A Commentary by Harley Bassman:

The Convexity Maven

Not a Product of Credit Suisse Research Value Concepts from the Credit Suisse Trading Desk
For Distribution to Institutional Clients Only April 17, 2013

“Skewered by Skew”

Walking past the derivatives trading desk the other day, my ears perked up as I
overheard snippets of a conversation focused upon the winning attributes of a
“Fat-Tail”. No, this was not a discussion about a Hollywood starlet, but rather an
earnest debate about how to model and evaluate the “Skew” of a disparate
package of options. In preview, this Commentary will detail the key drivers of
Skew and identify various trading opportunities. If you have inadvertently picked
up the wrong periodical, a quick click on TMZ.com should resolve this problem.

1
What is Skew ?

The Rosetta Stone of the options market is the famous Black-Scholes model,
created in 1973 at the University of Chicago. Given a set of inputs, the most
critical being the measure of Volatility, a “fair value” can be produced for an at-
the-money option. Most important for option market makers is the fact that
there exists an arbitrage free relationship between similar strike calls and puts;
this is known as Put::Call Parity.

Of the five inputs required by the model, four are transparent (Spot price, Strike
price, Interest Rate and Time) while the fifth, Implied Volatility, can be quite
ephemeral. That said, “ball-parking” an at-the-money (ATM) option price is
actually not that difficult (assuming you have a price history and an HP). Of all
the many risk vectors that drive an asset’s price, the most closely correlated to
its Implied Volatility is its recent Actual Volatility. The –shiraz line- below is the
Implied Volatility for a three-month into ten year straddle while the –vine line- is
the Actual Volatility for this rate over the trailing sixty business days. For this
series, the Implied Volatility averages about 7% over its paired Realized Volatility
and is generally within 10% of this spread. [See Commentary – Wall Street
Babylon, November 28, 2011 for a fuller discussion]

2
In fact, you will find the Implied Volatility on short-dated “gamma” options for
most tradable assets is usually about 10% above its matched Actual Volatility.
For the supremely active US Equity options market, the –aqua line- that charts
the Implied Volatility of three-month SPX options consistently traces slightly
above the –graphite line- of similarly dated Actual Volatility.

With this insight, a seasoned derivatives professional can make a fairly good
guess as to the price of an ATM option for most “liquid” assets; however, the
same cannot be said for options that are out-of-the-money (OTM). Thus we
introduce the concept of Skew. For our purposes, we will define Skew as the
difference between the Implied Volatility for an ATM option and a similar expiry
OTM option.

In general, there is no easy way to estimate what the Skew will be for a given
asset; the OTM Implied Volatility can be positive, negative, or even flat to its
ATM Implied Volatility. It must be understood that Put::Call Parity does not exist
for equidistant OTM options. So for instance, if there is a 10% positive skew for
a 25% OTM put option; this provides absolutely no information as to what
Implied Volatility might be appropriate for a 25% OTM call option. Put::Call
Parity only holds for options with the same strike and expiry date.

With this basic understanding under our belt, let’s explore what drives Skew.

Risk Preference

Humans are social beings who for either genetic or environmental reasons are
“risk averse”. A simple example of this notion is that most people feel more
emotional pain from losing $1,000 than they do pleasure from making a similar
sum. I will state this as a fact supported by many social science studies and I
shall not opine as to why this is the case. So let’s skip the math and examine
this concept heuristically.

3
Google is presently trading at $800 a share. Using exchange parameters, a
three-month ATM straddle costs about $68. The Implied Volatility of this price is
about 21.5%, which (not) coincidentally is nearly 10% above its three-month
Actual Volatility of 19.5%. Using coincident exchange parameters, a three-
month strangle with strikes of 700 and 900 would trade at $12; the Implied
Volatility here is a blended 23.0%, about 7% above the ATM level.

