeed the Fibonacci Retracement
lines, and thus are often used as levels for entry and exit positions. Usually one would expect a
trend to reverse to at least the 50 \% line if the trend is truly broken, however failure to do this
is indicative of a continuation of the current trend. One of the key advantages of Fibo
Retracement lines is that when you enter the market at one of these support lines, you can set
a very tight stop loss and thus minimize risk. [17]
This indicator is arguably used best with other indicators such as the MACD (describe in
point VI). If you see a currency pair retracing to a level and the MACD has flipped axis in a way
that agrees with the Fibonacci bounce direction this would be a good entry level. Additionally,
one can use the stochastic indicator as well to line up the Fibonacci retracement levels with
oversold indications from the stochastic to create buy entry levels [14].
2.3.11 Average True Range(ATR)

The Average True Range is a tool developed by Welles Wilder to measure volatility. The
ATR uses a measure called the true range and then plots an n-period moving average of the
values obtained. The true range is defined as the maximum of:

39

2.3.12 Directional Movement Indicator

The Directional Movement Indicator in conjunction with the ATR forms the basis of the
Wilder trading system. The directional movement index is split into two components the DM(minus DM) and the DM+ (plus DM). They are defined as follows:

The superscripts here are indices, whileidenotes the current bar and i+1 is the previous bar.
This is another unique indicator because it does not indicate price movement. Instead it
measures volatility of a pair. It takes 3 differences, between the recent high, low and previous
close and bases it off the largest value. Your trading station will then plot it automatically and
one can use the ATR to verify market strength to form decisions on your next position. Rather
than having the range being the difference between the days high and low, the true range
takes into account the previous days closing price if it happened to be outside of the current
days traditional range [10].
The ATR is simply the largest value of the minimum subtracted from the maximum. To
find the maximum it is the larger of the two values that include the previous closing price or the

40

high price of today, whichever is larger. Similarly, the minimum is found by taking the smaller
value of previous closing price, or the low of the recent day, whichever is smaller.
The smaller the ATR is displaying the more likely the market is or will be consolidating or
waiting to enter a breakout, that is a huge up or down trend after consolidation. If the value is
large it means the trend is strong or the breakout is actually happening and consolidation is
over.

Figure 15: From this figure one can see how increasing ranges support the idea that traders will keep selling or buying a stock
throughout the day, or let it consolidate if the range is decreasing. [52]

As seen on the graph provided by forexindicator.org, the value of the ATR is small while
the market isnt getting beyond support and resistance levels. Once the indicator increases the
trend breaks out and the resistance for this particular example was broken [10].

41

2.3.13 Average Directional Movement Indicator (ADX)

The Average Directional Index is an indicator developed by Welles Wilder which

incorporates all of his previous indicators. The ADX is a more complex method than MAs in
measuring the current trend of a currency pair. The basis for such an indicator is the directional
index (DI). The DI always assumes a positive number, however this value is assigned to one of
two variables, the plus-DI denoted DI+ or the minus-DI denoted DI-. Which variable takes on the
value is determined by the following equation:

The Directional Movement Index (DX) is defined as:

Then the ADXi is simply the exponential moving average of DX:

In MT4 the default is a 14 period EMA.

42

Figure 16: The ADX Indicator with DI in Blue, DI in Orange, and ADX in lime green

When the ADX is sloping downwards or below the 25 mark it usually means the markets
are consolidating. If it trends upwards above 25 this is usually a good time to buy [11]. Like the
CCI this also has divergence to present a possible trend reversal. If the ADX says the market is
consolidating and the pair is trending upward this could potentially be a good time to go short.
2.3.14 Parabolic Stop and Reverse (SAR)

The Parabolic SAR is yet another tool developed by Welles Wilder. It is similar to the
moving average; however it gives a slightly more obvious indication of trend and actually
calculates an expectation for the next bar into the future. The value of the SAR appears above
the price if it is a downward trend and below the price if it is an upward trend.

43

Where is the acceleration factor. MT4 sets by default = 0.02 and it doubles every bar
the current trend continues in that direction and returns to its default during a trend change.
The Pricei+1 is defined as the low of the previous bar if it was a downward trend and the high of
the previous bar if it was an upward trend.
The SAR is commonly used by many traders as an indicator for exit strategy, completing
their trades when the SAR changes its trend. In many cases it is also used as a trailing stop. [20]

Figure 17: Parabolic SAR

The Stop and Reverse is a popular indicator used mainly to identify the direction of
price movement. When the dots are underneath the candlesticks it indicates a time to buy and
indicates a sell when the signal is above the candlesticks [12].
44

This is a simple indicator that is easy to read in a simple glance. It works best with other
indicators to reaffirm trends. This indicator is not very helpful alone, especially if there is rapid
price movement. Another use of this indicator is setting a stop loss level. If you have decided to
buy, have your stop loss trail the previous Parabolic SAR dot. That way your profits will be
maximized and your losses will be minimized.
As shown on the figure previously, the SAR is simple to use. If the dots are above the
currency pair go short and if below go long. The steeper the slope from one dot to the next
shows trend strength.
2.3.15 Ichimoku Kinko Hyo

The Ichimoku Kinko Hyo system was developed by a Japanese newspaper man in
conjunction with several students whom he hired to do back testing. The base of the system
uses 5 different charting lines; it is a hyper visually based system and its name directly
translates to "Equilibrium chart at a glance". The lines are as follows:

45

Together the Senkou Span A and B form what is called the IchimokuKumo which
translates to cloud and is the area between these two lines. The Kumo is the shaded area in the
Ichimoku indicator and serves as the equilibrium range for the currency pair in question. A
trader utilizing this strategy will long pairs when the price breaks above the cloud and short
pairs when it breaks below the cloud.
The size of the Kumo is an indication of the current market volatility, where a wider
Kumo is a more volatile market. When the Senkou Span A is above the Senkou Span B this is
seen as a bullish signal akin to a MA-crossover strategy and Senkou A is similar to the fast-MA.
Likewise when the B is above the A it is a bearish sentiment.

Figure 18: Ichimoku Kinko Hyo with Tenkan-sen = Red, Kijun-sen = Blue, Chinkou Span = Lawn Green, Senkou Span A =
Orange, and Senkou Span B = Purple. The hatched areas are the Kumo. An orange Kumo is a bullish signal and a purple Kumo
is a bearish signal. Notice the Kumo is shifted forward and thus is an indicator of future sentiments.

46

Figure 19: Forces Index - Notice how long periods of small movement has little impact on

2.3.16 Forces Index-FI

The Forces Index is another trend tool which is removed from the regular chart window.
It is essentially a tool used to determine the current direction and strength of a trend. The
calculation for FI is:

Hereirefers to the current period, i+1 the previous, and MA anyn-period moving
average. The index in conjunction with a short moving average (2-period) is well suited to
finding exit and entrance positions. For higher order moving averages (13-period) the index
shows the trends and their changes. [24] Since this indicator incorporates volume it is a good
indicator of where the buy/sell pressure lies.

47

A typical FI trading strategy would involve buying when the FI becomes negative during
increasing tendency and selling when it becomes positive during decreasing tendency. When a
new trough/peak is reached in the FI indicator a continuation of the trend in that direction can
be expected. [24]

48

Chapter 3: Methodology
3.1 Trade Methodology
3.1.1 Personal Trade Strategy Adrian Delphia

After the first seven weeks of experimental trading and learning about the forex market and
different indicators, the next third of the course was dedicated to a more structured trading simulation;
this took place for the past recent seven weeks of the course. A new demo account valued at $100,000
leveraged 50:1 was setup and trades were made according to my plan so that they were like my real
capital for our company was invested. This account was less experimental than the one in the first part
of the course, and many things were learned now that I was trading according to a well-defined plan. It
was a dynamic plan that was modified along the way to increase profitability and reduce risk. The
changes made to the system and the reasons why are discussed in this section.
Trading these past few weeks taught me a lot of things about the forex market. I took trades
much more seriously this term and traded only to make a profit. In the first part of the course I made
many trades without even thinking just to see outcomes and try to gamble with the system. When I was
curious about a trade this term I would trade just a micro-lot just to see if the trend went the way I was
thinking but did not have enough confidence to put my real money into it (trade a standard lot or
more). For this reason I will only comment on the real trades I made for the purpose of furthering my
account profit which consists of trades where more than one standard lot was bought or sold.
One of the main things I learned from trading was the unpredictability and wildness of the
market. I did not think currencies could swing hundreds of pips against each other in every way!
Sometimes it even happened multiple times a day. It really shows how volatile the market can be, and
that there is definitely money to be made for the right trades. I also learned the importance of
fundamental news in the market, as many different things happening in Greece involving the debt crisis
49

immediately affected currencies everywhere. I never really thought it could have such a dramatic effect
as some did such as the currency strengthening significantly when Germany finally agreed to a bail out
voting in December for Greece.
My trading methodology had major modifications to it, but it mostly was simplified. In the first
third of the class my method was too confusing to follow or abide by. There were many conditions and
too many indicators involved to produce a clear method. I tried simplifying it for ease of use, but it
didnt add anything to my system, just made it less profitable. I have since simplified my system enough
that I can strictly trade by only its rules.
My current strategy consists of just trading off of support and resistance lines. At the start of my
trading I would just eyeball it and go look theres resistance so Im going to sell or buy if theres report.
This was the baseline for my strategy. However, as you will soon read, I have certain conditions in which
I am allowed to enter now.
Eventually, to make it more specific I needed to define exactly when I would buy or sell with
support and resistance. This is difficult to do when you dont know exactly how your mind is thinking but
I tried to narrow it down each week when I noticed another pattern or what was working and what was
not. My current entry position guidelines for my support and resistance system are as follows:
Support or resistance lines need to be obeyed at least more than 1 time unless it is a global
maximum or minimum in a 15 minute window or larger period
This is my golden rule, and every case of when I can and cannot enter a trade is outlined using the
following figure. I have labeled 8 points accordingly and describe whether or not I could enter a trade
and why.

50

Figure 20: A sample graph showing when it is allowed and not allowed to trade according to my system. [53]

For all below remarks; if it says allowed to trade that means if that line is tested again in the future
a position can be entered.

Point 1: Okay to trade that support line because it is obeyed more than once (like at point 2)

Point 2: Okay to trade that support line (See point 1)

Point 3: Okay to trade that support line because it is a global minimum

Point 4: Not okay to trade, local maxima obeyed only once

Point 5: Not okay to trade, local minima obeyed only once

Point 6: Okay to trade that support line because it is a global maximum

Point 7: Okay to trade that resistance line because it is obeyed more than once (like at point 8)

Point 8: Okay to trade that resistance line (See point 7)

51

In order for two peaks or two values have the same resistance or support lines, the difference in
height between the peaks in questions needs to bewithin the difference of the minor tick marks . For the
example figure used previously, the y axis ticks happen every 100 pips, sothe two peaks would need to
be within 100 pips of each other. This scales nicely for all window time frames you are using. Smaller
time scale means a smaller y axis range means the values need to be closer together to form a support
or resistance line which makes sense.
My exit position is a trailing stop loss between 10 and 20 pips. I used to not use trailing stop losses
and either set a discrete stop loss, or monitor the trade until I decided to exit for seemingly no reason.
Using the trailing stop loss every time was a great improvement to my plan since B term because it really
maximizes profits while minimizing losses. There are edge cases where a currency could theoretically
drop 21 pips before skyrocketing 200 pips making me lose all that profit however this is rare and not a
safe practice in trading and would affect my pull-down negatively.
These changes were necessary to make so I could figure out what was working and what was not.
Before if I was losing a trade I wouldnt know what to do differently because I didnt really know what I
did. If I just eyeballed a trade, and it didnt work, how could I know what to change to make a better
trade next time? Picking specific conditions and trading on them allowed me to see what worked and
what didnt. I then could change a specific part that didnt work and keep trading and evolving my
system. For example, in an earlier version of my system I could trade at any local minima or maxima. I
quickly figured out this led to many negative trades due to natural and common market noise and
volatility so I only trade on local minima and maxima if they are tested at least two times. This greatly
improved my profitable trade percentages.
Implementing the rule that a local peaks and valleys need to be hit more than once helped avoid
trading on false support and resistance due to normal market fluctuation and noise. I again modified this
rule a little more to exclude global minima and maxima. The reason for this is because if there is one
52

maximum or minimum over a huge period say 4 hours or something it is unlikely it will be busted
through so it is okay to have the option to trade on global peaks and valleys, especially if it is a line that
hasnt been reached in days and you think the market will continue obeying it.
My trade sizes have more or less stayed the same throughout my trading system this term. The
vast majority of my trades are one standard lot. Under certain conditions I do trade with two standard
lots but never more than that. My method for when to trade 2 standard lots is when the line has been
tested more than two times, and when viewed from a different time window, the line holds true still. A
line obeyed three times or more, especially across multiple time windows is more likely to hold true and
be profitable that it is worth risking more capital, 2 standard lots instead of 1.
My risk management also changed minimally through the past seven weeks. With being a new
trader, it is important to keep risk as low as possible to build and condition oneself to good habits. Low
risk is always better than high profit if one had to choose between the two. The backbone of my risk
management is always setting a stop loss. Whether it is a trailing stop loss or a fixed stop loss, I must
always place one to minimize risk. On short term trades it will be between 10 and 20 pips, my most
common trades. For long term trades it can be larger, around 100 or 200 pips, but I will never risk above
15% of my capital for my long term trades.
Trading an average of 1 standard lot on my initial investment of 100 thousand dollars means I
never risk more than 2% of my capital. Setting a stop loss to ten pips ensures I dont risk more than 5%
of my leveraged trade (assuming 1 pip is about 10 dollars). I dont like to enter trades often, but rather
wait for just the right trade to come along when I have almost no doubts. This will keep the risk low in
my account and lead to consistent trades.
Finally to reduce risk even more, once 2% of my capital is lost in a day trading will cease until
tomorrow. Additionally if I have made 3 consecutive good trades it is also time to stop. These two rules
will maximize profit while reducing risk at the same time. No reason to keep trading after a good day,
53

quit and do it again tomorrow rather than ending on a loss because I feel Im on a winning streak with
three positive trades in a row. Combining control with discipline is the backbone of my strategy with the
most important part being to strictly follow it, as I have designed it in a way where this can be done.
Here I will detail a sample trade made throughout B term and annotate it. You can see how well
it matches up to my methodology. To see a review of all trades made in B term refer to the Trade
Portfolio section of the report.
Out of all the trades I made this term I am going to explain the one that demonstrates my
trading methodology the best. It is swing trading according to my plan with the EURUSD currency pair.

Figure 21: EURUSD Scalping.

First, deciding if these similar peaks and valleys were support and resistance lines according to
my methodology was done. So the difference each minor tick value is 9.5 pips means that I can place a
support or resistance line if two peaks or valleys occur within 9.5 pips from one another. The support
line shown is easy to see that it follows my rule, and the resistance line follows the rule too but the
other peaks for the line appear off to the left cropped out of the picture to more accurately show the
trades made.

54

The left-most arrow was my first trade where it seemed to decline after resistance was nearly
met. I sold one lot and set a trailing stop of 20 pips and I closed the trade after a profit of 9 pips. I
continued watching this market data and again where the middle arrow indicates, the currency hit the
resistance line and started to go down. Sure enough the pair went down again and my trailing stop
activated profiting me 15 pips. I still further watched the market and after the currency bounced right
back up to the resistance line and started to decline I sold another lot. This one had a lot of jumping
around in it so I got out early happy with a profit of 4.5 pips. Looking forward in the future if I didnt go
to bed I couldve employed this strategy many more times during the consolidation to have made even
more pips!

Figure 22: Summary of three previous trades made.

The total profit was 28.5 pips all in 90 minutes of watching the forex market. What I have
learned is that consolidation can be a good thing if youre resistance and support lines are holding on.
Moreover I shouldve considering reversing my position when I bought the currency back at the lows,
that is buying the currency when I close my position. Then sell it when I open my next position. I
couldve in theory made twice the above amount! Finally, according to my rules I was allowed to sell two
standard lots on the last trade I made because the line was tested two or more times. I decided not to
which is fine anyways because I reached my performance for the week that da anyways, just good to
recognize for future trades.

55

3.1.2 Personal Trade Strategy Brendan Hamm

As a member of Empowerment Capital, we have been each given a $100,000 foreign exchange
account which we have been tasked with trading. As a member of this organization we were to develop
a trading strategy with which we were to adhere to in an attempt to create profit. As is expected for a
novice trader such as myself, such a trading strategy is not necessarily concrete, but rather evolves over
time. Over the course of the past couple of months my personal strategy has evolved in many facets for
a multitude of reasons.
Any trading strategy must match the goals of the investor. As part of Empowerment Capital, we
have created the goal of achieving a system which can generate 11% per year. With this relatively high
level of profit we can charge a management fee of 2% of our clients average annual account as well as a
20% performance fee. In part with this goal, minimizing drawdown is of utmost importance, as no
investor wants to experience large swings. For the purposes of this project, I would like to set a
maximum drawdown of 5% with the aim of achieving my goals in only 2%.
My trading strategy began outlined as follows:
1. Before beginning a trading session, making sure I am up to date on any current news will be of
utmost importance. I am not allowed to trade until we have read the Gartman Letter for that day.
In addition, I must check on ForexFactory.com for any news times which are scheduled. For any
currency pairs involved in scheduled news, I will not allow any open orders during a one-hour
block centered at the scheduled time. In addition, I will only trade major pairs which at least one
belongs to the current trading sessions region.
2. Multiple time frame charts will be brought up for the currency pairs in question. If any pair
exhibits similar trends on the 5-min, 15-min, 30-min, and hour timeframes, they will be
considered for trades in the following steps.
3. Every pair which has made it to this stage will be charted on the 5-minute time scale in
conjunction with two exponential moving averages of periods 5 and 15 respectively. In addition
56

a 50-period simple moving average will be used as a signal line. Only buy orders will be placed if
the ask price lies above the signal line and only sell orders if the bid price lies below the signal
line. The only exception to this SMA rule will be when a large movement in price takes place
followed by a retracement of over 50% in which case the signal line will be a 20-period moving
average. This adjustment will make trading possible during large oscillation cycles, which while
inducing greater risk also allow for greater returns. Due to this, all trades made using the 20period SMA will be subjected simply to the base lot size.
4. Since I am trading only pairs which have a general trend our next step is to create
support/resistance lines for the past 12 hours of trading. Emphasis will be placed on the most
recent support/resistance lines through the use of color (yellow) and lines which have been
broken during this period will be colored a dark blue.
5. I will then begin looking for entry positions both based on our EMA cross and signal lines, as
well as the resistance lines and candlestick patterns. I will not sell close to the lowest support line
and will not buy near the highest resistance level, unless they have been broken.
6. If a currency pair meets the entry position criteria then determining lot size will be the next step.
The base lot size will be 1 standard lot. If the chart incorporates candlestick patterns (i.e.
reversal/continuation patterns) favorable to my intended position this base lot size will be
doubled.
7. For long positions a stop loss will be placed below the next support line which lies below the
entry position. For short positions this s/l order will be placed above the next resistance line
which is above our entry position. The actual numerical value will be calculate/approximated as
5% the distance between that line and the next, below for long positions and above for short.
Similarly the take profit orders will be placed in a similar fashion using a distancing factor of
10%. These t/p values will lie above the next resistance line for long positions and below the
next support line for shorts.
57

8. If I am in a long position and that price crosses the resistance line which has t/p 10% above it, I
will update the s/l and take profit orders. This will be done by using the same 5% and 10%
calculations and adjusting them up to the next pivot point. In a similar fashion we will do this
with short positions. If a second update is made to any of the positions I will repeat all of the
previous analysis steps. If the pair still meets all of our criteria an additional lot may be added
using cognitive discretion. This extra lot will be treated in the same manner as all other trades.
9. If at any point there is indication (i.e. candlestick reversal) that a currency pair will not continue
in the direction of the trade it will be exited immediately.

My original strategy revolved heavily with the use of moving averages. However, as time progressed
I have reconsidered the heavy importance which the above trading strategy places upon these averages.
My original thought process in using them in the first place began with the notion that in capturing small
term price movements, a large volume of smaller trades could potentially see the returns I sought with a
low drawdown. However, just using moving average crosses seems to initiate a lot of trades during nonmovement periods and thus has a large number of small losses due to these false signals.

58

Figure 23: A Sample trade for the above trading strategy

For the reason of too many signals I had incorporated candlestick patterns to try and reduce the
amount of signals. It seemed a natural thing to do at the time, but for a novice trader with a myriad of
patterns, it is easy to confuse these candlestick patterns even though I keep graphics of them in front of
my computer at all times. In addition, this may have narrowed trade numbers down much more than I
anticipated and I find that in the majority of my sessions I did not trade.
For the reasons outlined above, I have taken a more pattern recognition approach to trading, while
negating the use of moving averages in my strategy. Throughout the course of this project we have
been utilizing WorldFXIQ.com to learn many basics of strategies for trading. Incorporated in this were
several lessons regarding patterns in price graphs. These patterns include Double Top/Bottom, Double
Pullbacks, Round Bottoms, Head & Shoulders, Gaps, Wedges, and Reversals. My final trading strategy
rested solely on the presence of these patterns.

59

Figure 24: An example of a Double Top pattern with a minor nested Double Top pattern and two minor nested Double
Bottom patterns.

A Double Top/Bottom pattern is one in which a currency pair experiences a reversal at a specified
level, referred to as a support for a bearish reversal or a resistance for a bullish reversal. The pair must
then reverse back again towards this support/resistance line. It is this point which becomes crucial to
this pattern as it will wither progress into this pattern or revert to be a Double Pullback. As can be seen
in(Figure 1), a Double Top occurs at the culmination of trend line C. At this point the price bounces
off of the resistance line. In this image the downward trend experiences a short period of sideways
trading, but as is often the case with this pattern in the EURUSD the final impact is tremendous, offering
a trader the possibility of a very profitable trade.
Similarly, there exist two nested Double Bottom patterns within this same trend pattern. During
these Double Bottom Patterns the bottom trend line acts as a support line and the price tends to rise

60

after reaching this level. A Double Pullback is also displayed in this same image as the support line
which the Double Bottom trend relies upon is broken at the far right of the chart.

Figure 25: A sample trade from my final strategy. In this trade a sell stop was placed with sufficient room to an un-shown
resistance line indicated by the top yellow line drawn by our program pivots_v1.07. Our trailing stop program then takes
over the trade, ensuring a good exit point.