Source: Investopedia

So let’s examine these two trades. In both cases, the buyer can only lose the
premium paid with the potential for an unlimited gain. On the other hand, the
seller’s profit is limited to the fee received while he is exposed to a potentially
huge loss. The model is supposed to produce prices that will have comparable,
ex ante, Net Present Values, but the cold reality is that while the strangle does
have 100 points of “cushion” over the straddle, they both ultimately have the
possibility for “unlimited” loss. With due respect to the almighty, only an atheist
could opine that Infinity plus one is greater than Infinity. As such, since the
potential losses are similar (infinite) yet the strangle seller books a fee barely one
sixth as large as the straddle seller, a risk averse strangle seller will demand a
higher NPV, and thus its Implied Volatility is elevated above that of the straddle.
While slightly convoluted, this is the best non-mathematical way I can describe
the classic “smile” one sees for many option profiles. The diagram above is a
textbook “smile”.

Kurtosis

Option models tend to employ standard Normal or Lognormal distribution curves

for ease and simplicity. The problem is that reality is not quite so tidy. As the
CEOs of most integrated financial institutions bitterly discovered in 2008, markets
tend to experience a “hundred year flood” about every seven years.
Consequently, the far left and right sides of the distribution tend to be “fatter” in

4
reality than expected under ideal theoretical circumstances. Kurtosis is the
mathematical name for measuring the plumpness of the tails in a distribution.

The figures below were created by statisticians in an attempt to contort a

standard distribution to better fit a sample. Ultimately, traders give up on the
“implicit” solution of squishing the distribution via tweaking the Kurtosis setting
to alter model option values. Instead, market convention is to employ an
“explicit” method of arching the Skew profile to change the Implied Volatility
input in such a manner that the model value more closely matches the market
price.

Source: Financial Planning – Body of Knowledge

Path Dependency (1): Speed

Path Dependency is the key differentiator in valuing a “static claim” and a

“contingent claim”. (That is geek-speak for pricing a bond versus an option on a
bond.) If a bond rises by 10% over a few weeks, you make a 10% profit no
matter how that move occurs. In contrast, the path makes a big difference in
option-land. An asset whose price rises by 1% a day for two weeks is a different
animal from one where the 10% price change occurs on a single day; speed
matters. Let’s consider how option pricing reflects this reality.

Unlike the Futures market where all longs must fully offset every short, the cash
markets do have a “net” position. Specifically, for listed Equities and liquid
Bonds, investors are massively net long. This is because the “shorts” are the
issuing corporations, US homeowners, or the US Government, who do not trade
their positions. This is why the MBS market is “net short Convexity”; investors
have sold the prepayment option to homeowners who do not “delta hedge” their
houses. [See Commentary – “The Convexity Vortex”, March 6, 2013]

Since everybody is long (in aggregate), a seller can only sell to someone who is
already long and is willing to become longer; this buyer often demands a
discount when markets are volatile. This partially explains why prices tend to

5
decline faster than they rise. The chart below details the fifty year history of
daily Equity returns. As shown, the left tail (large price declines) is greater than
the right tail (large price increases) as there are more extreme price drops than
similarly large price jumps.

Since short-dated “gamma” options are most impacted by the “speed” of a price
change (recall that both gamma and acceleration are second order derivatives),
investors who sell these types of options demand an extra risk premium. It is for
this reason that short expiry OTM equity options always exhibit a significantly
positive put skew. In the –chalkboard chart- below, the Implied Volatility of
25%-delta (OTM) puts has generally been about 21% above the Implied
Volatility of 50%-delta (ATM) puts for three-month options on the SPX Index.
SPX three-month expiry IVol of 25% delta put divided by 50% delta put Source: The Bloomberg

6
Path Dependency (2): Location

Path Dependency is a dual-pronged vector; now that we have examined the

consequence of speed, we must now consider the implication of direction. It is
for this reason that option modelers repeat a similar mantra to that of real estate
brokers: Location, location, location.

When pricing an OTM option, the trader/investor must consider not only the
probability of reaching the strike (before expiry), but also what will be the
Implied Volatility for an ATM once one arrives there. Since option trades are
frequently closed out before expiry (and almost always marked-to-market), this
is not a trivial concept. If one buys a nine-month option whose strike price is
20% away, superior investors will have a reasonable notion as to what ATM
Implied Volatility will be if that strike is reached in only three months.

In the equity markets, there is a logic that underpins the skew profile for short to
mid-dated options. If the S&P rises by 100 points, about 7%, over the next few
months, that would certainly imply happiness and goodness in the world; a world
with less risk and fear. As such, one would expect ATM Implied Volatility (such
as the VIX) to decline. It is for this reason that OTM calls usually exhibit a
negative skew. Rare is the day that the stock market rises on an expectation of
greater economic risk.