Of the patterns described above, I have found the Double Top/Bottom pattern to be the most useful
and profitable trading pattern. However, this pattern alone can take a long time to develop, so with
limited time it becomes difficult to catch. For this reason I wrote a program which will automatically
recognize this pattern and trade accordingly. I have had some success in my beta testing for this
program but only on short position trades. I feel that redefining the way it operates can further improve
upon it, specifically for long position.
Another thing I found necessary to change was my lot size. I began trading only single and double
lots as outlined in the strategy above. I found that the fluctuation in my account was much less than
61

what I would want if it ever were to end on the positive side of the fence. As I moved into November
the majority of my trades netted profits and losses of around $100, or 0.1% of the total equity. Around
this time my highest drawdown was less than 1.5, which was within the allowable drawdown of my
strategy. For this reason, I adopted a system of trading 2 standard lots as my base size, while expanding
up to 4 standard lots for optimal positions.
Finally, my stop loss and take profit levels regarding the pivots had changed as well. Before, it was
stated that I would set targets at 5% and 10% respectively from the pivot lines. However, in reality this
tended to be too close, so I wrote a program to determine the average bar size over the course of the
past hour. I then use this average to place my stop loss, at a factor of 1.5 average bars (on the 5 minute
level), from the pivot in the stop direction. Utilizing this formula I believe my trading became more
isolated from random noise, while still allowing me to exit if the initial assessment is proved wrong.
Regarding my Take Profit Levels I have wrote an expert which cycles through my open orders and sets a
Trailing Stop using this same formula. My Stop Loss feature only enables if the current stop loss is
further away than the newly calculated level. These are two things which I find wouldve been
cumbersome to do manually.
Finally, at the end of B-term I decided to switch the program I was using to execute my trades from
TradeStation 9.0 to Meta Trader 4. Within TradeStation I found that I made many mistakes simply due
to a lack of knowledge of that trading platform. I felt much more comfortable with the MT4 platform
and have been spending much of the project writing scripts to aid in my trading.
My risk strategy has been set such that I never have more than 2% of my account on the line at
any given time. Using the previous strategy, this actual level has never been surpassed or even
approached. With the stop loss system I was using before my total equity at risk never exceeded 1% and
was generally less than 0.5%. Utilizing these new levels for trade size and stop loss my exposure to the
market has yet to exceed 1%, but remains generally above the 0.5% level.
62

3.1.3 Trade Strategy for Srinivas Vasudevan

As members of Empowerment Capital, LLC, we given a $100,000 foreign exchange account that
we are assigned to effectively trade. We are responsible for developing a trading strategy that delivers
returns consistent returns with a low drawdown. As a first step towards achieving this goal, we
familiarized ourselves with common risk-management strategies such as employing stop losses and take
profits and learned about various fundamental and technical indicators. Then, we developed our own
indicators, such as support and resistance lines, entry and exit point calculators, and other tools. During
this time, as novice indicators, we also familiarized ourselves with existing indicators such as the MACD,
Bollinger Bands, simple moving averages, exponential moving averages, and the Parabolic SAR. Since we
were unfamiliar with trading, our strategies tended to evolve over time and we eventually were able to
achieve concrete trading goals and measure our performance against our objectives.
My trading strategy is to make one or two trades a week when all my trading conditions are
satisfied and to endeavor towards an 80% success rate for my trades. Since I make very few trades that
are generally short-term (within 30 minute), I trade with a lot size of 3 standard lots (6% of initial
amount of $100,000). I use a trend following trading strategy, meaning that I buy when the market is
moving upwards and sell when the market moves downwards, making trades only when my confidence
in the trend is very strong. My goal is to achieve an average profit of 10 pips on all trades (not just
profitable trades) with a stop loss of 15 pips Using these goals, I am able to project a yearly projected
profit would be around 12%. Thus far, during this project, I have made a profit of $2835, which would
yield a yearly profit of $11,340 so my achieved profit is very close to the profits estimated by my trading
goal.

63

My trade strategy is outlined below:

1.

Each day that I trade, I make sure that I am up-to-date with all fundamental issues that may
affect the value of the currencies that I plan to invest; I read the Gartman Letter, Forex
Factory, and also read the world news on Bing news, reading the most important world
events as well as searching the currencies that I plan on trading, usually the US Dollar, the
Euro, the Australian Dollar, and the Japanese Yen. The Gartman letter helps to provide me
with some intuition on what may be influencing the chart trends. When there are very
important news events that may cause higher-than-average market volatility, I avoid trading
as my trading strategy is based upon technical rather than fundamental analysis.

2.

The trading plan that I use is based upon a setup and a trigger. The setup is a particular set
of conditions on the chart under which I may consider to make a trade on a given currency
pair. The trigger is a particular event upon seeing which I execute the trade.
a. When I first open MQL4, I look up the currencies that I am considering trading given the
fundamental information that I have gathered. These pairs include EURUSD, AUDUSD,
USDJPY, and CADUSD. The spread on these pairs is quite low, which is very important
given that I generally plan to close my trade within 30 minutes.
b. The setup conditions for a buy is when a price falls to a support line and increases with
two or three consecutive bars, the MACD is positive or it has a very high derivative and
the simple moving average shows a positive trend. The setup conditions for a sell is
when a price increases to a resistance line two or three times, the MACD is negative or
has a very high derivative, and the simple moving average shows a negative trend. If it
moves in the opposite direction for a bar or two, I also look at the Parabolic SAR. Other
specific patterns that I try to identify are the double top and the double bottom patterns
which signal that the price may fall through or rise above the base value.
64

c. Before trading, I look at the plots of the prices as well as support and resistance lines
over using the 1-minute chart, the 5-minute chart, the 15-minute chart and the 30minute chart for the last day and see important support and resistance points and try to
identify support and resistance lines as well as the general overarching trend. I will also
determine whether the currency pair follows trends or goes against the trend. If there is
no coherent trend that I can follow, I avoid trading; if the prices generally seem to be
trend following, then I consider the trade further.
d. When I identify the setup that is good for trading, I then look for the trigger at which I
decide to actually execute the trade. The trigger is usually a crossover between the 9period and 21-period exponential moving average identified by the MACD indicator, the
double-top or double-bottom pattern, or a very strong trend that I determine will
continue to follow and will generate the target profit of 15 pips.
e. The exit strategy is based on support-and-resistance lines as well as stop-losses and
take-profits. When we are buying and we hit a resistance line that is shown by our
support-and-resistance indicator and we also notice that the price movement is slowing
or reversing, we exit the trade. Similarly, if we are selling and we hit a support line that
is strong and the price movement slows or reverses, we exit the trade. If there is
consolidation for an extended period of time and very limited price movement occurs,
we try to make sufficient pips to overcome the loss incurred by the spread and then
close the trade. This is because under periods of consolidation, we are highly unsure
about whether the support or resistance will eventually prevail. Finally, if a stop-loss of
15 pips is incurred, then the trade is closed. I do not explicitly specify take-profit
although if I lose around 30% or more of the profit on a trade, I close the trade. This exit
strategy works well to maximize gains and minimize losses.
65

When I began trading, I started with lot sizes of one or two lots. However, towards the end of
the project, I started buying or selling three standard lots of any given currency pair. The larger lot sizes
enable me to focus more on identifying optimal setup conditions rather than requiring me to make
many more trades in order to achieve my profit goals. However, if fewer than three of the setup
conditions have been met, I plan to trade with two lots, though this has not yet happened since I simply
do not trade under such conditions. If a trend is extremely profitable, meaning that it returns more than
25 pips of profit and if I can see the trend continuing, then I may add two more lots of the same pair,
though under no circumstances do I plan to trade more than 5 standard lots. Because my trades are very
short-term and I plan to exit within 30 minutes, I also do not trade multiple currency pairs at the same
time, allowing me to devote my entire undivided attention to the current pair I am trading.
The results of my trading strategy seem to be quite solid; as a result of continued refinement of
my plan, I eventually increased my stop-loss from 10 pips to 15 pips so I can catch some of the larger
price movements in the chart. This strategy worked quite well, helping me achieve a profit of 47.3 pips,
which over 3 lots is well over $1000 and can offset several negative trades at the stop loss. Trailing stop
losses and buy-limits and sell-limits are also an important addition to my plan that I plan to implement if
I continue trading at the professional level. A limit order is an order to buy a security at no more than a
specific price, or to sell a security at no less than a specific price; this can be useful if you are anticipating
the currency to rise above or fall below a certain value but do not wish to stay on the terminal waiting
for the conditions to occur. The disadvantage is that it is possible that these conditions may never occur,
but if they do, the profits may be very high. When these are combined with stop losses, trailing stop, or
variable stops, we have higher flexibility while still enforcing robust risk management.

66

Below we have included a sample trade where I executed my trading plan, buying 3 standard
lots of AUDUSD on 2011/11/17 at 05:35 for 1.00898 and selling on 2011/11/17 at 05:41 for 1.01073,
achieving a profit of 17.5 pips or $525.

Sell

Double-Bottom Pattern

Buy

Figure 26 Srinivass Second Trade

The decision to buy was based on the MACD, support and resistance lines, and the simple
moving average. The setup of the buy was a double-bottom pattern going downward with the price
trending back upward. We also noticed the price hitting a distinct support line. We executed the market
order when we saw that the price moved upwards for two consecutive bars and that a MACD crossing
between the 9-period and 25-period EMA occurred. The 9-period simple moving average, denoted by
the red line in the chart, also moved upward. We bought 3 standard lots with a stop loss of 10 pips. The
trade made a profit of 17.5 pips or $525 dollars, which was quite good. We ultimately decided to sell
noting a resistance line at the top. The price continued to increase further upward even after the
resistance line although it later rebounded; nonetheless, we made a good profit and there was no way
to be sure that the price would not rebound downwards strongly.

67

3.1.4 Trade Strategy for Samuel Veilleux

Over the last term, my trade strategy has evolved to only make trades which follow support and
resistance lines in the market. In their purest form, these lines represent prices at which currency pairs
tend to reverse their movement. These can be caused by (among other things) a large financial
institutions decision that a wise place to buy sits at fixed (low) value, and they do so every time the
currency pair reaches it. Resultantly, when the exchange rate falls to that value, the available liquidity at
that price is exhausted by the large buying power of the financial institution. The price accordingly
moves back upward. This may happen several times at the same price. A vigilant trader who notices
that pattern of upward trend after the specific low value is reached can enter a long position the next
time it happens, and keep the trade active while the trend continues.
Of course, this pattern cannot continue indefinitely, as even the funds of a large financial
institution are not unlimited. Resultantly, a trade strategy which yielded profit on several iterations of
the same support line will eventually fail. This failure is due to a breakout which could have been
caused by any number of reasons, and is usually indicatory that the reversal line will no longer be useful
in predicting trend reversals.

Figure 27: Support and Resistance Lines [54]

Sometimes more than one line can be observed in a chart, in the form of an upper and lower
boundary. A support line is one which supports the price of a currency pair from the low creating
local minima in the chart. Conversely, a resistance line resists upward movement, and will lead to local
68

maxima in the chart. These boundaries can be formed for the same or different reasons, and one may
be stronger than the other. The absolute strength of a support or resistance line is representative of
the likeliness of a currency pair to bounce off from it, as a result of the presence of people waiting to
buy (or sell) at that line.
Unfortunately, the absolute strength of any line cannot be determined without insider trading
knowledge! Thus, the apparent strength of a particular line correlated to how many times a currency
pair has already bounced off from it. Further, the apparent strength of a line is also correlated to how
tightly the bounces fall to the extreme-point. For instance, at a line of psychological importance (such
as 1.00000 in USDCAD or 1.30000 in EURUSD) may see bounces executed by programs, causing the price
to hit exactly 1.00000 or fall within a few tenths of a pip from there. Conversely, a weaker support line
may look just as strong from a distant zoom level, but close examination may show that the variance
among the local minima is as great as 5 pip.
A historically strong support line, once broken, can become a resistance line. This occurs
because during the time that the line was acting as support, sellers waiting for the price to fall BELOW
the line had to wait. Thus, after the breakthrough, sellers who had been waiting can now sell. Some
lines of support and resistance can be followed tightly for a few days, then broken cleanly (by a highvolume purchase which exceeds the absolute strength of the line during that time), only to return a few
days later and be obeyed again. Anomalies such as these can obscure a traders judgment of the
strength of a given line, and can also fool robots programmed to identify these lines.
The methods which I have adopted to govern the way I make trades based off of support and
resistance make up my trade strategy. There are several modules to this strategy, some of which can be
used independently of the others. I first lay down some rules for any and all market-entrances.
Following that I outline different strategies for making trades based on support and resistance lines,
including break-outs from boxes and reversal prediction.
69

Firstly, no matter which strategy I am using, I always wait for the price movement to indicate to
me that my prediction is being followed. Using the example of a strong support line I expect to continue
being obeyed: if I watch the price fall down to it, I do not enter a long position immediately. It is my
strategy to wait until I can see the price beginning to reverse rather than entering too quickly only to
find that it continues falling. Of course this method is not fool-proof, and will not lead to %100
profitable trades. One way to error-check this method is to look back at other times that the line in
question has been met, and observe the noisy signal on a narrow candle-width chart as it turns around.
This gives me a rough idea of what to expect a similar turn-around to look like.
Similarly, if Im planning to enter a trade as soon as a notable line is broken, I will not do so until
the break-out has at least exceeded the noise of the previous reversals. Often I will include a window of
confidence for these decisions a vertical distance I expect the price to move by after meeting my setup conditions to give me assurance that my prediction is coming to fruition. However, even with well
tested care-measures in that respect, the market cannot be predicted. A large price movement can
happen at any time as the result of some event somewhere in the world, and if this movement is not in
my favor it could bring a hard hit to my account. Accordingly, I include another safety measure (namely
the placement of a stop-loss) in all my trades. The value of my stop-loss varies with which method I am
using, as is outlines below.
The simplest application of support and resistance lines is the identification of a single line which
has been obeyed nor more times (for my personal trading Ive found n=3 sufficient), and trade on the
next time it is obeyed. To do this, as outlined above, I look at the previous bounces and get an idea of
how much noise to expect during the bounce, and look at the range which contains the absolute
extrema. Considering the example of a support line (characterized by similar local minima in recent
history), when the minima are very tightly grouped (i.e. within one pip) this method can be
implemented by way of a simple buy-limit. Although my trade strategy dictates the requirement that I
70

see the trend following my prediction before I enter the market, earlier (more simplistic) spins proved
this to be a relatively safe way to earn some pips.
In the example of a strong support line with very tightly grouped minima, a breakthrough will
come with the very quick warning of the price moving through the line identified by the previous
reversal points. If every preceding reversal fell within 1 pip of a certain line, any slight breech in this line
quickly raises a flag to the trader. Because that breech can be identified after a two-pip drop below the
support line, a stop-loss can be placed extremely close to the buy price because noise is not expected
to cause the price to fall that low during the reversal. Given that the maximum loss incurred by placing a
buy-limit at the support line with a 4 pip stop-loss is very small, it is reasonable to let the simple, built-in
pending-order functions of my trading platform execute the trade.
My exit strategy for this method has changed a lot this term. Initially, it was my decision to
place a take-profit at a level determined by looking at the previous bounces from the support line, and
estimating where I could expect the price to reach with reasonably certainty (under the assumption that
the bounce-pattern would continue.) However, I noticed major flaws in this (simple) methodology,
because I expected the trades to be completed both entry and exit - without my supervision. Every
bounce is new and unique, and subject to a very large number of unpredictable factors. Sometimes I
would see that the price moved up into a profitable region, but not far enough up, only to fall back
down to the stop-loss. Other times, I would see that the trade was profitable, but that the take-profit
level fell in the middle of a very obvious up-trend, and more money could have easily been made.
Accordingly, I implemented a sliding stop-loss. This helped catch the first failure, because after
a reasonably movement into profitability the trade is guaranteed to yield profit (even if small, still far
more welcome than the stop-loss lurking below). Similarly, a trailing stop-loss continues to stay below a
winning trade as it goes farther and farther in the preferred direction, eliminating the dismay brought by
the premature action of a take-profit. Choosing a value of this trailing stop-loss is not something I have
71

quantified yet, but my most successful tries have resulted from looking at previous bounces from the
same reversal-line and getting an idea of what ranges would have encompassed the noise in the signal.
This method has the value of permitting the trader to recognize a pattern in any time-frame,
and place the pending-order on his or her own time. Every aspect of this trade can be chosen and
designed according to any specific risk-loss management plan, usually without any rush, long before the
trade goes is made. It should be noted that most (if not all) platforms allow the user to create an
expiration date for pending orders such as the buy-limit, which means that the order is deleted and
therefore never placed if the condition is not met after a certain period of time. It should be further
noted that, while some brokers have this time-period defaulted to infinity, others do not. This was one
important lesson learned by my group when we found that a pending order expired and prevented the
placement of what was to be a very profitable trade.
An example of a trade I made this term following this strategy is shown in annotated
screenshots from ForeX Capital Markets (FXCM), below.

Figure 28: Setup for a Trade Based on One Resistance Line

As shown in the figure above, a resistance line was identified at 1.38438 (top, blue). Because it was
tested 5 times and stayed strong, I decided to set a sell-limit and to graphically determine a stop-loss
72

and a take-profit based on previous market activity. As outlined in my trade strategy, I configured every
aspect of this trade long before it actually initiated. The next figure shows the execution of this trade on
a tighter zoom-level.

Figure 29: Execution of a Trade Based on One Resistance Line

The sell-limit was met 26 minutes after the placement of the pending-order, while I was in class.
After the sale was made, the price rebounded as expected from the resistance line. It met the takeprofit line after 8 minutes. The profit from this sale (which was in the volume of two standard lots) was
$214.00 (10.7 pip). It can be seen on the chart that the rebound continued past my take-profit (lowest
blue line), but not by much. In this particular case, the graphically determined take-profit demonstrated
its value however in my other trades I have found it to be an overall unreliable method, as discussed
on the previous page.
An addition to my support-and-resistance trade strategy was developed to implement a box of
buy and sell stops around the critical bounce levels (both above and below, around sideways market
movement). It operates in anticipation that when the limits are broken (and exceeded) by a certain
specific amount, the market will continue moving in that direction. It is my stipulation that this method
only be implemented if and only if both lines are identified and tested at least twice each. Buy-stop and
73

sell-stop levels are selected at 25% outside of the support-resistance range. Figure 30 shows the
equations which govern the calculation of these values, along with the stop-loss levels.

Figure 30: Choosing stop-levels and exit strategy for market moves based on support and resistance levels

The constants 2 and 4 in these equations were selected graphically when I first decided to
outline this method, and have proven profitable over the little hand-testing which I have implemented.
The math which governs this strategy is simple. Sincethis strategy dictates the creation of pending
orders, rather than a split-second decision and market-entry, its implementation can be achieved by
hand or by a program. The only difficult part doing so is determining the values of support and
resistance lines effectively. Doing this by hand can be tedious and requires the user to check the market
frequently. However, a program which can effectively determine these levels would be very complex
and does not yet exist (or at least, is not freely available to the public). One project by our group this
semester was the creation of one such program. Its usefulness varied over different candle-widths and
the program did not convince its creators that it could be a substitute for by-hand analysis.
A trade exemplifying this trade-strategy can be found on the following page. Our groups
Support and Resistance Plotter analyzed the AUDUSD and identified a resistance line at 1.00385 (yellow)
and support at the psychologically important 1.00000 (blue). Buy-stop and sell-stop levels are identified
by dotted green lines.

74

Figure 31: Setup for a Trade Based on a One Support and One Resistance Line

The stop-loss entered for the pending order was set to half way between two levels calculated
using my updated trade strategy. The trades were then selected to have a trailing stop after the market
order was made, whose value was equal to half the range between the support and resistance.

Figure 32: Execution of a Trade Based on a One Support and One Resistance Line

When the resistance line was broken, it was soon followed by the breaking of my buy-stop line.
This event is marked in the figure above by the first red circle over the blue horizontal line. My trailing
stop of 11.2 pips caused the trade to close at 1.00725 a profit of 22.9 pips. Because this was a trade I
made for my account while also in control of the group account, the lot-size was 4 standard lots. The
profit was therefore $916.
75

3.2 Business Plan

We are collectively managing $800,000.00 USD for Empowerment Capital, LLC; as this is a very
large sum, we work hard to make thoughtful trades that align with our risk management strategy and
profit goals. We employ the highest ethical standards and leverage our technical skills and discipline to
attract and satisfy our clients.
3.2.1 Profit Allocation and Customer Relations

Overall goal of generating 11% profit per year

Use stop losses to avoid large losses, preserving investor confidence

Make reasonable profits for Empowerment Capital by charging the clients a 2% management
fee per annum, and a 20% performance fee per month

Empower society by donating 5% of our proceeds to charity

Provide clients with the other 75% of our profit

Provide investors with weekly performance reports on their investment, detailing all the specific
trades as well as the overall profit

Give customers security by facilitating transparency in our operation, allowing money to be

withdrawn in part or in full at any time

3.2.2 Trade Logistics and Money/Risk Management for our Group

Our account will be leveraged 50:1, and our average trade will be for 2% of the total account
capitol. Additionally, at any time there will be at most four trades open. Thereby, we never sink more
than 8% of our total assets into investments. With the average trade being 2% of our initial investment,
trades in our personal accounts will be in the volume of one standard lot. In the group account,
individual trade size is 4 standard lots. Thus, to reach our groups goal of 11% annual yield, we need
only to average (with a 250 day trading year estimate) 0.044% per day.
76

Trading 2% of the account at a time, with a leverage of 50:1, our groups account should show a profit

It should be noted that the calculation above does not include the effects of compounding, as it assumes
that the trade size will be 2% of the initial capital through the entire year of trading.
If any days trading activity nets more than three losses, our days trades will cease. We will stay
up-to-date on fundamental developments in the worlds economy, and how they are likely to affect the
foreign-exchange market. To do this we will read regularly The Gartman Letters, and keep updated on
some of the forex-news sites which we have come to trust. Although fundamental analysis is important,
particularly in terms of the impact that news can have on the immediate future of the market, the force
which drives our groups trades is that of technical analysis.
We typically use three different technical indicators to analyze the market. EMAs and the MACD
are used heavily, as well as support and resistance level identification. We never make a trade without a
stop-loss. The two currency pairs which we trade most frequently are EURUSD and AUDUSD mostly
because of their high refresh-rate and low spreads. Specific details of our group members individual
trading strategies can be found in Section 3.1 of this document.
Over the past two terms we have built strong familiarity with the use of the MQL4 programming
language in MT4. Happy with the processing power and response-time of our PCs trade abilities in
comparison to our own, we intend to continue improving our abilities with the language and adding
functionalities to the programs we have already written. Details of programming-related projects
completed by our group can be found in Appendix B.

77

Chapter 4: Trade Portfolio

4.1 Overall Performance Summary
Each team-members personal account of $100,000 USD, and the shared group account of
$300,000 USD, has been active since November 1st 2011. All trades were been documented each week
for record keeping and progress presentations. A summary of all account activity can be found in the
table below. The overall return on the virtual $700,000 investment was 1.835% ($12,843).
Table 1: Overall Trade Summary for Group 5

A performance chart showing every trade made by the group, including both individual trades
and trades made with the group account, is shown below. The y-axis is normalized over the $700,000 of
total capital available to these accounts. The slope of the linear trend-line added to the plot indicates
that the progression of the group capitol averaged 0.0145% per day, which can be extrapolated to
5.296% per year. While this investment return is less than our groups goal, it is still respectable and it is
our belief that it is sustainable. The Sharpe Ratio for this performance is 24.75.

78

Trades of All Accounts (Individual and Group)

2.00%
1.80%
1.60%
1.40%
1.20%
Profits
1.00%
(Percentage)
0.80%
0.60%
0.40%
0.20%
0.00%
24-Oct

y = 0.000145x - 5.914110
R = 0.816113

13-Nov

3-Dec

23-Dec

12-Jan

1-Feb

21-Feb

Date

Figure 33: Performance Chart for All Accounts

4.2

The Group Account

Trades made in the group account totaled a net profit of $2,918.42 USD (0.973% of initial

balance). In terms of basis points, 100.5 pips were gained in accounts favor. The average trade size was
2.17 standard lots. Trades in the group account were usually made when all group members could be
present, and a trade could be agreed upon before market-entry. Occasionally, when scheduling such a
meeting was too difficult, control of the group account was transferred to one group member who was
given the authority to make trades in the absence of the others. Resultantly, trades in the group
account sometimes mirror trades of individual members who made trades in both accounts
simultaneously. Trades of the group account are outlined in full below.

79

Table 2: Group Account Trades

The figure below shows a performance chart outlining the balance of this account over time,
with its y-axis normalized as percentages of the $300,000 initial balance.

Group Account Trades

3.50%
y = 0.00023x - 9.25732
R = 0.68080

3.00%
2.50%
Profits (Percentage)

2.00%
1.50%
1.00%
0.50%
0.00%
24-Oct

13-Nov

3-Dec

23-Dec

12-Jan

1-Feb

21-Feb

Date

Figure 34: Performance Chart for the Group Account

4.3 Adrian Delphias Account

Trades made in the Adrian Delphias account totaled a net profit of $2,911.00 USD (2.911% of
initial balance). An unabridged list of trades since November 1st 2011 is shown in the table below.

80

Table 3: Adrian Delphia's Trades

A performance chart indicating the profits in Adrians account over time is shown in the figure
below. The trend line on the plot indicates that profits averaged 2.8%% per day.

Adrian Delphia's Trades

4.00%
y = 0.00028x - 11.38267
R = 0.62796

3.50%
3.00%
2.50%
Profits (Percentage) 2.00%
1.50%
1.00%
0.50%
0.00%
24-Oct

13-Nov

3-Dec

23-Dec
Date

Figure 35: Performance Chart for Adrian Delphia

81

12-Jan

1-Feb

21-Feb

4.4 Brendan Hamms Account

Trades made in the Brendan Hamms account totaled a net profit of $3,320.07 USD (3.320% of
initial balance). An unabridged list of trades since November 1st 2011 is shown in the table below.
Table 4: Brendan Hamm's Trades

A performance chart indicating the profits in Brendans account over time is shown in the figure
below. The trend line on the plot indicates that profits averaged 3.5%% per day.

82

Brendan Hamm's Trades

4.00%
3.50%
3.00%
y = 0.00035x - 14.44215
2.50%
R = 0.66279
2.00%
Profits
1.50%
(Percentage)
1.00%
0.50%
0.00%
13-Nov
3-Dec
23-Dec
12-Jan
-0.50%
-1.00%
Date

1-Feb

21-Feb

Figure 36: Brendan Hamm's Trades

4.5 Srinivas Vasudevans Account

Trades made in the Srinivas Vasudevans account totaled a net profit of $2,397.00 USD (2.397% of initial
balance). An unabridged list of trades since November 1st 2011 is shown in the table below.
Table 5:Srinivas Vasudevan's Trades

A performance chart indicating the profits in Srinivass account over time is shown in the figure
below. The trend line on the plot indicates that profits averaged 2.3%% per day.

83

Srinivas Vasudevan's Trades

3.00%
2.50%
2.00%y = 0.00023x - 9.27253
R = 0.67585
Profits (Percentage) 1.50%
1.00%
0.50%
0.00%
24-Oct

13-Nov

3-Dec

23-Dec

12-Jan

1-Feb

Date

Figure 37: Srinivas Vasudevan's Trades

4.6 Samuel Veilleuxs Account

Trades made in the Samuel Veilleuxs account totaled a net profit of $1,297.00 USD (1.297% of initial
balance). An unabridged list of trades since November 1st 2011 is shown in the table below.
Table 6: Samuel Veilleux's Trades

84

A performance chart indicating the profits in Samuels account over time is shown in the figure below.
The trend line on the plot indicates that profits averaged 1.0%% per day.

Sam Veilleux's Trades

1.60%
1.40%
1.20%

y = 0.000100x - 4.066572
R = 0.537095

1.00%
Profits (Percentage) 0.80%
0.60%
0.40%
0.20%
0.00%
24-Oct

13-Nov

3-Dec

23-Dec
Date

Figure 38: Samuel Veilleux's Trades

85

12-Jan

1-Feb

21-Feb

Chapter 5: Parameters for Launching Our Company

5.1 Company Structure
Our team plans to launch a foreign exchange money management firm Oakwood
Investments. Oakwood Investments will be a hedge fund that will provide investors with an
expected return of 20 percent per year. Our strategy will employ complex risk management
strategies and will leverage the algorithmic and trading experience of our experienced team.
The firm will be structured as a limited liability corporation. We plan to run effective marketing
campaigns as well as recruiting wealthy investors during the next several years. Additionally, we
plan to establish a legal plan to protect our company and its assets.
The LLC structure has many advantages, which make it desirable for our trading firm.
We will have the liability protection of a corporation meaning that the partners in our
corporation cannot be held personally liable for debts unless they have signed a personal
guarantee. Furthermore, our leaders have the liability protection of a corporation. An LLC exists
as a separate entity much like a corporation; members cannot be held personally liable for
debts unless they have signed a personal guarantee. Unlike a partnership, we are free to select
various forms of distributing profits and are not bound to distribute returns evenly. This way,
our system can reward those who earn the highest returns, effectively creating an incentive
plan to attract adept traders. We also plan to file articles of organization with the Secretary of
State and hire a law firm to design an operating agreement.
The company will be funded by performance fees as well as management fees from
investors. Oakwood Investments will charge investors a flat 2% management fee per year to
cover the operating costs of the company as well as a flat performance fee of 20% for eligible
86

accounts. We plan to employ several trading strategies based on statistical signal processing,
advanced entry/exit triggers, and algorithm selection. Utilizing these several trading strategies
we can offer our clients several different packages with varying risks and returns. In this
manner we can attract investors with varying levels of risk management.
Our corporation will launch an extensive marketing campaign to seek private
investments by venture capitalists and wealthy investors who can extend our name and brand
recognition. We seek to manage several million dollars within the next few years, with
intentions of expanding once we establish a positive reputation. We also have start-up costs
and need to pay outside consultants, legal fees, and salaries to our employees; this means that
for our first few years we plan on reinvesting the majority of our profits to help attain positive
growth.

5.2 Findings Regarding Effective Marketing

In order to achieve and maintain profitability, it is imperative that we launch an
effective marketing campaign. There are several restrictions on hedge fund advertising and
marketing. Due to the broad mandates and relatively lenient registration and disclosure rules,
hedge funds in the United States are only allowed to accept investments from accredited
investors and institutions. Therefore, to gain visibility, we plan to use search engine
optimization to ensure that our hedge fund is returned near the top of prominent search
engines. One of the ways in which we can market the company is by speaking at conferences
and events within the industry.
Developing a brand image for Oakwood investments will be integral in earning the trust
of investors. We operate onshore and thus incur some regulation by the United States
87

government. We also plan to use external consultants where appropriate. These consultants
will be experts within raising capital within a specific channel, creating marketing materials and
promoting a marketing message. We plan to select carefully for consultants who we can trust
and who are effective at marketing hedge funds.
To start our company, we need seed capital. This capital will come from high net worth
individuals who are familiar with the ability of our team as well as our wide and comprehensive
portfolio management experience. Our excellent trading records and our principles of integrity
should allow us to attract many serious investors. Additionally, to startup we will attract
investments from family and friends whom are accredited investors, as well as faculty and staff
at Worcester Polytechnic Institute and other companies with which we have affiliations.

88

Works Cited
1. Brian Dolan and GalantMark.Currency Trading for Dummies.John Wiley and Sons, 2011.
2. Bank for International Settlements. Foreign Exchange and derivatives market activity in April 2010Preliminary Results, 2010.
3. Logan, Tina. Getting Started in Candlestick Charting. Hoboken, NJ: John Wiley & Sons, 2008.
4. "Moving Average." Wikipedia, the Free Encyclopedia. Web. 27 Oct. 2011.
<http://en.wikipedia.org/wiki/Moving_average>.
5."MACD." Wikipedia, the Free Encyclopedia. Web. 27 Oct. 2011. <http://en.wikipedia.org/wiki/MACD>.
6."Commodity Channel Index (CCI) - ChartSchool - StockCharts.com." StockCharts.com - Simply the
Web's Best Financial Charts.Web. 27 Oct. 2011.
<http://stockcharts.com/education/IndicatorAnalysis/indic_CCI.html>.
7."Bollinger Bands." Wikipedia, the Free Encyclopedia. Web. 27 Oct. 2011.
<http://en.wikipedia.org/wiki/Bollinger_Bands>.
8."Stochastic Indicator." Forex Indicators Guide.Web. 27 Oct. 2011. <http://forexindicators.net/stochastic>.
9."Commodity Channel Index (CCI) - ChartSchool - StockCharts.com." StockCharts.com - Simply the
Web's Best Financial Charts.Web. 27 Oct. 2011.
<http://stockcharts.com/education/IndicatorAnalysis/indic_CCI.html>.
10."Average True Range." Wikipedia, the Free Encyclopedia. Web. 27 Oct. 2011.
<http://en.wikipedia.org/wiki/Average_True_Range>.
11."Average Directional Index." Wikipedia, the Free Encyclopedia. Web. 27 Oct. 2011.
<http://en.wikipedia.org/wiki/Average_Directional_Index>.
12. "Parabolic SAR." Wikipedia, the Free Encyclopedia. Web. 27 Oct. 2011.
<http://en.wikipedia.org/wiki/Parabolic_SAR>.
89

13. "Relative Strength Index (RSI) - ChartSchool - StockCharts.com." StockCharts.com - Simply the Web's
Best Financial Charts.Web. 27 Oct. 2011.
<http://stockcharts.com/education/IndicatorAnalysis/indic_RSI.html>.
14."Forex Fibonacci Indicator Explained | Forex Indicator Guide." Best Forex Indicator - Forex Indicators Forex Training. Web. 27 Oct. 2011. <http://www.forexindicator.org/forex-fibonacci-indicatorexplained.html>.
15. Governing Board of the Swiss National Bank. Swiss National Bank Quarterly Bulletin,
29(Q3).
16. Dennis Gartman. September 7th. The Gartman Letter, 2011.
17.How to trade the Fibonacci retracement indicator. Indicator Forex LLC.Web.20 Oct 2011
<http://www.indicatorforex.com/content/how-trade-fibonacci-retracement-indicator>.2011.
18. Moving average convergence divergence. MetaQuotes Software.Web.22 Oct.
2011<http://codebase.mql4.com/258>, 2000.
19. "Is Social Security a Ponzi Scheme? - Room for Debate - NYTimes.com."The New York Times Breaking News, World News & Multimedia. 09 Sept. 2011. Web. 27 Oct. 2011.
<http://www.nytimes.com/roomfordebate/2011/09/09/is-social-security-a-ponzi-scheme>.
20. Parabolic stop and reverse. MetaQuotes Software. Web.22 Oct. 2011
<http://ta.mql4.com/indicators/trends/parabolic_sar>, 2000.
21. Censky, Annalyn. "Federal Reserve Launches Operation Twist - Sep. 21, 2011." CNNMoney Business, Financial and Personal Finance News. 21 Sept. 2011. Web. 27 Oct. 2011.
<http://money.cnn.com/2011/09/21/news/economy/federal_reserve_operation_twist/index.ht
m>.

90

22. Michael A. S. Joyce, Ana Lasaosa, Ibrahim Stevens, and Matthew Tong. The _nan-cial market impact
of quantitative easing in the u.k. International Journal of Central Banking, 7(3), page = 113-161),
2011.
23. The federal reserve system: Purposes and functions.
http://www.federalreserve.gov/pf/pdf/pf 4.pdf, 2005.
24. Force index indicator. MetaQuotes Software Web. 27 Oct. 2011. <http://codebase.mql4.com/299>,
2000.
25. Forex Factory. Web. 13 Jan. 2012. <http://www.forexfactory.com/news.php?do=news>.
26. Map of euro area 1999-2011. http://www.ecb.int/euro/intro/html/map.en.html, 2011.
27. Dennis Gartman. September 9th. The Gartman Letter, 2011.
28. Dennis Gartman. September 16th. The Gartman Letter, 2011.
29. William L. Watts. German lawmakers approve bigger efsf.
http://www.marketwatch.com/story/german-lawmakers-approve-bigger-efsf-201109-29, 2011.
30. Alexandra Hudson. ESFS could make germany liable for 465 billion euros: Ifo, September 2011.
31. "Crisis in Europe: Prepare for Repercussions from Standard & Poor's Credit Rating Downgrades Money Morning." Investment News: Money Morning - Only the News You Can Profit From.
Web. 05 Jan. 2012. <http://moneymorning.com/2012/01/13/crisis-in-europe-prepare-forrepercussions-from-standard-poors-credit-rating-downgrades/>.
32. All About...The Foreign Exchange Market in the United States. Federal Reserve Bank of New York,
1998.
33. John Williamson. The exchnage rate system.InPolicy Analyses in International Economics5, 1983.
34. Guide to Financial Markets. Exmouth House, 4 edition, 2006.
35. U.S. foreign exchange intervention. http://www.ny.frb.org/aboutthefed/fedpoint/fed44.html, 2007.
91

36. Frederic S. Mishkin. Inflation targeting in emerging market countries. National Journal of Economic
Research, (7618), March 2000.
37. Arturo Estrella and Frederic S. Mishkin.The yield curve as a predictor of US. recessions. Current Issues
in Economics and Finance, 2(7).
38. Q.F. Akram and O. Eitrheim.Flexible inflation targeting and financial stability. Journal of Banking and
Finance, 32(7), page = 1242-54), 2008.
39. Chris McMahon. Forex fundamentals: moving the markets. Futures Magazine Group, January 2007.
40. The Economic Consequences of Rising U.S. Government Debt: Priveleges at Risk, Departmental
Working Papers, Department of Economics, UC Santa Barbara, 2010.
41. Robert Wade. The asian debt-and-development crisis of 1997-?: Causes and consequences. World
Development, 26(8):1535{1553, 1998.
42. "CPI - Consumer Price Index - Forex Trading | MetaTrader Indicators and Expert Advisors Fundamental Analysis." Forex Trading | MetaTrader Indicators and Expert Advisors - Home.
Web. 26 Oct. 2011. <http://www.fxfisherman.com/forex/fundamental-analysis/a135-cpiconsumer-price-index/>.
43. "An Introduction to the Financial Markets." About.com US Economy.Web. 01 Mar. 2012.
<http://useconomy.about.com/od/themarkets/a/capital_markets.htm>.
44. "Tutorial: Stock Basics Tutorial." Investopedia.Web. 01 Feb. 2012.
<http://www.investopedia.com/university/stocks/>.
45. Madura, Jeff. Financial Markets and Institutions. Mason, OH: South-Western/Thomson Learning,
2003.
46. "Mutual Funds." About.com. Web. 20 Feb. 2012. <http://mutualfunds.about.com/>.
47. "Bond Markets & Prices." Bond Markets Defined. Web. 19 Feb. 2012.
<http://www.investinginbonds.com/marketataglance.asp?catid=31>.

92

48. Faerber, Esme, and EsmeFaerber. All about Bonds and Bond Mutual Funds: The Easy Way to Get
Started. New York: McGraw-Hill, 2000.
49. Galant, Mark, and Brian Dolan. Currency Trading for Dummies. Hoboken, NJ: Wiley Pub., 2007.
50. Bouchentouf, Amine. Commodities for Dummies. Indianapolis, IN: Wiley Pub., 2007.
51. Investopedia. Web. 05 Nov. 2011. <http://www.investopedia.com/>.
52. "Forex Indicator Guide." Best Forex Indicator. Web. 05 Nov. 2011. <http://www.forexindicator.org/>.
53. "Forex | Forex Trading | Forex News | Currency Trading." Forex.Web. 05 Mar. 2012.
<http://www.forexrazor.com/>.
54. "EDUCATION." Education Courses from ALGOSYS. Web. 05 Mar. 2012.
<http://algosys8.com/index.php/education/algosys_education/technical_analysis>.
55. Malcolm, Andrew. "Keystone XL Pipeline And Jobs Put Up Or Shut Up." News.investors.com.
Investors Business Daily. Web. 15 Feb. 2012.
<http://news.investors.com/article/597853/201201131851/chamber-of-commerce-boostskeystone-xl-pipeline.htm>

93

Appendix A: Recent Macroeconomic News

A.1 The Pegging of the Swiss Franc
Throughout the course of this year the Swiss National Bank (SNB), Switzerland's central bank,
had been unable to control the ever growing value of the franc (CHF). The effect of an ever growing
franc worked to deflate the currency, thus by proxy reduced the nation's ability to export goods and
fueled foreign investment in the nation. Such a trend was beginning to increase the nation's
unemployment. On September 6, 2011 the SNB announced that it would "peg" the CHF to the EUR in an
attempt to, "reduce the deflationary development that spring from massive overvaluation of the Swiss
franc." [15] The SNB set a target that the EUR/CHF should not be allowed to cross 1.200 and remarkably
this target has not only held since its inception, but has been surpassed reaching highs of about 1.24.
Within an hour since the decision was announced by the SNB the EUR/CHF shot up
approximately 1,000 pips. As Gartman puts it, "no where have we read of, been involved in, or even
considered the notion that a strong currency regime will tie itself, willingly, to that of a lesser, weaker
currency regime never ever." [16] This decision effectively ended the use of the CHF as a currency to
be used in international transactions and acts to increase the usage of the other largest currencies: the
British pound (GBP), the USD, and the Japanese Yen (JPY).
Pegging one currency to another is not an uncommon practice, in fact many nations around the
world peg their currencies to the USD. However, all of the nations which do this have a weaker economy
and partake in this practice as it alleviates the responsibilities of their central banks while still
maintaining a specific monetary goal. This was not a true definition of pegging though since the Euro
and Franc arent exactly equal all the time. However, this essentially kept the euro and franc at the
same value, due to the banks buying power to force this price. They will buy it at a minimum of 1.20 but
nothing less. Anything more is fine for them because they want the value of the CHF to decrease. This is
very risky for Switzerland because they are buying a large volume of other foreign currencies, and they
A1

will likely sell them at a loss eventually. With the current crisis in the euro zone markets, only time will
tell if such a move will have the long-term desired results.

Figure 39: This is the CHFEUR value during the pegging. Notice the massive drop instantaneously and how it isnt really a
peg since the SNB agreed to buy it at any lower price, and it isnt following the Euros value precisely.

Gartman believes that the president of the Swiss National Bank was trying to make up for his
previous 15 billion dollars in damage to their economy previously, and to do that he is taking huge risks.
An immediate result of devaluing the franc to the Euro caused the Euro to fall against most other
currencies. The national bank of Switzerland is promising a global weakening of their currency and said
they had to take extreme measures to do so, and was prepared to weaken it more if needed. The move
does make some economic sense because their currency was so strong their exports increased in cost
and lowered their tourism business.[15] The bank previously tried to weaken the currency by promising
near-zero interest rates and increasing the supply of the franc, but this had no effect. The dramatic
intervention was deemed necessary by the Swiss bank. [15]
Despite its global criticism, the Swiss is still trading near the euro and the bank is actually
arrogantly spurring investors to buy the franc and sell the euro to break their relationship, but it holds
strong. This is a macroeconomic issue because it affects the regional economy of all of Switzerland. This
A2

move put the risk of the entire Swiss economy crashing and the ruin of their financial industry. The
market has more money than the Swiss Bank does, and as Gartman points out, they will not be able to
keep buying all currencies to keep the relationship at their desired weakness level.[16]

A.2 The Largest Ponzi Scheme Ever Played

An issue that is brought and discussed more and more seriously as time progresses, social security and
its inevitable collapse will eventually have to be dealt with for real. It has been no secret that there is
not enough money to keep paying retirees and the funds are going dry. Older generations demand more
from it, and younger generations refuse to put their hard earned wages into knowing they will never get
it back. In Dennis Gartmans daily publication, the United States social security policy is alluded as a
Ponzi scheme numerous times by a variety of people.[16] People still pay into it and that money goes to
older people who paid into it first. Too much is being distributed that it has been deemed impossible to
sustain later generations. It sounds exactly like a Ponzi scheme except one is forced to pay into it.
Usually people enter Ponzi schemes on their own choice and free will, a position that both Gartman and
Dr. Krugman, an extreme leftist agree with. [16]
President Obama temporarily cut social security spending around the middle of September of
this year. This affected the entire economy of the United States because many older generations reacted
nervously to the news and likely would be prompted to spend less to save their money in the fear of
starvation by not being able to afford food. Gartman describes why this is a bad idea for a different
reason because rather than just saving this money it should be invested so we can pay people what they
deserve. [19]
Speaking of what people deserve, the first person to ever collect social security from the
government was Ida Fuller. She paid less than $25 into the system and received a modest 92,000%
return. That is no decimal point; she collected around $23,000 until she died, reaching triple digits in
age. No reasonable person would think she deserved this. However, the policy was new and it was still
A3

working out the kinks and loop holes right? A more recent policy enables a young lady who has married
an older gentleman to collect for the rest of her life if her spouse happens to perish. Is it reasonable that
she is entitled for that money? Her husband worked his whole life for it while shes barely started.
Governor Perry notices the idiocy of this concept and is one of the few emerging politicians who is
speaking strongly and negatively about social security. This of course guarantees the loss of the votes
from the older generation who fear not getting enough from the security, but strengthens his loyalties
from the younger generation who dont want to pay in anymore. His upcoming election will be
interesting for this split vote reasoning.[16]
Social Security has so many flaws, and the fact that people are forced to pay into it, and it has
been acknowledged they likely wont be able to collect when they retire is robbery. Social security has
left uneasy marks since its induction so many years ago. For instance, pre 1972 teach card was printed
with a phrase that said how the card is not to be used as a form of identification, and it was never
supposed to be. Suddenly that piece of information was taken off the card for design changes and now
its a massive tool for identification purposes. Social Security is a very prominent macroeconomic issue
in the United States that is here to stay. The closer and closer the funds get to being completely
depleted, the more openly it will be discussed. It affects the entire US economy tremendously, by
discouraging workers who fear they will never collect and scaring current collectors into not spending
any money except on essentials. It is embedded into our economy, and will cause a lot more major
problems as time continues and funds deplete.