Alternatively, fear and loathing almost always accompany lower prices; stock
prices pull back in anticipation of bad news. It is almost a given that lower
prices are associated with risk and worry, the main ingredients for higher option
prices.

Instead of a “smile”, the Skew profile for equity options tends to form a “smirk”
where the left side of the -raspberry line- is tipped up and the right side is
initially slanted down. The high correlation of Implied Volatility to location
dominates the other factors.

7
The consistently “smirked” skew for equities is not replicated in the debt
markets. Like most urban cities in the US, where central downtown has cycled
between au courant and slum as the demographic cycle flows, relative strike
locations swing between chic and shabby.

At the dawn of liquid option trading on Fixed Income securities, higher interest
rates were “bad” and lower rates were “good”. Higher rates (lower prices)
meant either Inflation or Liquidation, both negative for Total Rate of Return.
With Rates averaging 7% throughout the 1990s, a 5% rate would imply a stable
economy and higher stock prices while a 9% rate would mean a return of the
dreaded scourge of inflation. As such, OTM puts traded at a higher Implied
Volatility than equidistant OTM calls. [For quants, they traded fully Lognormal.
See Commentary – “Wall Street Babylon”]

Fast forward to early 2007, the eve of the Financial Crisis; Interest Rates were
circling around 5%. In this case, higher rates would confirm that the FED’s
tighter policy was indeed supported by a strong economy and that the housing
market would experience only a shallow retracement; higher rates would imply
less risk and lower Implied Volatility. However, much lower rates could only
occur if the most dire predictions of a housing bubble were true. Since the last
time national housing prices declined was during the Great Depression; fear
surrounded the prospect of lower rates which was reflected in the price of Skew.

In the chart above, the –kiwi line- is the Implied Normal Volatility for a six-month
into ten year 100bps OTM payer (put) minus the Implied Volatility for a 100bps
OTM receiver (call). The –huckleberry line- is the Sw10yr Rate. Notice how the
OTM call option became much dearer as the housing collapse gained steam in
late 2007. The market recognized that lower rates could only occur in a more
desperate environment; a world of much higher Implied Volatility.

8
Presently, lower yields would be concurrent with the endless boredom, and low
Volatility, of QE~ (infinity). On the flip side, consensus seems to be that a 3%-
handle on the Sw10yr Rate could only occur if the FED strongly hinted at the
prospect that they may soon scale back their buying program. Ipso, facto, a 3%
ten-year could only occur in a more uncertain environment as market forces
would be released from the FED’s repressive grip and would once again
independently manage the risk transfer process. Consequently, it should be no
surprise that put versus call skew is near its steepest level ever.

Supply versus Demand

No discussion of asset prices can ignore consideration of Supply versus Demand

dynamics. So briefly:

1) The Equity market almost always exhibits a “smirk”. Fundamental

considerations aside, there are large and well established option overlay
programs that employ “collars” and “covered calls”. As such, market
makers are almost always overly long OTM calls versus short OTM puts.
2) At one time, the Debt markets had a similar “smirk” profile. This was
reduced when the S&L crisis essentially ended “covered call” selling by
financial institutions. Regulatory changes, specifically in the form of FASB
122, converted MBS Servicing Rights from an off-balance sheet intangible
to an on-balance sheet asset. In tandem, these two “accounting” changes
have balanced out the supply::demand structure for Rates products.
3) Commodities and FX have relatively balanced flows, as such, the Skew
profile of these markets tend to most often exhibit the classic “smile”.

Trading Opportunities

Respect the MBS market

The largest participant in the USD options market is the Mortgage Servicing
Rights (MSR) business. And while the FED’s policy of Financial Repression has
temporarily dampened their hedging activity, the underlying of gross risk of MSR
has only increased in the past few years. Current option prices do not fully
indicate an appreciation of how the MBS market will react to higher interest
rates.

Buy 100mm 5 year into 5 year payer K = 3.65% 98nv

Sell 100mm 5 year into 5 year payer K = 4.50% 106nv
Sell 100mm 5 year into 5 year payer K = 5.50% 116nv

Enter trade for Zero cost; flat delta, flat gamma, positive carry.