A.3 United States Federal Bank Twisting the Yield Curve

This policy of asset purchases has come to be known as quantitative easing (QE). In general
terms, QE is normally defined as a policy that expands the central banks balance sheet in order to
increase the level of central bank money in the economy. This is sometimes contrasted with a policy of
changing the composition of the assets on the central banks balance sheet (often referred to as credit
A4

easing); for example, by shifting between short and longer maturity government bonds or by shifting
into riskier private assets, such as corporate bonds or equities. [22]
The yield curve is the relation between the interest rate of a loan, and the time to maturity or
term of the debt. [21] To start this process, the Federal Bank sold off most off their bonds (hundreds of
billions worth) and purchased long-term bonds with the money made. Most bonds they sold were close
to maturing, most within 3 years. The point was to lower yields on long term bonds and simultaneously
leaving the short-term rates unaffected. All of this was cleverly done without pumping more money into
the economy, just reusing already existent revenue.
At the end of September the Federal Reserve announced that it would be buying long-term
Treasury securities in an attempt to drive down these long-term rates. In effect, the decision was put in
place to attempt to reduce mortgage rates in the U.S. and spur the otherwise declining sector of home
sales.
Short-term interest rates, such as those on Treasury bills and commercial paper, are affected
not only by the current level of the federal funds rate but also by expectations about the overnight
federal funds rate over the duration of the short-term contract. As a result, short-term interest rates
could decline if the Federal Reserve surprised market participants with a reduction in the federal funds
rate, or if unfolding events convinced participants that the Federal Reserve was going to be holding the
federal funds rate lower than had been anticipated. Similarly, short-term interest rates would increase if
the Federal Reserve surprised market participants by announcing an increase in the federal funds rate,
or if some event prompted market participants to believe that the Federal Reserve was going to be
holding the federal funds rate at higher levels than had been anticipated. [23]
According to the Federal Reserve, "changes in long-term interest rates also affect stock prices,
which can have a pronounced effect on household wealth. Investors try to keep their investment returns
on stocks in line with the return on bonds, after allowing for the greater riskiness of stocks." [23] This is
A5

essentially what the Federal Reserve is attempting to do with the "twisting" of the yield curve by buying
back long term treasuries. The Fed also notes that as the interest rates are raised the dollar will bid up in
currency markets, precisely what happened following the announcement by the FOMC.
The major criticism of the decision by the Fed is that increasing the interest rates of short term
bonds could adversely affect many of the smaller banks in the nation who rely on short term loans as
the basis of their operations. However, when the FOMC announced the "twisting", the USD immediately
saw a boom, swelling up over 200 pips vs. many other currencies.

A.4 The Grecian Default The Euro and PIIGS

The euro (EUR) is the first currency created to be the base currency of a conglomerate of nations
under control of one central bank known as the European Central Bank (ECB). Seventeen different
nations from the European Union currently use the euro as their currency; they are: Belgium, Germany,
Estonia, Ireland, Greece, Spain, France, Italy, Cyprus, Luxembourg, Malta, The Netherlands, Austria,
Portugal, Slovenia, Slovakia, and Finland. [26]
Since this is historically the first time a group of nations has decided to use one currency under
the control of a board overseen by the union there have been many unseen problems which have
arisen. Several treaties have been enacted by the member nations of the euro block in an attempt to
properly regulate the currency, including the treaties of Lisbon and Maastricht.
The treaty of Maastricht explicitly sets the levels of economic activity allowed by the ECB. Since
the treaty's induction in 2000 the euro zone has breached the deficit and debt limit on 137 different
occasions. [30] The failing of so many nations to adhere to the limits which they agreed to has created a
partition of euro countries known as PIIGS, which are the nations which have been causing the greatest
turmoil. Portugal, Italy, Ireland, Greece and Spain are the nations whose debt and policies has been
causing so much turmoil in Europe.

A6

Of all the nations of PIIGS, Greece perhaps is the most troublesome. Their debt to GDP ratio is
by far the worst in the union at 157% [27] and every measure which they take to alleviate that debt
seems to fail. Over the past few months government workers have been laid off in droves and still
Greece cannot bring its debt under control.
Italy is next on this list with 129% debt to GDP ratio. [27] In addition to being plagued with a
financial crisis in the EU, Prime Minister Berlusconi has been accused of handing out big government
contracts in exchange for sexual favors, with as many as eight women in one night. [28]
As can be seen by the countries above treaties governing the financial policies of several of the
euro countries have been broken repeatedly. The European System of Financial Supervisors (ESFS), was
developed to help settle disputes and keep nations compliant with the outlined regulations.[28]
However, the original inception of the ESFS saw little impact as its ability to enforce regulations was
limited. Prompted by the need to do something, in September, the member nations of the ECB voted
overwhelmingly to expand the powers of the ESFS in an attempt to stabilize the euro. [27]
The enhanced ESFF has been designed to head off a deeper crisis, by allowing it to provide
support to euro-zone countries. It would also effectively take over the European Central Banks bondbuying duties, allowing the fund to take action in markets to hold down bond yields. This is an effort to
further insulate vulnerable countries from being engulfed by the debt crisis. [29]
The heavy burden of the bailouts being issued has been place on the hands of the largest
economy in the euro-zone, Germany. Despite much opposition from the German public, Angela Merkel,
the Prime Minister of Germany, has approved many loans to the nations of the PIIGS. It has been
estimated by the Institute for Economic Research (IFO), a German economic think tank, that the
expected cost of further bailouts to Greece alone is 465 billion euro. [30]
It is widely speculated that despite efforts to inject money into Greece, their inability to
implement a successful fiscal policy will inevitably lead to default. Greece has defaulted on loans on
A7

many occasions and given the current path it seems only a matter of time until it defaults again. The
major fear is that if Greece defaults then so will Italy, Spain, and the rest will follow. If this trend
continues only time will tell if the euro remains a currency in the coming years.
Continued political conferences and debates have been constantly going on about what to do in
Europe. Other Eurozone countries have reported their own financial difficulties, brought on by the
decreasing value of the Euro. The Euro has been steadily falling over the past year which has resulted in
more than 1000 pip loss (relative to the US dollar). Clearly the world has put less faith into the Euro
because of this crisis. As more and more people lose faith in the Euro its value perpetually decreases.
People decide to invest their money in a stronger currency so they buy one, so for example the CHF. This
then creates more demand for the CHF and less for the Euro. As more and more people do this with
more and more euro currency its value keeps going down. The crisis in Greece is affecting every country
which uses the Euro as the main currency, and it is hurting them all. Greece is replacing high ranking
officials as to of no avail yet, but something needs to change to pull out of the dangerous crisis in
Europe. [25]

A.5 Selling your Debt Italy and Spain Auctions

A lot of currency in the form of Spanish debt was being auctioned off and fetched a surprisingly
generous bid. This was the first bullish appeal the Euro has seen in quite some time. Investors realized
that they had become too bearish of European debt, largely as a consequence of the enormous amount
of liquidity of the European Central Bank, which lent nearly half a trillion dollars of three-year bonds
with similar offers earlier. The Spanish treasury said it sold 5.6 billion ($7.3 billion), which is far greater
than the 4.5 billion it had initially sought. [27] Investors demanded an interest rate of only 1.74 percent
to lend the government 3-month money, down sharply from 5.1 percent in the last such auction onNov.
22. [31] The rate for the 6-month bills was 2.44 percent, down from 5.22 percent. [31] The buyers were
mostly domestic banks and domestic individuals in Spain, along with the ECB. He fact that domestic
A8

buyers are supporting this shows how faith really has been restored in the Euro a little, because they are
definitely well aware of the state that the Eurozone has been in recently.
During the days of January 12 and January 13, Italy also managed to achieve a successful bond
auction, though its auction was not as successful as Spains auction. Italy auctioned 12 billion euros on
January 12 and another 4.5 billion euros on January 13 in treasury bills as borrowing costs plunged in the
countrys first debt sale of the year. This also restored confidence to Europe because Italy was second in
debt leaders only to Greece. The treasuries of Greece, Ireland and Portugal sought bailouts because
their bond yield rates were above seven percent, which is extremely high. Italys success in lowering its
yield rates and being able to attract investors to buy large quantities of debt shows that confidence in
the Italian economy is improving.
The fundamental impacts of the debt sale was on a macroeconomic scale and mostly positive;
stocks rose and the price of oil fell. The Euro strengthened against other currencies since faith was
starting to be put back into the Euro. The euro added 0.9 percent after the sales and rose to most main
currencies. The Dow Jones Industrial average also strengthened 21.57 points to 12,471.02. [31] This
does not mean the crisis is over in any way though, in fact we are far from it This is just one of many first
baby steps. Considering the S&P recently downgraded the government debt of France, Austria, Italy and
Spain, problems clearly exist and such consecutive downgrades are a horrific thing for Europe. At least
Germanys debt remains at AAA ranking to all three international debt-rating companies.

A.6 Oil Pipeline

A project proposal for the construction of an international oil pipeline has been a buzz around
the world recently. Indeed, the Keystone XL project has found abundant support among some import
and influential groups in the United States. Some include: the International Brotherhood of Electrical
Workers, the International Union of Operating Engineers, the Teamsters, the Laborers' International
Union, the Building and Construction Trades Department of the AFL-CIO and the United Association of
A9

Plumbers and Pipe Fitters for the United States & Canada.[27] This makes sense because many jobs
would be created for these groups. Furthermore, U.S. Chamber of Commerce President Tom Donohue
has publicly endorsed the project to President Obama, prompted as a response to his recent
announcement to award companies which bring jobs to the United States.
Like every coin, there are two sides to this project. Some adamant opponents of the project
have environmental concerns, as the initial pipeline course brought oil over the Ogallala Aquifer in
Nebraska. A secondary route was outlined by TransCanada to mitigate these concerns. However, the
project is still not underway. President Obama has been criticized1 to be waiting for election time to
receive the good public appeal which would come from pushing this project through. At a press
conference following Donohue's address, Bruce Josten, chamber vice president of government affairs,
noted that Keystone XL had already cleared an extensive three-year review and that only presidential
politics stood in the way. This is a classic example of a macroeconomic issue because it is affecting the
entire world at the same time, making way for all these debates and can result in both good and bad
things for many economies. [55]

A10

Appendix B: Programming in the Forex Market

B.1 Abstract
Programming is a useful way to make life simpler, faster, and more efficient in everyday tasks.
When applied to the foreign currency exchange market it can be used to create powerful tools that can
lead to high gains in some cases. Whether it is the newest indicator that nobody else is using, or the
latest automated trading robot that makes 20% a week, programming can be useful. Moreover it can be
used to present the data in a different way which can be in turn used to see trends and make custom
indicators, such as extracting chart data to MATLAB and applying useful functions to it.
One project created a variety of programs over the course of the 3 term project. They range
from adding MATLAB integration to MQL4 to exporting TradeStation data to a C++ environment. Some
indicators and expert advisors are created such as a moving average (for practice) and a double top
alerter respectively. Other things were programmed to just make trading easier, such as a program that
automatically draws support and resistance lines to one that adds a trailing stop loss to all your trades
automatically. The bottom line is that programming can be used everywhere in trading securities, and
forex is certainly no exception.

B.2 Programming for MQL4

B.2.1 MATLAB Interface
Our newest script uses a DLL to interface MQL4 and Matlab. It gives users the ability to access
variables from code in both programs, and to call Matlab functions from inside an MQL4-based script,
indicator, or robot. Thus, advanced analytical tools can be used on the discrete-time signal that is the
Forex market. Powerful numerical analysis of past data can be performed quickly and in a platform with
which we are very familiar.

B1

Physical Code
1. // MatlabMql.cpp : Defines the exported functions for the DLL application.
2. //
3.
4. #include "stdafx.h"
5.
6. Engine *__ep=NULL;
7. mxArray*T =NULL;
8.
9. _DLLAPI int _stdcall initMatlabEngine()
10. {
11.

if(__ep==NULL){

12.

if(!(__ep=engOpen("\0"))){

13.

fprintf(stderr, "\nCan't start MATLAB engine\n");

14.

returnEXIT_FAILURE;

15.

}else{

16.

returnEXIT_SUCCESS;

17.

18.

19.

returnEXIT_SUCCESS;

20. }
21.
22. _DLLAPI void __stdcall evalExpression(char* text)
23. {
24.

engEvalString(__ep, text);

25. }
26.
27.
28. _DLLAPI int _stdcall exportDoubleVector(char*var_name, double* vector, int size){
29.

T =mxCreateDoubleMatrix(1, size, mxREAL);

30.

memcpy((void*)mxGetPr(T), (void*)vector, size);

31.

intretval=engPutVariable(__ep, var_name, T);

32.

mxDestroyArray(T);

33.

returnretval;

34. }
35.
36. _DLLAPI int _stdcall exportDouble(char*var_name, double value){
37.

T =mxCreateDoubleScalar(value);

38.

intretval=engPutVariable(__ep, var_name, T);

39.

mxDestroyArray(T);

B2

40.

returnretval;

41. }
42.
43. _DLLAPI int _stdcall exportIntegerVector(char*var_name, int* vector, int size){
44.

T =mxCreateDoubleMatrix(1, size, mxREAL);

45.
46.

double*double_vector=(double*)(malloc(size *sizeof(double)));

47.
48.

for(inti=0;i<size;i++){

49.
50.

double_vector[i]=(double) vector[i];
}

51.
52.

memcpy((void*)mxGetPr(T), (void*)vector, size);

53.

intretval=engPutVariable(__ep, var_name, T);

54.

mxDestroyArray(T);

55.

free(double_vector);

56.

returnretval;

57. }
58.
59. _DLLAPI int _stdcall exportInteger(char*var_name, int value){
60.

T =mxCreateDoubleScalar((double) value);

61.

intretval=engPutVariable(__ep, var_name, T);

62.

mxDestroyArray(T);

63.

returnretval;

64. }
65.
66.
67. _DLLAPI int _stdcall exportString(char*var_name, char*str){
68.

T =mxCreateString(str);

69.

intretval=engPutVariable(__ep, var_name, T);

70.

mxDestroyArray(T);

71.

returnretval;

72. }
73.
74. _DLLAPI void _stdcall closeEngine(){
75.

engClose(__ep);

76. }
77.
78. _DLLAPI int __stdcall engineSetVisible(){
79.

returnengSetVisible(__ep, true);

B3

80. }
81.
82. _DLLAPI int __stdcall engineSetInvisible(){
83.

returnengSetVisible(__ep, false);

84. }

B.2.3 Trailing Stop For Live trade

During the course of this project we found that often time constraints forced us to close open
positions before any signals occurred which warrant this action. For this reason it became necessary to
create a program which would determine if conditions warranted such a close of our position, allowing
us to leave orders open while we were not physically available to check for such conditions. For this
reason we created a simple trailing stop program.
This program utilizes a user defined stop loss level (SL), but is set by default to 15 pips. At every
incoming tick it cycles through all open orders of the current selected currency pair. During each of
these cycles it checks to see if the current price at which the stop loss is set to trigger is further than the
SL away from the current price (Ask or Bid depending on direction). If it is indeed further then the
program updates the stop loss trigger level and then cycles on to the next open order. In this manner,
every time a tick goes in favor of the current trade direction the stop loss is moved proportionally along
with it, while if a tick goes against a trade it does nothing.
Physical Code

1. //+------------------------------------------------------------------+
2. //|

TrailStop4LiveTRade.mq4 |

3. //|
4. //|

Group 5

http://www.metaquotes.net |

5. //+------------------------------------------------------------------+

B4

6. #property copyright "Group 5"

7. #property link

"http://www.metaquotes.net"

8.
9. externdoubleTrailingStop

=100.0;//SL in pips

10. externboolFiveDigBroker

=True;//true if 5 dig; false if 4 dig broker

11.
12. //+------------------------------------------------------------------+
13. //| expert initialization function

14. //+------------------------------------------------------------------+
15. intinit()
16.

17. //---18.
19. //---20. return(0);
21.

22. //+------------------------------------------------------------------+
23. //| expert deinitialization function

24. //+------------------------------------------------------------------+
25. intdeinit()
26.

27. //---28.
29. //---30. return(0);
31.

32. //+------------------------------------------------------------------+
33. //| expert start function

34. //+------------------------------------------------------------------+

B5

35. int start()

36.

37. //---38. //if (FiveDigBroker ==True) TrailingStop=TrailingStop*10.0; //adjusts the level of stops to a
5 dig broker
39.
40. int total=OrdersTotal();//makes sure our loop runs for every open trade for respective
currency pair
41. for(int count=0;count<total;count++)
42.

43. OrderSelect(count,SELECT_BY_POS,MODE_TRADES);//select by pos

44. Print("Order Selected");
45. if(OrderType()<=OP_SELL &&OrderSymbol()==Symbol())// OrderTypes are indexed: buy, sell, pend
buy limit, p b stop, p s.l., p s.s.
46.

47. Print(Symbol());
48. if(OrderType()==OP_BUY)//work with long positions (as stops will be in opp direction as
shorts)
49.

50. if(TrailingStop>0)//check to make sure we don\'t have -SL

51.

52. if(OrderStopLoss()<Bid-Point*TrailingStop)//look to see if TS needs to be updated

53.

54. Print("Should modify now?");

55. OrderModify(OrderTicket(),OrderOpenPrice(),BidPoint*TrailingStop,OrderTakeProfit(),0,Green);//change SL
56. return(0);
57.
58.

}
}

B6

59.

60. if(OrderType()==OP_SELL)//do same steps for short positions

61.

62. if(TrailingStop>0)
63.

64. if(OrderOpenPrice()-Ask>Point*TrailingStop)
65.

66. if(OrderStopLoss()>Ask+Point*TrailingStop)
67.

68. OrderModify(OrderTicket(),OrderOpenPrice(),Ask+Point*TrailingStop,OrderTakeProfit(),0,Red);
69. return(0);
70.

71.

72.

73.

74.

75.
76.

77. //---78. return(0);

79.

80. //+------------------------------------------------------------------+

B.2.4 Draw Entry Exit Program

Throughout the course of this project our groups had to make periodic presentations to the
other members of the project and our professor. While these presentations contained many pieces of
information they included the trades we made for the week. For this reason it became necessary to
display the individual trades we made and the underlying reasons for making them. The process of
B7

gathering screenshots of all of the weekly trades could be quite a laborious one and thus we decided to
write a program which would automatically draw directly on the chart all of the trades within a specified
time frame. A check-mark is placed at the entry position and a cross at the close position with a
connecting dotted red line. The prices are indicated at the dotted orange lines (which highlights price
on right). Short trades are indicated by a downward arrow, while long positions are indicated by upward
arrows.

Figure 40: A Screenshot of the Draw entry exit Program. Two short position trades are shown in this chart. (Green and
yellow lines are from a separate indicator).

Physical Code

1. //+------------------------------------------------------------------+

2. //|

DrawEntryExit.mq4 |

3. //|

Group 5 |

4. //|

http://www.metaquotes.net |

B8

5. //+------------------------------------------------------------------+

6. #property copyright "Group 5"

7. #property link

"http://www.metaquotes.net"

8.

9. #property indicator_chart_window

10. #property indicator_buffers 2

11. #property indicator_color1 Green

12. #property indicator_color2 Red

13. #property indicator_width1 2

14. #property indicator_width2 2

15. //---input parameters

16. extern color opencolor

=Orange;

17. extern color closecolor

=Orange;

18. externintopenwidth

=1;

19. externintclosewidth

=1;

20. externintopenstyle

=STYLE_DASH;

21. externintclosestyle

=STYLE_DASH;

22. externintorderbuysig

=SYMBOL_ARROWUP;

23. externintordersellsig

=SYMBOL_ARROWDOWN;

24. externintorderclosesig

=SYMBOL_STOPSIGN;

B9

25. externintorderopensig

=SYMBOL_CHECKSIGN;

26. externintorderclosewidth

=3;

27. externintorderopenwidth

=3;

28. //--- buffers

29. double ExtMapBuffer1[];

30. double ExtMapBuffer2[];

31. //+------------------------------------------------------------------+

32. //| expert initialization function

33. //+------------------------------------------------------------------+

34. intinit()

35.

36. //----

37. SetIndexStyle(0,DRAW_LINE);

38. SetIndexBuffer(0,ExtMapBuffer1);

39. SetIndexStyle(1,DRAW_LINE);

40. SetIndexBuffer(1,ExtMapBuffer2);

41. //----

42. return(0);

43.

44. //+------------------------------------------------------------------+

B10

45. //| expert deinitialization function

46. //+------------------------------------------------------------------+

47. intdeinit()

48.

49. //----

50.

51. //----

52. return(0);

53.

54. //+------------------------------------------------------------------+

55. //| expert start function

56. //+------------------------------------------------------------------+

57. int start()

58.

59. //----

60.

61. inttotalclosed

=OrdersHistoryTotal();

62. bool Opens

= FALSE;

63. bool Closes

= FALSE;

64. intCurrentOTime, CurrentCTime, CurrentOPos, CurrentCPos, run;

B11

65. double rise, RiseOverRun;

66. stringhlineopen, hlineclose, openarrow, closearrow, opentime, closetime, dots;

67.

68. for(int count=0;count<totalclosed; count++)

69.

70. OrderSelect(count, SELECT_BY_POS, MODE_HISTORY);

71. if(OrderSymbol()==Symbol())

72.

73. hlineopen

="open "+OrderOpenTime()+count;

74. hlineclose="close "+OrderOpenTime()+count;

75. opentime

="OpenTime "+OrderOpenTime()+count;

76. closetime

="CloseTime "+OrderOpenTime()+count;

77. dots

="dots"+OrderOpenTime()+count;

78. openarrow

="oarrow"

+OrderOpenTime()+count;

79. CurrentOTime=OrderOpenTime();

80. CurrentCTime=OrderCloseTime();

81. ObjectCreate(hlineopen,OBJ_HLINE, 0, 0, OrderOpenPrice());

82. ObjectCreate(hlineclose,OBJ_HLINE, 0, 0, OrderClosePrice());

83. ObjectSet(hlineopen, OBJPROP_COLOR, opencolor);

84. ObjectSet(hlineclose, OBJPROP_COLOR, closecolor);

B12

85. ObjectSet(hlineopen, OBJPROP_WIDTH, openwidth);

86. ObjectSet(hlineclose, OBJPROP_WIDTH, closewidth);

87. ObjectSet(hlineopen, OBJPROP_STYLE, openstyle);

88. ObjectSet(hlineclose, OBJPROP_STYLE, closestyle);

89. ObjectCreate(closetime, OBJ_ARROW,0,OrderCloseTime(), OrderClosePrice()+18*Point);

90. ObjectSet(closetime,OBJPROP_ARROWCODE, orderclosesig);

91. ObjectSet(closetime,OBJPROP_WIDTH, orderclosewidth);

92. ObjectCreate(openarrow,OBJ_ARROW, 0, OrderOpenTime(),OrderOpenPrice()+10*Point);

93. ObjectSet(openarrow,OBJPROP_ARROWCODE,orderopensig);

94. ObjectSet(openarrow,OBJPROP_WIDTH,3);

95.

ObjectCreate(dots,OBJ_TREND,0,OrderOpenTime(),OrderOpenPrice(),OrderCloseTime(),OrderCloseP

rice());

96. ObjectSet(dots,OBJPROP_STYLE,STYLE_DASH);

97. ObjectSet(dots,OBJPROP_RAY,0);

98. if(OrderType()==OP_BUY)

99.

100.

ObjectCreate(opentime, OBJ_ARROW, 0, OrderOpenTime(), OrderOpenPrice()-75.0*Point);

101.

ObjectSet(opentime,OBJPROP_ARROWCODE, orderbuysig);

102.

ObjectSet(opentime,OBJPROP_WIDTH, orderopenwidth);

103.

ObjectSet(dots,OBJPROP_COLOR,Green);

B13

104.

105.

if(OrderType()==OP_SELL)

106.

107.

ObjectCreate(opentime, OBJ_ARROW, 0, OrderOpenTime(), OrderOpenPrice()+90.0*Point);

108.

ObjectSet(opentime,OBJPROP_ARROWCODE, ordersellsig);

109.

ObjectSet(opentime,OBJPROP_WIDTH, orderopenwidth);

110.

ObjectSet(dots,OBJPROP_COLOR,Red);

111.

112.

113.

114.

//----

115.

return(0);

116.

117.