9
While this trade looks like a “classic option ladder” with a maximum payoff
between 4.50% and 5.50% and a lower bound breakeven of 6.35%, the plan
here is be out of the trade well before expiry with a rather tidy profit.

The 5yr-5yr rate is presently at 3.00% while the Par MBS rate is near 2.50%. A
125bps rise in rates, a notion that is not totally insane since that would only push
the cash ten-year to 2.95%, would also make FN 3s a discount bond and FN 3.5s
a current coupon security. The FED’s policy of Financial Repression has “re-
couponed the stack” such that the point of Maximum Convexity (the Convexity
Vortex) has been repositioned.

Assuming a parallel shift, the –redwood line- 5yr-5yr would be at 4.25% and the
3.65% strike payer option you are long would be 60bps in-the-money. As per
Put::Call Parity, this option is similar in Volatility to a 60bp OTM call option.

So here is the magic: (1) Since rates would now be on the other side of the
Convexity Vortex, MSR hedgers will need to buy OTM calls, not puts. As such,
the 3.65% strike will not be negatively skewed by 8nv to the 4.50% strike you
are short. (2) Additionally, since the 5yr-5yr has historically rotated around the
–leaf line- of 5.00%, speculators will not be willing to pay a 10nv premium
between the 4.50% strike and the 5.50% strike. Ultimately, there will NOT be
an 18nv Skew across this ladder. The big profits will not be earned from delta or
theta, they will be booked as vega profits when the Skew profile compresses and
ultimately inverts.

It is never different this time

Until the FED implemented ZIRP and yanked the Fed Funds rate to zero, the
ratio of the Implied Volatility of 6m-5yr versus 6m-10yr averaged 105% and
rarely breached the parity barrier. This is a long way from the current ratio of
70%. While it is unclear when rates will rise, what we can almost guarantee is
that on that distant date when rates move significantly higher, this ratio will once
again return to a premium.

10
Buy 100mm 18 month into 5yr Payer K = 2.25% 79nv
Sell 53mm 18 month into 10yr Payer K = 3.26% 88nv

Enter trade for Zero cost with a 90% Volatility Ratio

Below, the –azure line- is he Implied Volatility ratio of 6m-5yr versus 6m-10yr
while the –clementine line- is the Sw10yr rate. Critical concepts:

1) The ratio was almost always above 1.00 until ZIRP;

2) Even during ZIRP, the ratio expanded when rates jumped;
3) A Sw10yr rate above 3.26% was associated with a ratio above 1.00

While this trade might look like your standard “conditional bear flattener”, the
magic here is that you can enter the trade at a 90% ratio at a strike level where
the ratio should be above parity. The plan here is to book at least a 10nv vega
profit as the Implied Volatility ratio expands with rising rates. Any curve profits
will be gravy.

Still the BEST trade for 2013

If I pound the table any harder on this trade, I will fracture my wrist.

If you have actually read the preceding ten pages, then you hopefully appreciate
the fact that unlimited loss is feared and that unlimited loss with only the hope
for a small gain tends to increase Skew. So why is it that the Skew for long-
dated equity options seems to completely reverse this logic ?

11
Let’s conduct a thought exercise. If one buys $100mm of the S&P, the most one
can lose is $100mm. This scenario may be a bit servere, nonetheless, the loss is
limited. Now consider the potential outcome of shorting $100mm of the S&P.
The loss potential over a long period of time is substantially larger. I will not say
infinite, but two times to five times over a decade is not out of bounds.

So riddle me this: How is it that a five-year expiry 40% OTM put on the S&P, an
option on the limited loss side of the distribution, trades with a +25% Skew
premium to the ATM while an equidistant 40% OTM call, an option on the
UNLIMITED loss side of the distribution, trades at a -15% Skew discount ?

Buy 100mm five-year expiry call option, K = 2000

Sell 100mm five-year expiry put option, K = 1000
Enter the trade at Zero cost, Spot S&P ~~ 1575

Reality Check #1: How is possible that the dollar cost to own 28% of the payoff
distribution (the delta of the call) is the same as the dollar cost of 15.8% of the
distribution (the delta on the put) ?