//+------------------------------------------------------------------+

B.2.5 Double Tops

The program DoubleTopsv3.mql4 is one which was created to recognize and trade the Double
Tops pattern. The Double Tops pattern can be summarized into 3 parts.
A. An upward trending pattern
B. A reversal to a downward trend creating a resistance line
C. A second reversal back to the resistance line
B14

Figure 41: Double Top Pattern contains 3 trend lines denoted A, B, and C

After these three parts have been exhibited the price can be expected to reverse yet again off of the
resistance line, else if it breaks this line it would be considered a Double Top Pullback. The Double
Bottom pattern is just the inverted version of this, in which case the price would be bouncing off of a
support line.
This program is set by default to look at the past 24 bars, and meant to be used on a 5-minute
chart, thus utilizing only data for the past 2 hours. This value of 24 was determined by data mining past
charts for occurrences of this pattern. It then stores both the highest and lowest points and waits until
the price begins to return to one of these levels. If the price approaches within 10% (user-adjustable) of
the difference between the high and low for this period then the program initiates a trade in the
respective direction. Only one trade is allowed to be open at a time, and a user defined trailing stop is
initialized (default of 15 pip).
This program was back tested using 5 minute data for the EURUSD downloaded from the
MetaQuotes history center. This program achieved some of the best results we saw out of any of our
B15

expert advisors. It has both a 57% profit ratio, as well as an 8.652% total profit over the year of 2011,
with a maximal drawdown under 1.8%. However, for reasons which remain unknown the program
achieved much better results for short positions than it did for long ones.

Figure 42:Results of the Double Top/Bottom program "DoubleTopsv3"

1. //+------------------------------------------------------------------+
2. //|

Double Tops.mq4 |

3. //|

Group 5

4. //|

http://www.metaquotes.net |

5. //+------------------------------------------------------------------+

B16

6. #property copyright "Group 5"

7. #property link

"http://www.metaquotes.net"

8.
9. externintperiod

24;

10. externdoubleReqDiff

10;

11. externdoubleTrailingStop=
12. externdouble SL

150;
=

50;

13. //+------------------------------------------------------------------+
14. //| expert initialization function

15. //+------------------------------------------------------------------+
16. intinit()
17.

18. //---19.
20. //---21. return(0);
22.

23. //+------------------------------------------------------------------+
24. //| expert deinitialization function

25. //+------------------------------------------------------------------+
26. intdeinit()

B17

27.

28. //---29.
30. //---31. return(0);
32.

33. //+------------------------------------------------------------------+
34. //| expert start function

35. //+------------------------------------------------------------------+
36. int start()
37.

38. //---39. intticket,volume,slippage,magic;

40. double highs[2];
41. double lows[2];
42. inthighspos[2];
43. intlowspos[2];
44. doublecurrenthigh, currentlow, TP;
45. volume =1;
46. slippage =3;
47. magic =10101;

B18

48. TP=15000;
49. int total=OrdersTotal();//makes sure our loop runs for every open trade for respective
currency pair
50. currenthigh=High[0];
51. currentlow=Low[0];
52. for(inti=0;i<period;i++)
53.

54. if(High[i]>currenthigh)
55.

56. currenthigh=High[i];
57. highs[1]=highs[0];
58. highspos[1]=highspos[0];
59. highs[0]=currenthigh;
60. highspos[0]=i;
61.

62. if(Low[i]<currentlow)
63.

64. currentlow=Low[i];
65. lows[1]=lows[0];
66. lowspos[1]=lowspos[0];
67. lows[0]=currentlow;

B19

68. lowspos[0]=i;
69.

70.

71. if(total<1)
72. {
73. if(highs[0]>highs[1]&& highs[0]-lows[0]>=ReqDiff*10.0*Point
&&highspos[0]>lowspos[0]&&highspos[1]<lowspos[0])
74.

75. if(High[0]>highs[0]-0.025*(highs[0]-lows[0]))
76.

77.

ticket=OrderSend(Symbol(),OP_SELL,volume,Bid,slippage,0,0,NULL,magic,0,Red);

78. OrderModify(ticket, OrderOpenPrice(),Ask+(SL*10*Point),Ask-TP*Point,0,Red);

79.
80.

}
}

81. if(lows[0]<lows[1]&& highs[0]-lows[0]>=ReqDiff*10*Point

&&lowspos[0]>highspos[0]&&lowspos[1]<highspos[0])
82.

83. if(Low[0]<lows[0]+0.025*(highs[0]-lows[0]))
84.
85.

{
ticket=OrderSend(Symbol(),OP_BUY,volume,Ask,slippage,0,0,NULL,magic,0,Green);

86. OrderModify(ticket, OrderOpenPrice(),Bid-(SL*10*Point),Bid+TP*Point,0,Green);

B20

87.

88.

89. }
90. for(int count=0;count<total;count++)//basic formula for a "for" loop
91.

92. OrderSelect(count,SELECT_BY_POS,MODE_TRADES);//select by pos

93. if(OrderType()<=OP_SELL &&OrderSymbol()==Symbol())// OrderTypes are indexed: buy, sell, pend
buy limit, p b stop, p s.l., p s.s.
94.

95. if(OrderType()==OP_BUY)//work with long positions (as stops will be in opp direction as
shorts)
96.

97. if(TrailingStop>0)//check to make sure we don\'t have -SL

98.

99. if(OrderStopLoss()<Bid-Point*TrailingStop)//look to see if TS needs to be updated

100.
101.

{
OrderModify(OrderTicket(),OrderOpenPrice(),Bid-

Point*TrailingStop,OrderTakeProfit(),0,Green);//change SL
102.

return(0);

103.
104.

}
}

B21

105.
106.

}
if(OrderType()==OP_SELL)//do same steps for short positions

107.
108.

{
if(TrailingStop>0)

109.

110.

if(OrderOpenPrice()-Ask>Point*TrailingStop)

111.

112.

if(OrderStopLoss()>Ask+Point*TrailingStop)

113.
114.

{
OrderModify(OrderTicket(),OrderOpenPrice(),Ask+Point*TrailingStop,OrderTakeProfit(),0,R

ed);
115.

return(0);

116.

117.

118.

119.

120.

121.

122.

//----

123.

return(0);

124.

B22

125.

//+------------------------------------------------------------------+

B.2.6 Support and Resistance Indicator: Identifying Critical Levels

First, a small fixed range is decided (ex: candles in window 10). In the example below there
are 6 candles in this range. That range is now a window which is stepped back through the candles
under analysis. Each time it is moved, the local extrema are determined. The value is discarded if it
appears at either end of the window because that can be indicatory of a trend, and the next windowlocation will pick it up if it is a true local extreme.

Figure 43: Sliding Window

After the slide is complete, all possibly (but not definitely) noteworthy local extrema have been
identified and stored. The figure below shows only the minima, for simplicity.

B23

Extrema are identified and stored in an array

Only minima are identified in this picture, for simplicity
Figure 44: Identified Extrema

Next, the price-values of those minima are stored in an array and sorted. Graphically, this is
equivalent to looking at all the vertical heights on one axis (as shown on the left side of the figure below)

Figure 45: Extrema Sorted

Now an algorithm must look through the sorted array of local minima and decide which are
important or noteworthy. Graphically, a group of lines close together (in vertical spacing) is more

B24

important than a group of lines which are farther apart. A simple way to implement this is to break up
the vertical scale into a series of buckets, and see which contain the most horizontal lines.

Figure 46: Applying Buckets

The second bucket from the bottom has the most minimums in it therefore we associate it
with a meaningful support line. The output of this function is a horizontal line located at the average of
the minimums from within that bucket. This is shown in the next figure.

Figure 47: Support Output

B25

While the previously described process is occurring, very similar lines of code are applying the
same processing technique to local maximum values. Buckets of similar nature are created, and local
maxima are determined. Currently, out algorithm plots lines for all buckets with three or more values in
them. However, this has the flaw of drawing a huge number of lines on the screen when a lot of candles
are being processed. A more sophisticated output might include plotting the three fullest buckets.
Demonstration of the Programs Functionality

Figure 48: The support and resistance lines drawn automatically using our expert advisor.

Note the yellow line indicates a resistance line and the blue line a support level. One can see
how the yellow line was placed due to catching so many peaks in its bucket described earlier, 4 on the
graph. The same can be seen for the blue support line. The program updates with every tick so it is
constantly drawing you the best 1, 2, or n number of support and resistance lines for that specific
window of data, determined by the user.
Physical Code

1. //+------------------------------------------------------------------+
2. //|

Support and Resistance Lines.mq4 |

3. //|
4. //|

Group 5 |
http://www.metaquotes.net |

B26

5. //+------------------------------------------------------------------+
6. #property copyright "Group 5
7. #property link

"http://www.metaquotes.net"

8.
9. #property indicator_chart_window
10. //+------------------------------------------------------------------+
11. //| Custom indicator initialization function

12. //+------------------------------------------------------------------+
13.
14. intminima_plot_index;
15. intmaxima_plot_index=0;
16.
17. intinit()
18.

19.
20.
21. //---- indicators
22. //---23.
24.

return(0);
}

25. //+------------------------------------------------------------------+
26. //| Custom indicator deinitialization function

27. //+------------------------------------------------------------------+
28. intdeinit()
29.

30. //---31.

// remove minima objects

32.

/*

33.

for(inti = 0; i<minima_plot_index; i++) {

34.
35.

ObjectDelete("Minima" + i);
}

B27

36.

*/

37.
38.

// remove maxima objects

39.
40.

for(int j =0; j <maxima_plot_index; j++){

41.

ObjectDelete("Maxima"+ j);

42.

43. //---44.
45.

return(0);
}

46. //+------------------------------------------------------------------+
47. //| Custom indicator iteration function

48. //+------------------------------------------------------------------+
49. int start()
50.

51.
52.

intcounted_bars=IndicatorCounted();

53.
54.

intbars_count

=WindowBarsPerChart();

55.
56.

intsliding_window_size=bars_count/10;

57.
58.

double min;

59.

double max;

60.

double range;

61.
62.

intwindow_min_index=0, window_max_index=0;

63.
64.

double minima[1000];

65.

double maxima[1000];

66.

doubleminima_copy[1000];

B28

67.

doublemaxima_copy[1000];

68.
69.

doubleminima_weight[1000];

70.

doublemaxima_weight[1000];

71.

doublesorted_maxima_weight[1000];

72.

doublesorted_minima_weight[1000];

73.
74.

intminima_index=0;

75.

intmaxima_index=0;

76.
77.

for(inti=0;i<bars_count-sliding_window_size;i++)

78.

79.

min =1000000;

80.

max =-1000000;

81.
82.

for(int j =0; j <sliding_window_size; j++)

83.

84.

if(min > Low[i+j]){

85.

min = Low[i+j];

86.

window_min_index= j;

87.

88.
89.

if(max < High[i+j]){

90.

max = High[i+j];

91.

window_max_index= j;

92.
93.

}
}

94.
95.

range = max - min;

96.
97.

// if max and min are at boundaries, then we ignore them

B29

98.
99.

if((window_max_index!=sliding_window_size&&window_max_index!=0)&&((maxima_index==0)||(M
athAbs(max - maxima[maxima_index-1]))>0.0001)){

100.

maxima[maxima_index]= High[i+window_max_index];

101.

maxima_weight[maxima_index]= range *5;

102.

maxima_index++;

103.

104.
105.

if((window_min_index!=sliding_window_size&&window_min_index!=0)&&((minima_index=

=0)||(MathAbs(min - minima[minima_index-1]))>0.0001)){
106.

minima[minima_index]= min;

107.

minima_weight[minima_index]=10*MathSqrt(range);

108.

minima_index++;

109.

110.
111.

112.
113.

ArrayCopy(minima_copy, minima);

114.

ArrayCopy(maxima_copy, maxima);

115.
116.

minima_plot_index=minima_index;

117.
118.

/*

119.

for(int n = 0; n <minima_index; n++) {

120.

ObjectCreate("Minima" + n, OBJ_HLINE, 0, NULL, minima[n]);

121.

ObjectSet("Minima" + n, OBJPROP_COLOR, Blue);

122.

ObjectSet("Minima" + n, OBJPROP_STYLE, STYLE_SOLID);

123.

ObjectSet("Minima" + n, OBJPROP_WIDTH, 3);

124.
125.

}
*/

126.

B30

127.

ArraySort(maxima, maxima_index, 0, MODE_ASCEND);

128.

ArraySort(minima, minima_index, 0, MODE_ASCEND);

129.
130.

int w =0;

131.
132.

for(w =0; w <maxima_index; w++){

133.

intsorted_maxima_index=ArrayBsearch(maxima, maxima_copy[w], maxima_index);

134.

sorted_maxima_weight[sorted_maxima_index]=maxima_weight[w];

135.

136.
137.

for(w =0; w <minima_index; w++){

138.

intsorted_minima_index=ArrayBsearch(minima, minima_copy[w]);

139.

sorted_minima_weight[sorted_minima_index]=minima_weight[w];

140.

141.
142.
143.
144.

doubleoutlier_adjusted_range=MathAbs(maxima[maxima_index-1]- minima[0]);

145.
146.

intmaxima_plot_index=0;

147.
148.

// if there exist maxima

149.

if(maxima_index>0)

150.

151.

// the index of the current maxima we are comparing against

152.

intcurrent_maxima_index=0;

153.

// the current maxima itself

154.

doublecurrent_maxima= maxima[0];

155.

// the average of the cluster of values

156.

doubleaverage_current_maxima=0;

157.

// the average weight of this cluster

B31

158.

doubleaverage_weight=0;

159.
160.

int k;

161.
162.
163.

for(k =0; k <maxima_index; k++){

Print("outlier_adjusted_range", "::", outlier_adjusted_range, "::",

10*(maxima[k]-current_maxima));
164.

if(10*MathAbs(maxima[k]-current_maxima)>outlier_adjusted_range){

165.

// determine whether values are significant and plot

166.

average_current_maxima=average_current_maxima/(k -current_maxima_index);

167.

// assume only one maxima with exact same value

168.

average_weight=average_weight/MathSqrt(MathSqrt(k -current_maxima_index));

169.
170.

Print("average_weight: ", average_weight);

171.

if(average_weight>0.05){

172.

ObjectCreate("Maxima"+maxima_plot_index, OBJ_HLINE, 0, NULL,

average_current_maxima);
173.

ObjectSet("Maxima"+maxima_plot_index, OBJPROP_COLOR, Yellow);

174.

ObjectSet("Maxima"+maxima_plot_index, OBJPROP_STYLE, STYLE_SOLID);

175.

ObjectSet("Maxima"+maxima_plot_index, OBJPROP_WIDTH, 3);

176.

maxima_plot_index++;

177.

178.
179.

Print("Average weight ", average_weight);

180.

current_maxima= maxima[k];

181.

current_maxima_index= k;

182.

average_current_maxima= maxima[k];

183.

}else{

184.

// assume only one maxima with a specific value

185.

average_current_maxima=average_current_maxima+ maxima[k];

186.

average_weight=average_weight+sorted_maxima_weight[k];

B32

187.

188.

189.

190.
191.

// repeat code:

192.

// determine whether values are significant and plot

193.

Print("debug2 ", k, " ", current_maxima_index, " ", maxima_index);

194.

average_current_maxima=average_current_maxima/(k -current_maxima_index);

195.

// assume only one maxima with exact same value

196.

average_weight=average_weight/(k -current_maxima_index);

197.
198.

Print("Data: ", average_current_maxima, ":", average_weight);

199.
200.
201.

return(0);
}

B.2.7 Pivots

Similar to the Support Resistance program displayed earlier, the Pivots indicator also creates
horizontal lines at key points in a chart. However, the support resistance program calculates those
prices which are returned to several times whereas the pivot program is designed to create one line for
each bucket. In this manner regardless of where the current price is this program can identify key points
around that price. The program also displays 2 additional lines at the chart extrema displayed in a bold
green line.
Additionally this program only considers the most recent bars whereas the support resistance
program updates when the window is scrolled. It operates much in the same manner, looking at 6 bar
increments for highs and lows and storing them. It continues in this manner until all bars in the window
B33

have been considered. The fact that this window does not update allows you to consider these same
points as time progresses and was used heavily in the latter formulation of Brendan Hamms trading
strategy. The number of buckets to be looked at is user controlled, so if one feels that too many points
are being considered, they need only reduce the buckets.

Figure 49: Pivots indicator set to find 10 pivot points (yellow lines). The extrema are shown as bolded green lines. In this
instance the lowest point is the only point of interest for the bottom bucket, thus only 9 yellow lines appear.

Physical Code
1. //+------------------------------------------------------------------+
2. //|

Pivotsv1.07.mq4 |

3. //|

Brendan Hamm |

4. //|

http://www.metaquotes.net |

5. //+------------------------------------------------------------------+
6. #property copyright "Brendan Hamm"

B34

7. #property link

"http://www.metaquotes.net"

8.
9. #property indicator_chart_window
10. //--- input parameters
11. externint

shift=6;

12. externint

num_of_buckets=10;

13. //+------------------------------------------------------------------+
14. //| Custom indicator initialization function

15. //+------------------------------------------------------------------+
16. intinit()
17.

18. //---- indicators

19. //---20.
21.

return(0);
}

22. //+------------------------------------------------------------------+
23. //| Custom indicator deinitialization function

24. //+------------------------------------------------------------------+
25. intdeinit()
26.

27. //---28.
29. //---30.
31.

return(0);
}

32. //+------------------------------------------------------------------+
33. //| Custom indicator iteration function

34. //+------------------------------------------------------------------+
35. int start()
36.

37.

int

counted_bars=IndicatorCounted();

38.

int

pos=Bars-counted_bars;

39.

double high[];

40.

ArrayResize(high, shift);

41.

double low[];

42.

ArrayResize(low, shift);

43.

int

highpos,lowpos;

44.

int

bars_count=WindowBarsPerChart();

45.

int

first=bars_count-WindowFirstVisibleBar();

46.

B35

47.

string hlinehigh, hlinelow;

48.

string highs[];

49.

ArrayResize(highs, bars_count/shift+1);

50.

string lows[];

51.

ArrayResize(lows, bars_count/shift+1);

52.

string globalhigh="globalhigh";

53.

string globallow

54.

doublemaxhigh, current1, current2;

55.

doubleminlow=4;

="globallow";

56.
57.
58.

for(int k=first; k<bars_count+first; k=k+shift)

59.

60.

for(inti=0;i<shift;i++)

61.

62.

high[i]=High[i+k];

63.

if(Low[i+k]!=0)

64.

65.

low[i]=Low[i+k];

66.

67.

68.

highpos=ArrayMaximum(high);

69.

lowpos=ArrayMinimum(low);

70.

hlinehigh="hlinehigh_"+(highpos+k);

71.

hlinelow="hlinelow_"+(lowpos+k);

72.

highs[k/shift]=hlinehigh;

73.

lows[k/shift]=hlinelow;

74.

ObjectCreate(hlinehigh,OBJ_HLINE,0,0,high[highpos]);

75.

ObjectCreate(hlinelow,OBJ_HLINE,0,0,low[lowpos]);

76.

ObjectSet(hlinelow, OBJPROP_COLOR, Blue);

77.

78.

minlow=ObjectGet(lows[0],1);

79.

for(int h=0; h<bars_count/shift+1; h++)

80.

81.

current1=ObjectGet(highs[h],1);

82.

if(current1>maxhigh)maxhigh=current1;

83.

current2=ObjectGet(lows[h],1);

84.
85.

if(current2<minlow&& current2!=0)

86.

minlow=current2;

B36

87.
88.
89.
90.

ObjectCreate(globalhigh,OBJ_HLINE,0,0,maxhigh);

91.

ObjectCreate(globallow,OBJ_HLINE,0,0,minlow);

92.

ObjectSet(globalhigh, OBJPROP_COLOR, Lime);

93.

ObjectSet(globallow, OBJPROP_COLOR, Lime);

94.

ObjectSet(globalhigh, OBJPROP_WIDTH, 3);

95.

ObjectSet(globallow, OBJPROP_WIDTH, 3);

96.
97.

doublebucketsize=(maxhigh-minlow)/num_of_buckets;

98.

int count;

99.

doubleavgprices;

100.

string avg;

101.
102.

for(int p=1;p<=num_of_buckets;p++)

103.

104.

avgprices=0;

105.

count=0;

106.

for(int q=0;q<bars_count/shift+1;q++)

107.

108.

current1=ObjectGet(highs[q],1);

109.

current2=ObjectGet(lows[q],1);

110.

if(current1<=minlow+(p*bucketsize)&& current1>minlow+((p-1)*bucketsize))

111.

112.

avgprices=avgprices+current1;

113.

count=count+1;

114.
115.

116.

if(current2<=minlow+(p*bucketsize)&& current2>minlow+((p-1)*bucketsize))

117.

118.

avgprices=avgprices+current2;

119.

count=count+1;

120.
121.

122.
123.

}//end q loop

124.

if(p==1)

125.

126.

avgprices=avgprices+minlow;

B37

127.

count=count+1;

128.
129.

130.

if(avgprices>0)

131.

132.

avgprices=avgprices/count;

133.

avg="avg_"+p;

134.

ObjectCreate(avg,OBJ_HLINE,0,0,avgprices);

135.

ObjectSet(avg, OBJPROP_COLOR, Yellow);

136.

ObjectSet(avg, OBJPROP_WIDTH, 1);

137.

138.
139.

}//end p loop

140.

for(int r=0; r<bars_count/shift+1; r++)

141.

142.

ObjectDelete(highs[r]);

143.

ObjectDelete(lows[r]);

144.

145.

WindowRedraw();

146.
147.
148.

//----

149.
150.
151.
152.
153.

//---return(0);
}
//+------------------------------------------------------------------+

B.2.8 Bars

After having experimented with several different forms of expert advisors we began to feel that
the various variables which are held constant might better suit the intentions of a program if they were
adjusted to the current market conditions. For instance, when determining the level at which to place a
stop loss, a fixed value of 5 pips will not be a good value for a currency pair experiencing a period of high
volatility, but may be suitable in a rather calm environment. Thus to adjust such values to the current
market conditions it can be very useful to have an idea of what the average bar size over a given period
B38

is. Thus, we have created a program which does exactly this and displays this information on the chart
screen. While it may not be useful to most users who are manually trading, the functionality of this
indicator can be combined with an expert advisor to enhance its capabilities.
Since this program was intended to be used for a myriad of purposes, we decided to calculate
average bar sizes both from wick to wick, as well as from open to close. The open close measurements
can be compared to the total bar size to give the user an idea of how much noise is being experienced
during the time interval in question.

Figure 50:The Bars Program is displayed in a separate window from the chart. The blue histogram displays the body size for
each individual bar, while the red line displays the total bar size. The black and purple lines are for a 5-period moving
average of the body and total size respectively. Similarly, the green and orange lines are for a 12-period moving average of
the body and total size respectively.

Figure 51: Zoom of the Bars Indicator using the same values as the previous figure.

B39

Physical Code

1. //+------------------------------------------------------------------+

2. //|

BarSize.mq4 |

3. //|

Group5 |

4. //|

http://www.metaquotes.net |

5. //+------------------------------------------------------------------+

6. #property copyright "Group 5"

7. #property link

"http://www.metaquotes.net"

8.

9. #property indicator_separate_window

10. #property indicator_buffers 6

11. #property indicator_color1 Blue

12. #property indicator_color2 Red

13. #property indicator_color3 Chartreuse

14. #property indicator_color4 Orange

15. #property indicator_color5 Yellow

16. #property indicator_color6 Purple

17. //--- input parameters

18. externdouble N=12;

B40

19. externdouble M=5;

20. //--- buffers

21. double ExtMapBuffer1[];

22. double ExtMapBuffer2[];

23. double ExtMapBuffer3[];

24. double ExtMapBuffer4[];

25. double ExtMapBuffer5[];

26. double ExtMapBuffer6[];

27.

28. //+------------------------------------------------------------------+

29. //| Custom indicator initialization function

30. //+------------------------------------------------------------------+

31. intinit()

32.

33. //---- indicators

34.

SetIndexStyle(0,DRAW_HISTOGRAM);

35.

SetIndexBuffer(0,ExtMapBuffer1);

36.

SetIndexStyle(1,DRAW_LINE);

37.

SetIndexBuffer(1,ExtMapBuffer2);

B41

38.

SetIndexStyle(2,DRAW_LINE);

39.

SetIndexBuffer(2,ExtMapBuffer3);

40.

SetIndexStyle(3,DRAW_LINE);

41.

SetIndexBuffer(3,ExtMapBuffer4);

42.

SetIndexStyle(4,DRAW_LINE);

43.

SetIndexBuffer(4,ExtMapBuffer5);

44.

SetIndexStyle(5,DRAW_LINE);

45.

SetIndexBuffer(5,ExtMapBuffer6);

46.

stringshort_name="We got Bar Sizes!";

47.

//IndicatorShortName(short_name);

48. //----

49.

50.

return(0);

51. //+------------------------------------------------------------------+

52. //| Custom indicator deinitialization function

53. //+------------------------------------------------------------------+

54. intdeinit()

55.

56. //----

B42

57.

58. //----

59.

60.

return(0);

61. //+------------------------------------------------------------------+

62. //| Custom indicator iteration function

63. //+------------------------------------------------------------------+

64. int start()

65.

66.

int

counted_bars=IndicatorCounted();

67.

doubleBarSize;

68.

doubleMaxSize;

69.

doubleAvgMaxSize=0;

70.

doubleAvgBarSize=0;

71.

doubleOutputB;

72.

doubleOutputM;

73.

int j=0;//used to count to 12

74.

75. //---- check for possible errors

B43

76.

if(counted_bars<0)return(-1);

77. //---- last counted bar will be recounted

78.

if(counted_bars>0)counted_bars--;

79. //---- DEfine the position to look at

80.

intpos=Bars-counted_bars;

81.

82.

while(pos>=0)

83.

84.

OutputB=0;

85.

OutputM=0;

86.

BarSize=MathAbs(Open[pos]-Close[pos])*10000;

87.

MaxSize=(High[pos]-Low[pos])*10000;

//calculate the body size

//Calculate the size from wick to wick

88.

89.

ExtMapBuffer2[pos]=MaxSize;//send to buffer2 to plot

90.

ExtMapBuffer1[pos]=BarSize;//send it to buffer to plot

91.

92.

for(int k=0; k<N; k++)

93.

94.

OutputB=OutputB+ExtMapBuffer1[pos+k];

B44

95.

OutputM=OutputM+ExtMapBuffer2[pos+k];

96.

97.

ExtMapBuffer3[pos]=OutputB/N;

98.

ExtMapBuffer4[pos]=OutputM/N;

99.

OutputB=0;

100.

OutputM=0;

101.

for(k=0; k<M; k++)

102.

103.

OutputB=OutputB+ExtMapBuffer1[pos+k];

104.

OutputM=OutputM+ExtMapBuffer2[pos+k];

105.

106.

ExtMapBuffer5[pos]=OutputB/M;

107.

ExtMapBuffer6[pos]=OutputM/M;

108.

109.

110.

pos--;

111.

112.

113.

//----

B45

114.

115.

116.

117.

118.

//----

return(0);

//+------------------------------------------------------------------+

B.2.9 Order History Write

The program OrderHistoryWrite.mq4 was created to export trade data to a .csv file. This
program is useful for when you need to export all data from your trades to create a spreadsheet of the
progress of a portfolio. In the MT4 platform one can select the range of history they are interested in
and run the script. As is the case with all scripts in MQL4it runs only once creating a file called
tradetracker.csv located in ROOT/experts/files which contains the following information in each row:
OrderOpenTime, Order Type, Order Lots, OrderOpenPrice, OrderStopLoss, OrderTakeProfit,
OrderCloseTime, OrderClose Price, OrderPipProfitand,OrderProfit. The fields are defined as follows:
Order Open Time = yyyy.mm.dd hh.mm at which the transaction was executed (for some brokers this
value may not be the same time zone as the users)
Order Type = buy/sell
Order Lots = integer value in standard lots
Order Open Price = Price at which the transaction was executed

B46

Order Stop Loss = Last known value of stop loss applied to transaction (unfortunately any changes to a
stop loss are notrecorded and thus unavailable)
Order Take Profit = Last known value of take profit level (also does not record changes during
transaction)
Order Close Time = yyyy.mm.dd hh.mm at which the transaction was terminated (for some brokers this
value may not be the same time zone as the users)
Order Close Price = Price at which transaction was terminated
Order Pip Profit = Number of total pips made, expressed as:

Order Profit = Total Profit of the trade (currently only available for USD based currencies)
For an example of the output, refer to any of the trading portfolios for our group.
Physical Code

1. //+------------------------------------------------------------------+
2. //|

OrderHistoryWrite.mq4 |

3. //|
4. //|

Group 5|
http://www.metaquotes.net |

5. //+------------------------------------------------------------------+
6. #property copyright "Group 5"
7. #property link

"http://www.metaquotes.net"

8.
9.
10. bool Comma = True;

// Some Spread Sheets want "," instead of "." for cent

11. string Doub2String (doubledoub, int digits)

B47

12.

13.

string num, temp;

14. num=DoubleToStr(doub,digits);
15.

temp =num;

16.

if(Comma)

17.

18. doub=StringFind(num,".",0);
19.
20.

if(doub!=1)

21.

22.

temp =StringSubstr(num,0,doub);

23.

24.

25.

return(temp);

26.

27. string Typ2String(int type)

28.

29.

string typ;

30.

if(type == OP_BUY)

31.

if(type == OP_SELL)typ="sell";

32.

return(typ);

33.

typ="buy";

34.
35. voidWriteOrder(int handle)
36.

37. int type;

38.

doublepipprofit;

39.

type =OrderType();

40.

if(type==OP_BUY || type==OP_SELL)

41.

42.

if(type == OP_BUY)

43.

B48

44. pipprofit=(OrderClosePrice()-OrderOpenPrice())*10000;
45.

46.

if(type == OP_SELL)

47.

48. pipprofit=(OrderOpenPrice()-OrderClosePrice())*10000;
49.

50. FileWrite(handle,TimeToStr(OrderOpenTime()), Typ2String(OrderType()),

51.

Doub2String(OrderLots(),3), Doub2String(OrderOpenPrice(),5),

52.

Doub2String(OrderStopLoss(),5),Doub2String(OrderTakeProfit(),5),

53. TimeToStr(OrderCloseTime()), Doub2String(OrderClosePrice(),5),

54.

Doub2String(pipprofit,5), Doub2String(OrderProfit(),5));

55.

56.

57. //+------------------------------------------------------------------+
58. //| script program start function

59. //+------------------------------------------------------------------+
60. int start()
61.

62. //---63. int handle, total, error;

64. total=OrdersHistoryTotal();
65. handle =FileOpen("TradeTracker.csv", FILE_CSV|FILE_WRITE,\',\');
66. FileWrite(handle, "OrderOpenTime", "OrderType", "OrderLots", "OrderOpenPrice",
"OrderStopLoss", "OrderTakeProfit", "OrderCloseTime", "OrderClosePrice", "OrderPipProfit",
"OrderProfit");
67. for(int count=0;count<total;count++)
68. {
69.
70.

if(OrderSelect(count,SELECT_BY_POS, MODE_HISTORY)==true)
{

71. WriteOrder(handle);
72.

73.

else

B49

74.

75.

error=GetLastError();

76.

Print("Last Error was ",error);

77.

78. }
79. FileClose(handle);
80. //---81.
82.

return(0);
}

83. //+------------------------------------------------------------------+

B.2.10 Currency Meter

General Overview
For this interactive qualifying project, our team has implemented a simple currency meter that
determines and assigns strength to each currency using chart values for each currency and displays the
calculated value graphically to the user. This software will help the user to decide which currency pair to
trade as it makes sense to trade a strong currency against a weak currency. The user can specify
approximately how many bars will be used in the calculation as well as the period of the bars (which can
be 1 minute, 5 minutes, 15 minutes, etc.). To report the strength of the currency, the tool will calculate a
numeric strength and quantize it an integer strength index. Then a graphical icon denoting the
strength index is displayed to the user in the form of a number of upward or downward facing bars, with
upward facing bars indicating strength and downward facing bars indicating weakness.

B50

Strength Description
Exceptionally strong

Icon Description
Five upward-facing bars

Extremely Strong

Four upward-facing bars

Very Strong

Three upward-facing bars

Strong

Two upward-facing bars

Slightly strong

One upward-facing bar

Slightly weak

One downward-facing bars

Weak

Two downward-facing bars

B51

Icon

Very weak

Three downward-facing
bars

Extremely weak

Four downward-facing bars

Exceptionally weak

Five downward-facing bars

The currency meter features tiles for all currencies that are supported by the individual broker
and also features a display mode, in which all the currency pairs are shown which can be sorted by
spread, bid price, ask price, and the spread. The currency meter shows the current state of connection
with the MT4 client (i.e., whether it is connected or not and if it is connected, how many frames it has
received).
Below is the screenshot of the software in tabular mode, where we can see each currency pair,
the strength, buy and ask values, and the spread.

B52

Figure 52 - Currency Meter Running in Table Mode

Here, we have adjusted for the spread of currencies that include the Japanese Yen as well as
rounding the spread values to the nearest third decimal place. The strength value is also shown and was
used extensively for debugging purposes.
The software also runs in a graphical mode, in which we can see a strength indication for each
currency. We can also change the number of bars used in the calculation as well as the period of the
calculation (M1, M5, M15, H1, etc.). The software sends the new symbol period to the server and gets
new base data, to which new bars are added.

B53

Figure 53 - Currency Meter Running in Graphical View

Algorithmic Overview
The software uses a robust algorithm to determine the strength of each currency based on
pairwise currency strength comparisons. We keep track of the last 1440 bars (which is equivalent to one
day when each bar represents one minute). However, the user can select the number of bars that are
B54

actually used to determine the strength calculations that are presented to the user. When a new bar is
added, an old bar is removed. This enables us to use a finite amount of memory and avoid leaks. The
latest tick values are also included in the display interface though they are not stored due to memory
constraints.
One of the core ideas behind this project is to define strength. Currencies that are stronger
tend to rise in terms of the other currencies. We determine a basic intrinsic strength of a currency
as the weighted sum of its change over the entire period with respect to the US dollar. The
weighting depends on how much data you wish to consider. To determine the exponential
weighting, we determine the value based on the bar N, where the Nth bar has the weighting of
the first bar. We select n as the number of bars the user wishes to use in the calculation.
Mathematically, we would show as follows.

We then simulate exchange rates at the open and close times by dividing the target currency
by another currency (this is repeated for all currencies). The close price is subtracted from the open
price and divided by the close price quotient to normalize the values. Correlation is used to
determine the similarity of movement of the open-to-close price and the high-to-low price for each
currency; the two correlation values are averaged. The pairwise strength between currencies that
are less correlated is weighted higher. We make the assumption that our process has an underlying
normal distribution so we use Pearson product-moment correlation over every bar.
We determine a correlation matrix, which is symmetric due to the relation
The correlation coefficient can be calculated as below.

B55

We calculate the matrix of weightings P as follows, knowing that it is symmetric across the
main diagonal, which are ones because each currency is perfectly correlated with itself. Perfect
correlation has zero weighting as weighting is 1 minus the correlation value. We divide the raw
value for a currency by the sum of the entries in its corresponding row.

The intrinsic strength is calculated as the movement of every other currency with respect to
the US dollar. Fortunately, every currency has a pair with the US dollar. If the other currency is the
first in the pair with the dollar, then we invert the currency so that it is normalized to the dollar,
which is the base currency. For the US dollar or a possibility of a currency without a pair against the
US dollar, we normalize both currencies with respect to the Euro or Japanese Yen. We also
normalize the currency with respect to itself in order to get a percentage gain so that weightage is
based on the correlation alone. However, we use the alpha parameter to discriminate based on how
long ago the bar occurred. The actual strength is calculated as follows:

There is one caveat, which is to invert the currency if the US Dollar is the first term; for
example, we would invert USDJPY to get the simulated value of JPYUSD. The resulting strengths are
then quantized into icons displaying strength using quantization buckets that we determined by
visualizing large sets of data.

B56

To make the strengths more relevant, we can also add a weightage factor based on the
trading period and integrating the auto-correlation

, where x is the strength of the currency

and y is the difference between open and close. This makes sense because if the currency strength
as a whole is not correlated with the increase in the currency after

units, then the statistic is less

meaningful. If we know this information, we can also incorporate the spread into the calculation (its
usefulness depends on trading duration).
Software Overview
The software is designed to operate with the MQL4 trading platform though it can be easily adapted to
TradeStation or any other platform with C++ integration. Since we already figured out the task of
C++/TradeStation integration, we could easily implement the client side module to interface with
TradeStation, although for demonstration, the work that we have completed suffices.
The basic operation is that the trading software first sends the last 1440 bars of open, close, high, low,
bid, and ask values for each currency pair over to the currency meter software. This is done using a
named pipe, which is an inter-process communication mechanism in Windows. Each of these messages
is sent in ASCII over a 64 kB frame. Then, for each bar and for each tick, new data is sent. If the data is
sent for a bar, then it is added to the circular buffer and the first bar recorded is removed. If the data is
sent for a tick, then the tick data is merely shown on the tabular display of the runtime.
The algorithm is calculated in each process and the tables as well as the graphical view are
updated instantaneously to changes in the number of bars to use. The bar period is also incorporated
into the system; this value is sent to MQL immediately after a new data frame has been received (for a
tick or bar). If the period changed then the initialization data of the first bars is sent from MQL to the
Currency Meter. There is code in the MQL4 end that sends new bars according to the period of the bar.

B57

1. // MatlabMql.cpp : Defines the exported functions for the DLL application.

2. //
3.
4. #include "stdafx.h"
5. #include "GlobalPipeVars.h"
6. #include "PipeUtils.h"
7. #include <sstream>
8.
9. using namespace std;
10.
11. struct MqlStr {
12.

int

len;

13.

char

*string;

14. };
15.
16. _DLLAPI int __stdcall initialSendDataToServer(int startIndex, int numBarsToDownload, int
numSymbols, char* symbols, double* high, double* low, double* open, double* close);
17. _DLLAPI void __stdcall sendStringToServer(char* textToSend);
18. /*
19. _DLLAPI void __stdcall testSendDataToServer() {
20.

char** symbols = {"EURUSD", "AUDUSD", "CADUSD", "GBPUSD", "JPYUSD"};

21.

double transportCosts[25] = { 1, 2, 3, 4, 5, 1, 2, 3, 4, 5, 1, 2, 3, 4, 5, 1, 2, 3, 4,
5, 1, 2, 3, 4, 5 };

22.

initialSendDataToServer(10, 5, 5, symbols, transportCosts, transportCosts, transportCosts,

transportCosts);

23.

//sendStringToServer("BING");

24. }
25. */
26.
27. _DLLAPI int __stdcall initialSendDataToServer(int startIndex, int numBarsToDownload, int
numSymbols, char* symbols, double* high, double* low, double* open, double* close) {
28.

DWORD cbWritten;

29.
30.

stringstream ss (stringstream::in | stringstream::out);

31.
32.

int code_1 = 100;

33.

ss << "

" << code_1 << " ";

34.
35.

ss << startIndex << " ";

36.

ss << numBarsToDownload << " ";

B58

37.

ss << numSymbols << " ";

38.
39.

string str = symbols;

40.

ss << str << " ";

41.
42.

for(int i = 0; i < numSymbols * numBarsToDownload; i++) {

43.

ss << high[i] << " ";

44.

45.
46.

for(int i = 0; i < numSymbols * numBarsToDownload; i++) {

47.

ss << low[i] << " ";

48.

49.
50.

for(int i = 0; i < numSymbols * numBarsToDownload; i++) {

51.

ss << open[i] << " ";

52.

53.
54.

for(int i = 0; i < numSymbols * numBarsToDownload; i++) {

55.

ss << close[i] << " ";

56.

57.
58.

string textToSendStr = ss.str();

59.
60.

const char* textToSendArr = textToSendStr.c_str();

61.

int bytesTextToSend = textToSendStr.length();

62.
63.

if (!WriteFile(

64.

hPipe,

// Handle of the pipe

65.

textToSendArr,

// Message to be written

66.

bytesTextToSend + 1,

// Number of bytes to write

67.

&cbWritten,

// Number of bytes written

68.

NULL

// Not overlapped

69.

))

70.

71.

dwError = GetLastError();

72.

cleanUp();

73.
74.

return 100;
}

75.
76.

return 90;

B59

77. }
78.
79. _DLLAPI int __stdcall sendNewBarToServer(int numSymbols, double* high, double* low, double*
open, double* close, double* currentBid, double* currentAsk) {
80.

DWORD cbWritten;

81.
82.

stringstream ss (stringstream::in | stringstream::out);

83.

int code_1 = 1000;

84.
85.

ss << "

" << code_1 << " ";

86.
87.

ss << numSymbols << " ";

88.
89.

for(int i = 0; i < numSymbols; i++) {

90.

ss << high[i] << " ";

91.

92.
93.

for(int i = 0; i < numSymbols; i++) {

94.

ss << low[i] << " ";

95.

96.
97.

for(int i = 0; i < numSymbols; i++) {

98.
99.

ss << open[i] << " ";

}

100.
101.

for(int i = 0; i < numSymbols; i++) {

102.

ss << close[i] << " ";

103.

104.
105.

for(int i = 0; i < numSymbols; i++) {

106.
107.

ss << currentBid[i] << " ";

}

108.
109.

for(int i = 0; i < numSymbols; i++) {

110.
111.

ss << currentAsk[i] << " ";

}

112.
113.

string textToSendStr = ss.str();

114.

const char* textToSendArr = ss.str().c_str();

115.

int bytesTextToSend = textToSendStr.length();

B60

116.
117.

if (!WriteFile(

118.

hPipe,

// Handle of the pipe

119.

textToSendArr,

// Message to be written

120.

bytesTextToSend + 1,

// Number of bytes to write

121.

&cbWritten,

// Number of bytes written

122.

NULL

// Not overlapped

123.

))

124.

125.

dwError = GetLastError();

126.

cleanUp();

127.

return 1;

128.

129.
130.
131.

return 0;
}

132.
133.
134.

_DLLAPI int __stdcall sendNewTickToServer(int numSymbols, double* high, double* low,

double* open, double* close, double* currentBid, double* currentAsk) {

135.

DWORD cbWritten;

136.
137.

stringstream ss (stringstream::in | stringstream::out);

138.

int code_1 = 10000;

139.
140.

ss << "

" << code_1 << " ";

141.
142.

ss << numSymbols << " ";

143.
144.

for(int i = 0; i < numSymbols; i++) {

145.

ss << high[i] << " ";

146.

147.
148.

for(int i = 0; i < numSymbols; i++) {

149.

ss << low[i] << " ";

150.

151.
152.

for(int i = 0; i < numSymbols; i++) {

153.
154.

ss << open[i] << " ";

}

B61

155.
156.

for(int i = 0; i < numSymbols; i++) {

157.

ss << close[i] << " ";

158.

159.
160.

for(int i = 0; i < numSymbols; i++) {

161.

ss << currentBid[i] << " ";

162.

163.
164.

for(int i = 0; i < numSymbols; i++) {

165.

ss << currentAsk[i] << " ";

166.

167.
168.

string textToSendStr = ss.str();

169.

const char* textToSendArr = ss.str().c_str();

170.

int bytesTextToSend = textToSendStr.length();

171.
172.

if (!WriteFile(

173.

hPipe,

// Handle of the pipe

174.

textToSendArr,

// Message to be written

175.

bytesTextToSend + 1,

// Number of bytes to write

176.

&cbWritten,

// Number of bytes written

177.

NULL

// Not overlapped

178.

))

179.

180.

dwError = GetLastError();

181.

cleanUp();

182.

return 1;

183.

184.
185.
186.

return 0;
}

187.
188.

_DLLAPI int __stdcall sendDataToServer(int numSymbols, int tickNumber, int barNumber,

double* high, double* low, double* open, double* close, double* currentBid, double*
currentAsk) {
189.

DWORD cbWritten;

190.
191.

stringstream ss (stringstream::in | stringstream::out);

192.

B62

193.

int code_1 = 0x00000000;

194.

ss << code_1 << " ";

195.

ss << numSymbols << " ";

196.

ss << tickNumber << " ";

197.

ss << barNumber << " ";

198.
199.

for(int i = 0; i < numSymbols; i++) {

200.

ss << high[i] << " ";

201.

202.
203.

for(int i = 0; i < numSymbols; i++) {

204.

ss << low[i] << " ";

205.

206.
207.

for(int i = 0; i < numSymbols; i++) {

208.

ss << close[i] << " ";

209.

210.
211.

for(int i = 0; i < numSymbols; i++) {

212.

ss << currentBid[i] << " ";

213.

214.
215.

for(int i = 0; i < numSymbols; i++) {

216.

ss << currentAsk[i] << " ";

217.

218.
219.

string textToSendStr = ss.str();

220.

const char* textToSendArr = ss.str().c_str();

221.

int bytesTextToSend = textToSendStr.length();

222.
223.

if (!WriteFile(

224.

hPipe,

// Handle of the pipe

225.

textToSendArr,

// Message to be written

226.

bytesTextToSend + 1,

// Number of bytes to write

227.

&cbWritten,

// Number of bytes written

228.

NULL

// Not overlapped

229.

))

230.

231.

dwError = GetLastError();

232.

cleanUp();

B63

233.

return 90;

234.

235.
236.
237.

return 100;
}

238.
239.

_DLLAPI void __stdcall sendStringVectorToServer(char** textToSend, int numStrings) {

240.

DWORD cbWritten;

241.
242.

char* last_str = textToSend[numStrings - 1];

243.

int lengthOfFinalStr = strlen(last_str);

244.

int bytesTextToSend = (&(textToSend[numStrings - 1])) - textToSend + 1;

245.
246.

if (!WriteFile(

247.

hPipe,

// Handle of the pipe

248.

textToSend,

// Message to be written

249.

bytesTextToSend,

// Number of bytes to write

250.

&cbWritten,

// Number of bytes written

251.

NULL

// Not overlapped

252.

))

253.

254.

dwError = GetLastError();

255.

cleanUp();

256.
257.

}
}

258.
259.

_DLLAPI void __stdcall sendStringToServer(char* textToSend) {

260.

DWORD cbWritten;

261.

size_t bytesTextToSend = sizeof(textToSend);

262.
263.

if (!WriteFile(

264.

hPipe,

// Handle of the pipe

265.

textToSend,

// Message to be written

266.

bytesTextToSend,

// Number of bytes to write

267.

&cbWritten,

// Number of bytes written

268.

NULL

// Not overlapped

269.

))

270.
271.

{
dwError = GetLastError();

272.

B64

273.

std::ofstream m;

274.

m.open ("C:\\\\Users\\\\Srinivas Vasudevan\\\\example.txt");

275.

m << dwError;

276.

m.close();

277.
278.

cleanUp();

279.