Reality Check #2: For the same dollar price, you can be long an unlimited gain
option struck 425 points away versus being short a limited loss option struck 575
points away.

Reality Check #3: The -robin’s egg line- above is the Skew profile of five-year
expiry S&P options while the –over-ripe avocado line- is the “forever” average of
the VIX Index. Implicit in this trade, one is buying an unlimited gain “tail option”
for a 10% lower Implied Volatility than the lifetime average. This is almost silly.

Since outright execution of this trade may be problematic, we suggest you

consider executing this trade as an “asset substitution” for your “beta exposure”
to the equity market. Let us detail the superior characteristics of this strategy.

12
Conclusion

While trading duration (or any asset with a linear return) is not easy, at least you
only have to manage one risk vector. The added complication of trading options
is that the risk profile now has more than a single dimension.

The key message here is not that everyone should become an option trader, but
that there is significant information embedded in the skew profile of an asset.
Often is the time that Skew will rotate in advance of a large market move.

Both our CS Rates and Equity strategy teams employ various Skew constructions
as key indicators to linear risk. Even Barron’s quotes Skew on the VIX in their
weekly options column. As such, Skew can be used either directly in trading or
indirectly in risk construction to increase portfolio returns.

We welcome your questions on this topic. Please call your CS representative.

Harley S. Bassman
Credit Suisse US Rates Trading
April 17, 2013

Clear Notice: This Commentary reflects the view of its author and is independent of CS Strategy.

This material has been prepared by individual sales and/or trading personnel of Credit Suisse AG or its subsidiaries or affiliates (collectively "Credit
Suisse") and not by Credit Suisse's research department. It is not investment research or a research recommendation for the purposes of FSA rules as it
does not constitute substantive research. All Credit Suisse research recommendations can be accessed through the following hyperlink: https://s.research-
and-analytics.csfb.com/login.asp subject to the use of approved login arrangements. This material is provided for information purposes, is intended for
your use only and does not constitute an invitation or offer to subscribe for or purchase any of the products or services mentioned. Any pricing
information provided is indicative only and does not represent a level at which an actual trade could be executed. The information provided is not
intended to provide a sufficient basis on which to make an investment decision. Credit Suisse may trade as principal or have proprietary positions in
securities or other financial instruments that are the subject of this material. It is intended only to provide observations and views of the said individual
sales and/or trading personnel, which may be different from, or inconsistent with, the observations and views of Credit Suisse analysts or other Credit
Suisse sales and/or trading personnel, or the proprietary positions of Credit Suisse. Observations and views of the salesperson or trader may change at any
time without notice. Information and opinions presented in this material have been obtained or derived from sources believed by Credit Suisse to be
reliable, but Credit Suisse makes no representation as to their accuracy or completeness. Credit Suisse accepts no liability for loss arising from the use of
this material. Nothing in this material constitutes investment, legal, accounting or tax advice, or a representation that any investment or strategy is suitable
or appropriate to your individual circumstances. Any discussions of past performance should not be taken as an indication of future results, and no
representation, expressed or implied, is made regarding future results. Trade report information is preliminary and subject to our formal written
confirmation.

CS may provide various services to municipal entities or obligated persons ("municipalities"), including suggesting individual transactions or trades and
entering into such transactions. Any services CS provides to municipalities are not viewed as “advice” within the meaning of Section 975 of the Dodd-
Frank Wall Street Reform and Consumer Protection Act. CS is providing any such services and related information solely on an arm’s length basis and
not as an advisor or fiduciary to the municipality. In connection with the provision of the any such services, there is no agreement, direct or indirect,
between any municipality (including the officials, management, employees or agents thereof) and CS for CS to provide advice to the municipality.
Municipalities should consult with their financial, accounting and legal advisors regarding any such services provided by CS. In addition, CS is not acting
for direct or indirect compensation to solicit the municipality on behalf of an unaffiliated broker, dealer, municipal securities dealer, municipal advisor, or
investment adviser for the purpose of obtaining or retaining an engagement by the municipality for or in connection with Municipal Financial Products,
the issuance of municipal securities, or of an investment adviser to provide investment advisory services to or on behalf of the municipality.

13