280.
281.

std::ofstream m;

282.

m.open ("C:\\\\Users\\\\Srinivas Vasudevan\\\\example.txt");

283.

m << "success";

284.

m.close();

285.

286.
287.

_DLLAPI void _stdcall sendDoubleVectorToServer(double* vector, int size) {

288.

DWORD cbWritten;

289.

size_t bytesTextToSend = sizeof(double) * size;

290.
291.

if (!WriteFile(

292.

hPipe,

293.

vector,

294.

bytesTextToSend,

295.

&cbWritten,

// Number of bytes written

296.

NULL

// Not overlapped

297.

))

298.

// Message to be written
// Number of bytes to write

299.

dwError = GetLastError();

300.

cleanUp();

301.
302.

// Handle of the pipe

}
}

303.
304.

_DLLAPI void _stdcall sendDoubleToServer(double value) {

305.

DWORD cbWritten;

306.

size_t bytesTextToSend = sizeof(int);

307.
308.

if (!WriteFile(

309.

hPipe,

310.

&value,

311.

bytesTextToSend,

312.

&cbWritten,

// Handle of the pipe

// Message to be written
// Number of bytes to write
// Number of bytes written

B65

313.

NULL

314.

))

315.

316.

dwError = GetLastError();

317.

cleanUp();

318.
319.

// Not overlapped

}
}

320.
321.

_DLLAPI void _stdcall sendIntVectorToServer(int* vector, int size) {

322.

DWORD cbWritten;

323.

size_t bytesTextToSend = sizeof(int) * size;

324.
325.

if (!WriteFile(

326.

hPipe,

327.

vector,

328.

bytesTextToSend,

329.

&cbWritten,

// Number of bytes written

330.

NULL

// Not overlapped

331.

))

332.

// Message to be written
// Number of bytes to write

333.

dwError = GetLastError();

334.

cleanUp();

335.
336.

// Handle of the pipe

}
}

337.
338.

_DLLAPI void _stdcall sendIntToServer(int value) {

339.

DWORD cbWritten;

340.

size_t bytesTextToSend = sizeof(int);

341.
342.

if (!WriteFile(

343.

hPipe,

344.

&value,

345.

bytesTextToSend,

346.

&cbWritten,

// Number of bytes written

347.

NULL

// Not overlapped

348.

))

349.

// Message to be written
// Number of bytes to write

350.

dwError = GetLastError();

351.
352.

// Handle of the pipe

cleanUp();
}

B66

353.

1. // dllmain.cpp : Defines the entry point for the DLL application.

2. #include "stdafx.h"
3. #include "PipeUtils.h"
4. #include "GlobalPipeVars.h"
5.
6. bool initialize();
7.
8. BOOL APIENTRY DllMain( HMODULE hModule,
9.

DWORD

ul_reason_for_call,

10.

LPVOID lpReserved

11.

12. {
13.

switch (ul_reason_for_call)

14.

15.

case DLL_PROCESS_ATTACH:

16.

return initialize();

17.

case DLL_THREAD_ATTACH:

18.

case DLL_THREAD_DETACH:

19.

case DLL_PROCESS_DETACH:

20.
21.

break;
}

22.
23.

return true;

24. }
25.
26. bool initialize() {
27.

//

28.

// Send a request from client to server

29.

//

30.

while (TRUE)

31.

32.

hPipe = CreateFile(

B67

33.

FULL_PIPE_NAME,

// Pipe name

34.

GENERIC_READ | GENERIC_WRITE,

// Read and write access

35.

0,

// No sharing

36.

NULL,

// Default security attributes

37.

OPEN_EXISTING,

// Opens existing pipe

38.

0,

// Default attributes

39.

NULL

// No template file

40.

);

41.
42.

// If the pipe handle is opened successfully ...

43.

if (hPipe != INVALID_HANDLE_VALUE)

44.

45.

wprintf(L"The named pipe (%s) is connected.\\n", FULL_PIPE_NAME);

46.

break;

47.

48.
49.

dwError = GetLastError();

50.
51.

// Exit if an error other than ERROR_PIPE_BUSY occurs.

52.

if (ERROR_PIPE_BUSY != dwError)

53.

54.

wprintf(L"Unable to open named pipe w/err 0x%08lx\\n", dwError);

55.

goto Cleanup;

56.

57.
58.

// All pipe instances are busy, so wait for 5 seconds.

59.

if (!WaitNamedPipe(FULL_PIPE_NAME, 5000))

60.

61.

dwError = GetLastError();

62.

wprintf(L"Could not open pipe: 5 second wait timed out.");

63.

goto Cleanup;

64.
65.

}
}

66.
67.

return true;

68.
69.

Cleanup:

70.
71.

// Centralized cleanup for all allocated resources.

72.

if (hPipe != INVALID_HANDLE_VALUE)

B68

73.

74.

CloseHandle(hPipe);

75.

hPipe = INVALID_HANDLE_VALUE;

76.

77.
78.

return false;

79. }

1. #include "stdafx.h"
2. #include "mql4ipc.h"
3. #include "Form1.h"
4. #include "globals.h"
5.
6. /****************************** Module Header ******************************\\
7. * Module Name:

CppNamedPipeServer.cpp

8. * Project:

CppNamedPipeServer

9. * Copyright (c) Microsoft Corporation.

10. *
11. * Named pipe is a mechanism for one-way or duplex inter-process communication
12. * between the pipe server and one or more pipe clients in the local machine
13. * or across the computers in the intranet:
14. *
15. * PIPE_ACCESS_INBOUND:
16. * Client (GENERIC_WRITE) ---> Server (GENERIC_READ)
17. *
18. * PIPE_ACCESS_OUTBOUND:
19. * Client (GENERIC_READ) <--- Server (GENERIC_WRITE)
20. *
21. * PIPE_ACCESS_DUPLEX:
22. * Client (GENERIC_READ or GENERIC_WRITE, or both) <-->
23. * Server (GENERIC_READ and GENERIC_WRITE)
24. *
25. * This code sample demonstrates calling CreateNamedPipe to create a pipe
26. * named "\\\\.\\pipe\\SamplePipe", which supports duplex connections so that both
27. * client and server can read from and write to the pipe. The security
28. * attributes of the pipe are customized to allow Authenticated Users read and
29. * write access to a pipe, and to allow the Administrators group full access
30. * to the pipe. When the pipe is connected by a client, the server attempts to
31. * read the client\'s message from the pipe by calling ReadFile, and write a

B69

32. * response by calling WriteFile.

33. *
34. * This source is subject to the Microsoft Public License.
35. * See http://www.microsoft.com/opensource/licenses.mspx#Ms-PL.
36. * All other rights reserved.
37. *
38. * THIS CODE AND INFORMATION IS PROVIDED "AS IS" WITHOUT WARRANTY OF ANY KIND,
39. * EITHER EXPRESSED OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE IMPLIED
40. * WARRANTIES OF MERCHANTABILITY AND/OR FITNESS FOR A PARTICULAR PURPOSE.
41. \\***************************************************************************/
42.
43. void cleanup(HANDLE hNamedPipe);
44. void calculate_strength(int n);
45.
46. using namespace std;
47.
48. vector<boost::circular_buffer<double> *> high;
49. vector<boost::circular_buffer<double> *> low;
50. vector<boost::circular_buffer<double> *> open;
51. vector<boost::circular_buffer<double> *> close;
52. vector<boost::circular_buffer<double> *> bid_bar;
53. vector<boost::circular_buffer<double> *> ask_bar;
54.
55. string strarray2[100];
56.
57. double latest_high[100];
58. double latest_low[100];
59. double latest_open[100];
60. double latest_close[100];
61. double latest_bid[100];
62. double latest_ask[100];
63.
64. // used for computation
65. double data1[NUM_BARS_TO_USE];
66. double data2[NUM_BARS_TO_USE];
67. double data3[NUM_BARS_TO_USE];
68. double data4[NUM_BARS_TO_USE];
69.
70. double processed_strength_vector[MAX_NUM_SYMBOLS];
71.

B70

72. std::map<std::string, std::vector<std::string>*> currency_pair_map;

73. std::set<std::string> currencies;
74. std::map<std::string, int> currency_pair_index_map;
75. std::map<std::string, double> processed_strength_vector_map;
76.
77. vector<string> symbols_vector;
78.
79. PCURRENCYDATA pData;
80. DWORD

dwThread;

81. HANDLE

hThread;

82.
83. string textToUse = "Windows";
84.
85. int code = 0, startIndex = 0, numBarsToDownload = 0, numSymbols = 0;
86.
87. DWORD WINAPI getDataFromMql(LPVOID lpParam)
88. {
89.

DWORD dwError = ERROR_SUCCESS;

90.

HANDLE hNamedPipe = INVALID_HANDLE_VALUE;

91.
92.

// Initialize array of Global strarray

93.

Globals::strarray = gcnew array<System::String^>(100);

94.
95.

// Create the named pipe.

96.

hNamedPipe = CreateNamedPipe(

97.

FULL_PIPE_NAME,

// Pipe name.

98.

PIPE_ACCESS_DUPLEX,

// The pipe is duplex; both server and

99.

// client processes can read from and

100.

// write to the pipe

101.

PIPE_TYPE_MESSAGE |

// Message type pipe

102.

PIPE_READMODE_MESSAGE |

// Message-read mode

103.

PIPE_WAIT,

// Blocking mode is enabled

104.

PIPE_UNLIMITED_INSTANCES,

// Max. instances

105.

BUFFER_SIZE,

// Output buffer size in bytes

106.

BUFFER_SIZE,

// Input buffer size in bytes

107.

NMPWAIT_USE_DEFAULT_WAIT,

// Time-out interval

108.

NULL

// Security attributes

109.

);

110.
111.

if (hNamedPipe == INVALID_HANDLE_VALUE)

B71

112.

113.

dwError = GetLastError();

114.

wprintf(L"Unable to create named pipe w/err 0x%08lx\\n", dwError);

115.

cleanup(hNamedPipe);

116.
117.

return dwError;
}

118.
119.

wprintf(L"The named pipe (%s) is created.\\n", FULL_PIPE_NAME);

120.
121.

// Wait for the client to connect.

122.

wprintf(L"Waiting for the client\'s connection...\\n");

123.

if (!ConnectNamedPipe(hNamedPipe, NULL))

124.

125.

if (ERROR_PIPE_CONNECTED != GetLastError())

126.

127.

dwError = GetLastError();

128.

wprintf(L"ConnectNamedPipe failed w/err 0x%08lx\\n", dwError);

129.

cleanup(hNamedPipe);

130.

return dwError;

131.
132.

}
}

133.
134.

//

135.

// Let the system know that the client has been connected

136.

//

137.

Globals::textToUse = gcnew System::String("Client is connected.");

138.

Globals::form->Invoke(gcnew

System::Windows::Forms::MethodInvoker((CurrencyMeter::Form1 ^)Globals::form,
&CurrencyMeter::Form1::UpdateStatusStrip));
139.
140.

// we initialize the number of messages to zero

141.

int numMessages = 0;

142.
143.

//

144.

// Receive a request from client.

145.

//

146.
147.

while(true) {
BOOL fFinishRead = FALSE;

148.
149.

double temp = 0;

B72

150.
151.

string symbols = ""; // initialize to ""

152.
153.

char chRequest[BUFFER_SIZE];

154.

DWORD cbRequest, cbRead;

155.

cbRequest = sizeof(chRequest);

156.

memset (chRequest, 0, BUFFER_SIZE);

157.
158.

// read from the input

159.

fFinishRead = ReadFile(

160.

hNamedPipe,

// Handle of the pipe

161.

chRequest,

// Buffer to receive data

162.

cbRequest,

// Size of buffer in bytes

163.

&cbRead,

// Number of bytes read

164.

NULL

// Not overlapped I/O

165.

);

166.
167.

// check if there are any errors

168.

if (!fFinishRead && (ERROR_MORE_DATA != GetLastError()))

169.

170.

dwError = GetLastError();

171.

printf("Receive %ld bytes from client: \\"\\"\\n", cbRead);

172.

wprintf(L"ReadFile from pipe failed w/err 0x%08lx %d\\n",

dwError, cbRead);
173.

cleanup(hNamedPipe);

174.

return dwError;

175.

176.
177.

printf("Receive %ld bytes from client: \\"%s\\"\\n", cbRead,

chRequest);
178.
179.

// if the number of received bytes is greater than one

180.

if(cbRequest > 0) {

181.

stringstream s (stringstream::in | stringstream::out);

182.

s << chRequest;

183.
184.

// WARNING: KLUDGE ignore first 20 characters

185.

s.ignore(20, \' \');

186.
187.

// get these constants from the frame

B73

188.

s >> code;

189.
190.

if(code == 100) {

191.

s >> startIndex;

192.

s >> numBarsToDownload;

193.

s >> numSymbols;

194.

s >> symbols;

195.
196.

// export data into the trade evaluation engine

197.

if(Globals::numSymbols == 0) {

198.

Globals::numSymbols = numSymbols;

199.

for(int i = 0; i < numSymbols; i++) {

200.

string str1 = symbols.substr(i*6, 3);

201.

string str2 = symbols.substr(i*6 + 3,

3);
202.

Globals::strarray[i] = gcnew

System::String(symbols.substr(i * 6, 6).c_str());
203.

strarray2[i] = symbols.substr(i * 6,

6);
204.

currency_pair_index_map[symbols.substr(

i * 6, 6)] = i;
205.
206.

if(currency_pair_map.find(str1) ==

currency_pair_map.end()) {
207.

currency_pair_map[str1] = new

vector<string>();
208.

209.
210.

currency_pair_map[str1]-

>push_back(str2);
211.
212.

if(currency_pair_map.find(str2) ==

currency_pair_map.end()) {
213.

currency_pair_map[str2] = new

vector<string>();
214.

215.
216.

currency_pair_map[str2]-

>push_back(str1);
217.

B74

218.

219.
220.

// get the list of all the currencies

221.
222.

for (std::map<std::string,

std::vector<std::string>*>::iterator it = currency_pair_map.begin(); it !=
currency_pair_map.end(); it++) {
223.

currencies.insert((*it).first);

224.

225.

Globals::form->Invoke(gcnew

System::Windows::Forms::MethodInvoker((CurrencyMeter::Form1 ^)Globals::form,
&CurrencyMeter::Form1::AddRowsToDataGridView));
226.

Globals::form->Invoke(gcnew

System::Windows::Forms::MethodInvoker((CurrencyMeter::Form1 ^)Globals::form,
&CurrencyMeter::Form1::AddSymbolsToListView));
227.

228.
229.

// if the vectors are uninitialized

230.

if(high.size() == 0) {

231.

for(int i = 0; i < numSymbols + 1; i++) {

232.

high.push_back(new

boost::circular_buffer<double>(NUM_BARS_TO_USE));
233.

low.push_back(new

boost::circular_buffer<double>(NUM_BARS_TO_USE));
234.

open.push_back(new

boost::circular_buffer<double>(NUM_BARS_TO_USE));
235.

close.push_back(new

boost::circular_buffer<double>(NUM_BARS_TO_USE));
236.

bid_bar.push_back(new

boost::circular_buffer<double>(NUM_BARS_TO_USE));
237.

ask_bar.push_back(new

boost::circular_buffer<double>(NUM_BARS_TO_USE));
238.

239.
240.

for(int i = 0; i < NUM_BARS_TO_USE; i++) {

241.

high[numSymbols]->push_front(1.00);

242.

low[numSymbols]->push_front(1.00);

243.

open[numSymbols]->push_front(1.00);

244.

close[numSymbols]->push_front(1.00);

245.

high[numSymbols]->push_front(1.00);

B75

246.

low[numSymbols]->push_front(1.00);

247.
248.

}
}

249.
250.

for(int i = 0; i < numSymbols; i++) {

251.

for(int j = startIndex; j < startIndex +

numBarsToDownload; j++) {
252.

s >> temp;

253.

high[i]->push_front(temp);

254.
255.

}
}

256.
257.

for(int i = 0; i < numSymbols; i++) {

258.

for(int j = startIndex; j < startIndex +

numBarsToDownload; j++) {
259.

s >> temp;

260.

low[i]->push_front(temp);

261.
262.

}
}

263.
264.

for(int i = 0; i < numSymbols; i++) {

265.

for(int j = startIndex; j < startIndex +

numBarsToDownload; j++) {
266.

s >> temp;

267.

open[i]->push_front(temp);

268.
269.

}
}

270.
271.

for(int i = 0; i < numSymbols; i++) {

272.

for(int j = startIndex; j < startIndex +

numBarsToDownload; j++) {
273.

s >> temp;

274.

close[i]->push_front(temp);

275.
276.

}
}

277.
278.
279.

} else if (code == 1000) {

s >> numSymbols;

280.
281.

for(int i = 0; i < numSymbols; i++) {

B76

282.

s >> temp;

283.

high[i]->push_front(temp);

284.

285.
286.

for(int i = 0; i < numSymbols; i++) {

287.

s >> temp;

288.
289.

low[i]->push_front(temp);
}

290.
291.

for(int i = 0; i < numSymbols; i++) {

292.

s >> temp;

293.

open[i]->push_front(temp);

294.

295.
296.

for(int i = 0; i < numSymbols; i++) {

297.

s >> temp;

298.

close[i]->push_front(temp);

299.

300.
301.

for(int i = 0; i < numSymbols; i++) {

302.

s >> temp;

303.

bid_bar[i]->push_front(temp);

304.

305.
306.

for(int i = 0; i < numSymbols; i++) {

307.

s >> temp;

308.

ask_bar[i]->push_front(temp);

309.

310.
311.

calculate_strength(120);

312.
313.

Globals::form->Invoke(gcnew

System::Windows::Forms::MethodInvoker((CurrencyMeter::Form1 ^)Globals::form,
&CurrencyMeter::Form1::UpdateDataInTable));
314.

Globals::form->Invoke(gcnew

System::Windows::Forms::MethodInvoker((CurrencyMeter::Form1 ^)Globals::form,
&CurrencyMeter::Form1::UpdateListView));
315.
316.

} else if(code == 10000) {

s >> numSymbols;

317.

B77

318.

for(int i = 0; i < numSymbols; i++) {

319.

s >> temp;

320.

latest_high[i] = temp;

321.

322.
323.

for(int i = 0; i < numSymbols; i++) {

324.

s >> temp;

325.

latest_low[i] = temp;

326.

327.
328.

for(int i = 0; i < numSymbols; i++) {

329.

s >> temp;

330.

latest_open[i] = temp;

331.

332.
333.

for(int i = 0; i < numSymbols; i++) {

334.

s >> temp;

335.

latest_close[i] = temp;

336.

337.
338.

for(int i = 0; i < numSymbols; i++) {

339.

s >> temp;

340.

latest_bid[i] = temp;

341.

342.
343.

for(int i = 0; i < numSymbols; i++) {

344.

s >> temp;

345.

latest_ask[i] = temp;

346.

347.
348.

calculate_strength(120);

349.
350.

Globals::form->Invoke(gcnew

System::Windows::Forms::MethodInvoker((CurrencyMeter::Form1 ^)Globals::form,
&CurrencyMeter::Form1::UpdateDataForTickInTable));
351.

Globals::form->Invoke(gcnew

System::Windows::Forms::MethodInvoker((CurrencyMeter::Form1 ^)Globals::form,
&CurrencyMeter::Form1::UpdateListView));
352.

353.

B78

354.

numMessages++;

355.

Globals::textToUse = gcnew System::String("Receieved data frame

# " + numMessages);
356.

Globals::form->Invoke(gcnew

System::Windows::Forms::MethodInvoker((CurrencyMeter::Form1 ^)Globals::form,
&CurrencyMeter::Form1::UpdateStatusStrip));
357.

358.

359.
360.

DisconnectNamedPipe(hNamedPipe);

361.
362.

getchar();

363.
364.

return 0;

365.
366.

367.
368.

template <class T>double mean(int n, T *data)

369.

370.

double m=0;

371.

int i=n/8;

372.

while (i>0)

373.

374.

m += data[0];

375.

m += data[1];

376.

m += data[2];

377.

m += data[3];

378.

m += data[4];

379.

m += data[5];

380.

m += data[6];

381.

m += data[7];

382.

data += 8;

383.

i--;

384.

385.
386.

switch (n%8)

387.

388.

case 7: m+=data[6];

389.

case 6: m+=data[5];

390.

case 5: m+=data[4];

B79

391.

case 4: m+=data[3];

392.

case 3: m+=data[2];

393.

case 2: m+=data[1];

394.

case 1: m+=data[0];

395.

396.
397.
398.

return m/n;
}

399.
400.

template <class T>double pearson(int N, T *data, T *data1)

401.

402.

double med = mean<T>(N, data),

403.

med1 = mean<T>(N, data1),

404.

rez = 0, rez1 = 0, rez2 = 0;

405.
406.

for(int i=0; i<N; i++)

407.

408.

rez+=(data[i]-med)*(data1[i]-med1);

409.

rez1+=(data[i]-med)*(data[i]-med);

410.

rez2+=(data1[i]-med1)*(data1[i]-med1);

411.

412.
413.
414.

return rez/sqrt(rez1*rez2);
}

415.
416.

double calculate_covariance_of_price_changes(string currency1, string currency2) {

417.

bool invert_currency1 = true;

418.

bool invert_currency2 = true;

419.
420.

double mean1 = 0;

421.

double mean2 = 0;

422.

double covariance = 0;

423.
424.

int ind1;

425.

int ind2;

426.
427.

std::map<std::string, int>::iterator it1;

428.

std::map<std::string, int>::iterator it2;

429.
430.

// check whether or not to invert

B80

431.

if(currency1 != "USD") {

432.

if((it1 = currency_pair_index_map.find(currency1 + "USD")) !=

currency_pair_index_map.end()) {
433.

invert_currency1 = false;

434.

ind1 = (*it1).second;

435.

} else {

436.

invert_currency1 = true;

437.

it1 = currency_pair_index_map.find("USD" + currency1);

438.

ind1 = (*it1).second;

439.
440.

}
} else {

441.
442.

ind1 = numSymbols;
}

443.
444.

if(currency2 != "USD") {

445.

if((it2 = currency_pair_index_map.find(currency2 + "USD")) !=

currency_pair_index_map.end()) {
446.

invert_currency2 = false;

447.

ind2 = (*it2).second;

448.

} else {

449.

invert_currency2 = true;

450.

it2 = currency_pair_index_map.find("USD" + currency2);

451.

ind2 = (*it2).second;

452.
453.

}
} else {

454.
455.

ind2 = numSymbols;
}

456.
457.
458.

for(int i = 0; i < NUM_BARS_TO_USE; i++) {

data1[i] = invert_currency1 ? (close[ind1]->at(i) - open[ind1]->at(i))

: (1 / (close[ind1]->at(i))) - (1 / (open[ind1]->at(i)));
459.

data1[i] = (data1[i] == data1[i]) ? data1[i] : 0;

460.

data2[i] = invert_currency2 ? (close[ind2]->at(i) - open[ind2]->at(i) )

: (1 / (close[ind2]->at(i))) - (1 / (open[ind2]->at(i)));
461.

data2[i] = (data2[i] == data2[i]) ? data2[i] : 0;

462.

data3[i] = invert_currency1 ? (high[ind1]->at(i) - low[ind1]->at(i)) :

(1 / (high[ind1]->at(i))) - (1 / (low[ind1]->at(i)));
463.

data3[i] = (data3[i] == data3[i]) ? data3[i] : 0;

464.

data4[i] = invert_currency2 ? (high[ind2]->at(i) - low[ind2]->at(i)) :

(1 / (high[ind2]->at(i))) - (1 / (low[ind2]->at(i)));

B81

465.

data4[i] = (data4[i] == data4[i]) ? data4[i] : 0;

466.

467.
468.

double output = (pearson(NUM_BARS_TO_USE, data1, data2) +

pearson(NUM_BARS_TO_USE, data3, data4)) / 2;

469.
470.

// if value is nan, return 0

471.

if(output != output) {

472.

return 0;

473.

} else {

474.

return output;

475.
476.

}
}

477.
478.

void calculate_strength(int n) {

479.

int numCurrencies = currencies.size();

480.

int i = 0; int j = 0;

481.

double covariance_matrix[MAX_NUM_SYMBOLS][MAX_NUM_SYMBOLS];

482.

double raw_strength_vector[MAX_NUM_SYMBOLS];

483.

double alpha = pow(0.5, 1.0/double(n));

484.
485.

double alpha_sum = 0;

486.
487.

for(int i = 0; i < MAX_NUM_SYMBOLS; i++) {

488.
489.

alpha_sum = alpha_sum + pow(alpha, i);

}

490.
491.

for(set<string>::iterator it1 = currencies.begin(); it1 != currencies.end();

it1++) {
492.

i = 0;

493.

for(set<string>::iterator it2 = currencies.begin(); it2 !=

currencies.end(); it2++) {
494.

if(*it1 == "USD" && *it2 == "USD") {

495.

covariance_matrix[i][j] = 1;

496.

} else if(*it1 == "USD" || *it2 == "USD") {

497.

covariance_matrix[i][j] = 0;

498.

} else {

499.

covariance_matrix[i][j] =

calculate_covariance_of_price_changes(*it1, *it2);
500.

B82

501.
502.

// i must always be incremented

503.

i++;

504.

505.
506.

// raw_strength vector starts at 0 and is incremented

507.

raw_strength_vector[j] = 0;

508.
509.

// run code to calculate strength vector

510.

// use alpha as exponential weighting factor

511.

for(int k = 0; k < NUM_BARS_TO_USE; k++) {

512.

raw_strength_vector[j] = raw_strength_vector[j] + data1[k] *

pow(alpha, k);
513.

514.
515.

// divide by weighting factor

516.

raw_strength_vector[j] = raw_strength_vector[j] / alpha_sum;

517.
518.

// j must always be incremented

519.

j++;

520.

521.
522.

double covariance_sum ;

523.

double current_strength;

524.
525.

for(int i = 0; i < numCurrencies; i++) {

526.

current_strength = 0;

527.

covariance_sum = 0;

528.

for (int j = 0; j < numCurrencies; j++) {

529.

current_strength = current_strength + raw_strength_vector[j] *

covariance_matrix[i][j];
530.

covariance_sum = covariance_sum + covariance_matrix[i][j];

531.

532.

processed_strength_vector[i] = current_strength * 10000 /

covariance_sum;
533.

534.
535.

i = 0; // outer counter loop

536.

for(set<string>::iterator it2 = currencies.begin(); it2 != currencies.end();

it2++) {

B83

537.

processed_strength_vector_map[*it2] = processed_strength_vector[i];

538.

i = i++;

539.
540.

}
}

541.
542.

void cleanup(HANDLE hNamedPipe) {

543.

if (hNamedPipe != INVALID_HANDLE_VALUE)

544.

545.

CloseHandle(hNamedPipe);

546.

hNamedPipe = INVALID_HANDLE_VALUE;

547.
548.

}
}

1. #pragma once
2.
3. #include "mql4ipc.h"
4. #include "globals.h"
5.
6. namespace CurrencyMeter {
7.
8.

using namespace System;

9.

using namespace System::ComponentModel;

10.

using namespace System::Collections;

11.

using namespace System::Windows::Forms;

12.

using namespace System::Data;

13.

using namespace System::Drawing;

14.
15.

/// <summary>

16.

/// Summary for Form1

17.

/// </summary>

18.

public ref class Form1 : public System::Windows::Forms::Form

19.

B84

20.

public:

21.

Form1(void)

22.

23.
24.

InitializeComponent();
}

25.
26.

protected:

27.

/// <summary>

28.

/// Clean up any resources being used.

29.

/// </summary>

30.

~Form1()

31.

32.

if (components)

33.

34.

delete components;

35.
36.

}
}

37.

private: System::Windows::Forms::TabControl^

38.

protected:

TabularView;

39.
40.

private: cli::array<System::Windows::Forms::ListViewItem^>^ listViewItemList;

41.

private: System::Windows::Forms::ImageList^ largeImageList;

42.

private: System::Windows::Forms::TabPage^

DataTab;

43.

private: System::Windows::Forms::TabPage^

GraphicalTab;

44.

protected:

45.
46.

protected:

47.
48.

protected:

49.
50.

protected:

51.
52.
53.

private: System::Windows::Forms::StatusStrip^

StatusStrip;

54.
55.

private: System::Windows::Forms::DataGridView^

56.

private: System::Windows::Forms::DataGridViewTextBoxColumn^

Symbol;

57.

private: System::Windows::Forms::DataGridViewTextBoxColumn^

Bid;

58.

private: System::Windows::Forms::DataGridViewTextBoxColumn^

Ask;

59.

private: System::Windows::Forms::DataGridViewTextBoxColumn^

Spread;

B85

dataGridView1;

60.
61.

private: System::Windows::Forms::ToolStripStatusLabel^

LoadingStatus;

62.

private: System::Windows::Forms::DataGridViewTextBoxColumn^

63.

private: System::Windows::Forms::ListView^

Strength;

listView1;

64.
65.
66.

private:

67.

// Implements the manual sorting of items by columns.

68.

ref class ListViewItemComparer: public IComparer

69.

70.

private:

71.

int col;

72.
73.

public:

74.

ListViewItemComparer()

75.

76.

col = 0;

77.

78.
79.

ListViewItemComparer( int column )

80.

81.

col = column;

82.

83.
84.

virtual int Compare( Object^ x, Object^ y )

85.

86.

return (Double::Parse((dynamic_cast<ListViewItem^>(y))->SubItems[ col ]>Text) - (Double::Parse((dynamic_cast<ListViewItem^>(x))->SubItems[ col ]->Text))) > 0.0;

87.
88.

}
};

89.
90.

private:

91.

/// <summary>

92.

/// Required designer variable.

93.

/// </summary>

94.

System::ComponentModel::Container ^components;

95.
96. #pragma region Windows Form Designer generated code
97.

/// <summary>

98.

/// Required method for Designer support - do not modify

B86

99.

/// the contents of this method with the code editor.

100.

/// </summary>

101.

void InitializeComponent(void)

102.

103.

this->TabularView = (gcnew

System::Windows::Forms::TabControl());
104.

this->DataTab = (gcnew System::Windows::Forms::TabPage());

105.

this->dataGridView1 = (gcnew

System::Windows::Forms::DataGridView());
106.

this->Symbol = (gcnew

System::Windows::Forms::DataGridViewTextBoxColumn());
107.

this->Bid = (gcnew

System::Windows::Forms::DataGridViewTextBoxColumn());
108.

this->Ask = (gcnew

System::Windows::Forms::DataGridViewTextBoxColumn());
109.

this->Spread = (gcnew

System::Windows::Forms::DataGridViewTextBoxColumn());
110.

this->Strength = (gcnew

System::Windows::Forms::DataGridViewTextBoxColumn());
111.

this->GraphicalTab = (gcnew System::Windows::Forms::TabPage());

112.

this->StatusStrip = (gcnew

System::Windows::Forms::StatusStrip());
113.

this->LoadingStatus = (gcnew

System::Windows::Forms::ToolStripStatusLabel());
114.

this->listView1 = (gcnew System::Windows::Forms::ListView());

115.

this->TabularView->SuspendLayout();

116.

this->DataTab->SuspendLayout();

117.

(cli::safe_cast<System::ComponentModel::ISupportInitialize^

>(

this->dataGridView1))->BeginInit();
118.

this->GraphicalTab->SuspendLayout();

119.

this->StatusStrip->SuspendLayout();

120.

this->SuspendLayout();

121.

//

122.

// TabularView

123.

//

124.

this->TabularView->Anchor =

static_cast<System::Windows::Forms::AnchorStyles>((((System::Windows::Forms::AnchorStyles::Top
| System::Windows::Forms::AnchorStyles::Bottom)
125.

| System::Windows::Forms::AnchorStyles::Left)

126.

| System::Windows::Forms::AnchorStyles::Right));

B87

127.

this->TabularView->Controls->Add(this->DataTab);

128.

this->TabularView->Controls->Add(this->GraphicalTab);

129.

this->TabularView->Location = System::Drawing::Point(0, 0);

130.

this->TabularView->Name = L"TabularView";

131.

this->TabularView->SelectedIndex = 0;

132.

this->TabularView->Size = System::Drawing::Size(563, 316);

133.

this->TabularView->SizeMode =

System::Windows::Forms::TabSizeMode::FillToRight;
134.

this->TabularView->TabIndex = 0;

135.

//

136.

// DataTab

137.

//

138.

this->DataTab->Controls->Add(this->dataGridView1);

139.

this->DataTab->Location = System::Drawing::Point(4, 22);

140.

this->DataTab->Name = L"DataTab";

141.

this->DataTab->Padding = System::Windows::Forms::Padding(3);

142.

this->DataTab->Size = System::Drawing::Size(555, 290);

143.

this->DataTab->TabIndex = 0;

144.

this->DataTab->Text = L"Tabular Data";

145.

this->DataTab->UseVisualStyleBackColor = true;

146.

//

147.

// dataGridView1

148.

//

149.

this->dataGridView1->AllowUserToAddRows = false;

150.

this->dataGridView1->AllowUserToDeleteRows = false;

151.

this->dataGridView1->Anchor =

static_cast<System::Windows::Forms::AnchorStyles>((((System::Windows::Forms::AnchorStyles::Top
| System::Windows::Forms::AnchorStyles::Bottom)
152.

| System::Windows::Forms::AnchorStyles::Left)

153.

| System::Windows::Forms::AnchorStyles::Right));

154.

this->dataGridView1->ColumnHeadersHeightSizeMode =

System::Windows::Forms::DataGridViewColumnHeadersHeightSizeMode::AutoSize;
155.

this->dataGridView1->Columns->AddRange(gcnew cli::array<

System::Windows::Forms::DataGridViewColumn^

>(5) {this->Symbol,

156.

this->Bid, this->Ask, this->Spread, this->Strength});

157.

this->dataGridView1->Location = System::Drawing::Point(3, 3);

158.

this->dataGridView1->Name = L"dataGridView1";

159.

this->dataGridView1->ReadOnly = true;

160.

this->dataGridView1->RowHeadersVisible = false;

161.

this->dataGridView1->Size = System::Drawing::Size(549, 284);

B88

162.

this->dataGridView1->TabIndex = 0;

163.

//

164.

// Symbol

165.

//

166.

this->Symbol->HeaderText = L"Symbol";

167.

this->Symbol->Name = L"Symbol";

168.

this->Symbol->ReadOnly = true;

169.

//

170.

// Bid

171.

//

172.

this->Bid->HeaderText = L"Bid";

173.

this->Bid->Name = L"Bid";

174.

this->Bid->ReadOnly = true;

175.

//

176.

// Ask

177.

//

178.

this->Ask->HeaderText = L"Ask";

179.

this->Ask->Name = L"Ask";

180.

this->Ask->ReadOnly = true;

181.

//

182.

// Spread

183.

//

184.

this->Spread->HeaderText = L"Spread";

185.

this->Spread->Name = L"Spread";

186.

this->Spread->ReadOnly = true;

187.

//

188.

// Strength

189.

//

190.

this->Strength->HeaderText = L"Strength";

191.

this->Strength->Name = L"Strength";

192.

this->Strength->ReadOnly = true;

193.

//

194.

// GraphicalTab

195.

//

196.

this->GraphicalTab->Controls->Add(this->listView1);

197.

this->GraphicalTab->Location = System::Drawing::Point(4, 22);

198.

this->GraphicalTab->Name = L"GraphicalTab";

199.

this->GraphicalTab->Padding =

System::Windows::Forms::Padding(3);
200.

this->GraphicalTab->Size = System::Drawing::Size(555, 290);

B89

201.

this->GraphicalTab->TabIndex = 1;

202.

this->GraphicalTab->Text = L"Graphical View";

203.

this->GraphicalTab->UseVisualStyleBackColor = true;

204.

//

205.

// StatusStrip

206.

//

207.

this->StatusStrip->Items->AddRange(gcnew cli::array<

System::Windows::Forms::ToolStripItem^

>(1) {this->LoadingStatus});

208.

this->StatusStrip->Location = System::Drawing::Point(0, 319);

209.

this->StatusStrip->Name = L"StatusStrip";

210.

this->StatusStrip->Size = System::Drawing::Size(566, 22);

211.

this->StatusStrip->TabIndex = 1;

212.

this->StatusStrip->Text = L"statusStrip1";

213.

//

214.

// LoadingStatus

215.

//

216.

this->LoadingStatus->Name = L"LoadingStatus";

217.

this->LoadingStatus->Size = System::Drawing::Size(199, 17);

218.

this->LoadingStatus->Text = L"Waiting for connection from

client...";
219.

//

220.

// listView1

221.

//

222.

this->listView1->Anchor =

static_cast<System::Windows::Forms::AnchorStyles>((((System::Windows::Forms::AnchorStyles::Top
| System::Windows::Forms::AnchorStyles::Bottom)
223.

| System::Windows::Forms::AnchorStyles::Left)

224.

| System::Windows::Forms::AnchorStyles::Right));

225.

this->listView1->Location = System::Drawing::Point(0, 0);

226.

this->listView1->Name = L"listView1";

227.

this->listView1->Size = System::Drawing::Size(555, 290);

228.

this->listView1->TabIndex = 0;

229.

this->listView1->UseCompatibleStateImageBehavior = false;

230.

//

231.

// Form1

232.

//

233.

this->AutoScaleDimensions = System::Drawing::SizeF(6, 13);

234.

this->AutoScaleMode =

System::Windows::Forms::AutoScaleMode::Font;
235.

this->ClientSize = System::Drawing::Size(566, 341);

B90

236.

this->Controls->Add(this->StatusStrip);

237.

this->Controls->Add(this->TabularView);

238.

this->Name = L"Form1";

239.

this->Text = L"Form1";

240.

this->Load += gcnew System::EventHandler(this,

&Form1::Form1_Load);
241.

this->TabularView->ResumeLayout(false);

242.

this->DataTab->ResumeLayout(false);

243.

(cli::safe_cast<System::ComponentModel::ISupportInitialize^

this->dataGridView1))->EndInit();
244.

this->GraphicalTab->ResumeLayout(false);

245.

this->StatusStrip->ResumeLayout(false);

246.

this->StatusStrip->PerformLayout();

247.

this->ResumeLayout(false);

248.

this->PerformLayout();

249.
250.

251.

#pragma endregion

252.
253.

public: void UpdateStatusStrip() {

254.
255.

LoadingStatus->Text = Globals::textToUse;
}

256.
257.

public: void AddRowsToDataGridView() {

258.

for(int i = 0; i < Globals::numSymbols; i++) {

259.

dataGridView1->Rows->Add(gcnew

System::String(Globals::strarray[i]), gcnew System::String(L""), gcnew System::String(L""),

gcnew System::String(L""));
260.
261.

}
}

262.
263.

public: void AddSymbolsToListView() {

264.

listView1->View = View::LargeIcon;

265.
266.

// Create columns for the items and subitems.

267.

// Width of -2 indicates auto-size.

268.

listView1->Columns->Add("Currency", -2,

HorizontalAlignment::Left);
269.

listView1->Columns->Add("Strength", -2,

HorizontalAlignment::Left);

B91

>(

270.
271.

largeImageList = gcnew ImageList;

272.

largeImageList->ImageSize = System::Drawing::Size(64, 64);

273.
274.

// Initialize the ImageList objects with bitmaps.

275.

largeImageList->Images-

>Add(Bitmap::FromFile("C:\\\\currencymeter\\\\very_high_strength.png"));
276.

largeImageList->Images-

>Add(Bitmap::FromFile("C:\\\\currencymeter\\\\high_strength.png"));
277.

largeImageList->Images-

>Add(Bitmap::FromFile("C:\\\\currencymeter\\\\strength.png"));
278.

largeImageList->Images-

>Add(Bitmap::FromFile("C:\\\\currencymeter\\\\weak_strength.png"));
279.

largeImageList->Images-

>Add(Bitmap::FromFile("C:\\\\currencymeter\\\\weak_weakness.png"));
280.

largeImageList->Images-

>Add(Bitmap::FromFile("C:\\\\currencymeter\\\\weakness.png"));
281.

largeImageList->Images-

>Add(Bitmap::FromFile("C:\\\\currencymeter\\\\high_weakness.png"));
282.

largeImageList->Images-

>Add(Bitmap::FromFile("C:\\\\currencymeter\\\\very_high_weakness.png"));
283.

// largeImageList->Images-

>Add(Bitmap::FromFile("C:\\\\currencymeter\\\\no_indication.png"));
284.
285.

listView1->LargeImageList = largeImageList;

286.
287.

// Initialize the tile size.

288.

this->listView1->TileSize = System::Drawing::Size(128, 128);

289.
290.

listViewItemList = gcnew array<

System::Windows::Forms::ListViewItem^>(currencies.size());
291.
292.

int i = 0;

293.
294.

for(std::set<std::string>::iterator iter = currencies.begin();

iter != currencies.end(); iter++) {

295.

listViewItemList[i] = gcnew ListViewItem(gcnew

System::String((*iter).c_str()));
296.

listViewItemList[i]->SubItems-

>Add(System::Convert::ToString(0));

B92

297.

// listViewItemList[i]->ImageIndex = 8;

298.

i++;

299.

300.
301.

// Add the items to the ListView.

302.

listView1->Items->AddRange(listViewItemList);

303.

304.
305.

public: void UpdateListView() {

306.

int numCurrencies = currencies.size();

307.

int i = 0;

308.
309.

for(std::set<std::string>::iterator iter = currencies.begin();

iter != currencies.end(); iter++) {

310.

ListViewItem^ currentItem = listView1-

>FindItemWithText(gcnew System::String((*iter).c_str()));
311.

currentItem->SubItems[1]->Text =

System::Convert::ToString(processed_strength_vector[i]);
312.
313.

if(currentItem != nullptr) {

314.

double str = processed_strength_vector[i];

315.

if(str > 5) {

316.

currentItem->ImageIndex =

317.
318.

currentItem->ImageIndex =

319.

1;

} else if (str > 0.2) {

320.

currentItem->ImageIndex =

321.

2;

} else if (str > 0) {

322.

currentItem->ImageIndex =

323.

3;

} else if (str > -0.2) {

324.

currentItem->ImageIndex =

325.

4;

} else if (str > -0.5) {

326.

currentItem->ImageIndex =

327.

5;

} else if (str > -5) {

328.
329.

currentItem->ImageIndex =

6;

currentItem->ImageIndex =

7;

} else {

330.
331.
332.

0;

} else if (str > 0.5) {

}
}

333.

B93

334.

i++;

335.

336.
337.

this->listView1->ListViewItemSorter = gcnew

ListViewItemComparer( 1 );
338.
339.

340.
341.

public: void UpdateDataInTable() {

342.

for(int i = 0; i < Globals::numSymbols; i++) {

343.

std::string str1 = strarray2[i].substr(0, 3);

344.

std::string str2 = strarray2[i].substr(3, 3);

345.

dataGridView1->Rows[i]->Cells[0]->Value =

Globals::strarray[i];
346.

dataGridView1->Rows[i]->Cells[1]->Value =

System::Convert::ToString(bid_bar.at(i)->front());
347.

dataGridView1->Rows[i]->Cells[2]->Value =

System::Convert::ToString(ask_bar.at(i)->front());
348.

if(str2 != "JPY" && str1 != "JPY") {

349.

dataGridView1->Rows[i]->Cells[3]->Value =

System::Convert::ToString((ask_bar.at(i)->front() - bid_bar.at(i)->front()) * 10000);

350.

} else if (str2 == "JPY") {

351.

dataGridView1->Rows[i]->Cells[3]->Value =

System::Convert::ToString((ask_bar.at(i)->front() - bid_bar.at(i)->front()) *
1000);
352.

} else {

353.

dataGridView1->Rows[i]->Cells[3]->Value =

System::Convert::ToString((ask_bar.at(i)->front() - bid_bar.at(i)->front()) * 100000);

354.

355.

dataGridView1->Rows[i]->Cells[4]->Value =

System::Convert::ToString(processed_strength_vector_map[str1] processed_strength_vector_map[str2]);
356.

357.

dataGridView1->Update();

358.

359.
360.
361.

public: void UpdateDataForTickInTable() {

for(int i = 0; i < Globals::numSymbols; i++) {

362.

std::string str1 = strarray2[i].substr(0, 3);

363.

std::string str2 = strarray2[i].substr(3, 3);

B94

364.

dataGridView1->Rows[i]->Cells[0]->Value =

Globals::strarray[i];
365.

dataGridView1->Rows[i]->Cells[1]->Value =

System::Convert::ToString(bid_bar.at(i)->front());
366.

dataGridView1->Rows[i]->Cells[2]->Value =

System::Convert::ToString(ask_bar.at(i)->front());
367.

if(str2 != "JPY" && str1 != "JPY") {

368.

dataGridView1->Rows[i]->Cells[3]->Value =

System::Convert::ToString((ask_bar.at(i)->front() - bid_bar.at(i)->front()) * 10000);

369.

} else if (str2 == "JPY") {

370.

dataGridView1->Rows[i]->Cells[3]->Value =

System::Convert::ToString((ask_bar.at(i)->front() - bid_bar.at(i)->front()) *
1000);
371.

} else {

372.

dataGridView1->Rows[i]->Cells[3]->Value =

System::Convert::ToString((ask_bar.at(i)->front() - bid_bar.at(i)->front()) * 100000);

373.

374.

dataGridView1->Rows[i]->Cells[4]->Value =

System::Convert::ToString(processed_strength_vector_map[str1] processed_strength_vector_map[str2]);
375.

376.

dataGridView1->Update();

377.

378.
379.

private: System::Void Form1_Load(System::Object^

sender, System::EventArgs^

e) {

380.
381.

Globals::form = this;

382.
383.

pData =

(PCURRENCYDATA) HeapAlloc(GetProcessHeap(),

HEAP_ZERO_MEMORY, sizeof(PCURRENCYDATA));
384.
385.

hThread = CreateThread(

386.

NULL,

// default security attributes

387.

0,

// use default stack size

388.

getDataFromMql,

389.

pData,

// thread function name

// argument to

thread function
390.

0,

// use default creation flags

391.

&dwThread);

thread identifier

B95

// returns the

392.
393.

if( pData == NULL ) {

394.

// If the array allocation fails, the system is out of

memory
395.

// so there is no point in trying to print an error

message.
396.

// Just terminate execution.

397.

ExitProcess(2);

398.

399.
400.

if (hThread == NULL)

401.

402.

ExitProcess(3);

403.

404.

405.

};

406.

B96