The Institutional Nature of Price Bubbles*

By SHEEN S. LEVINE and EDWARD J. ZAJAC

Price bubbles remain a puzzle for economic theory, particularly given their
appearance in experimental markets with high efficiency and minimized
uncertainty and noise. We propose that bubbles are caused by the
institutionalization of social norms, when individuals observe and adopt the
behavior of others. Explanations of bounded rationality or individual bias
appear insufficient as we show experimentally that (1) participants pricing
skills are better ex-ante than ex-post and (2) that individual discrepancies
between intrinsic values and market prices become increasingly serially
correlated during trading. We also find no support for the Greater Fool
explanation. (94 words)

JEL: B52 (Institutional), C91 (Laboratory, Individual Behavior), D41(Perfect

Competition), D53 (Financial Markets), D83 (Search; Learning; Information
and Knowledge; Communication; Belief), G12 (Asset Pricing; Trading
volume; Bond Interest Rates)

Markets sometimes develop price bubbles, i.e., they trade in high volumes at prices that
are considerably at variance from intrinsic values (Ronald R. King et al., 1993). Cases such as the
stock market crash of 1929 (Eugene N. White, 1990) demonstrate the enormous effect of
bubbles on individuals, firms, markets and even nations, and explain the interest they draw from
economists as well as the public (Charles MacKay, 1841; Charles P. Kindleberger, 1978). While
important in their consequences, the causes of bubbles are not well understood. Theoreticians
have suggested that bubbles may be rational (J. Bradford De Long et al., 1990; Peter M. Garber,
1990), intrinsic (Kenneth A. Froot and Maurice Obstfeld, 1991), and contagious (Richard Topol,
1991), but there is no widely accepted theory to explain their occurrence.
The existence of market bubbles seems at odds with common assumptions regarding the
efficiency of financial markets. Even more puzzling is the finding that bubbles occur not only in

*
Levine: Singapore Management University, 50 Stamford Road, Singapore 178899 (e-mail: [email protected]);
Zajac: Northwestern University, 2001 Sheridan Road, Evanston, Illinois 60208 (e-mail: e-
[email protected]). Robert Kurzban has contributed greatly through conversations and advice. Martin
Dufwenberg and Tobias Lindqvist have shared their instrument and advice with us most collegially. We also
benefited from the comments of Martin Conyon, Massimiliano Landi, Niro Sivanathan, and participants at the
Institutional Theory Conference at the University of Alberta, 2006 and the meeting of the American Economic
Association, Chicago 2007. Oi Ying Lam, Sun Li, and Junjie Tong provided able research support.
C 2007 Some rights reserved. You are free to copy, distribute, display, and make derivative works, as long as you
attribute the work to the authors and use it only in a non-commercial manner. For details, see Creative Commons
attribution-noncommercial 2.5 license.

Word count (including in-text citations and excluding appendixes): 2988 Last saved: 20-Jun-07 12:23:00

Electronic copy available at: http://ssrn.com/abstract=960178

real-world markets, where uncertainty and noise can lead to diverging rational expectations, but
also in highly predictable experimental markets (e.g. Vernon L. Smith, Gerry L. Suchanek and
Arlington W. Williams, 1988). In a typical laboratory study, participants engage in double-auction
trading of assets that are defined to have a finite lifespan and a known distribution of dividends.
Uncertainty is eliminated and participants should be able to calculate the intrinsic value of the
assets simply by examining the expected stream of dividends. Nevertheless, bubbles have been
observed repeatedly in experimental markets, even with sophisticated participants such as
business students, managers, and professional traders. Experimental bubbles have proven robust
to a variety of conditions, including short-selling, margin buying, equal portfolio endowment,
brokerage fees, the presence of informed insiders, dividend certainty, constant value, and limit
price change rule (Ronald R. King et al., 1993; David P. Porter and Vernon L. Smith, 2003).
A. Bounded rationality. While a complete explanation of experimental bubbles is still in
the making, research findings seem to imply that the phenomena may be the result of bounded
rationality, as evident in the behavioral versus rational debate (e.g. Colin F. Camerer, 1989;
Peter M. Garber, 1990; Eugene N. White, 1990; Thomas Lux, 1995). It has been shown that
bubbles abated when participants traded repeatedly within the same group (Ronald R. King et al.,
1993; Mark V. van Boening, Arlington W. Williams and Shawn LaMaster, 1993). This finding can
lead one to hypothesize that bounded rationality gives rise to bubbles in the short term as people
may have cognitive difficulty in applying a theoretical pricing model in practice (Herbert A.
Simon, 1955; Richard Cyert and James March, 1963). Arguably, initial mispricing may be
decreasing due to individual learning processes, which leads to improved pricing in subsequent
periods and an overall abatement of bubbles. If bubbles are indeed caused by bounded
rationality and alleviated by individual learning, then we should see indications that participants
improve their asset pricing skills as they trade. For instance, we would expect that average asset
prices will be farther from intrinsic values ex ante and move closer to intrinsic values ex post. We
would also expect that average price deviation from intrinsic value to be higher ex ante and that
price amplitude will be higher ex ante.
B. Greater Fool explanation. A second possible explanation is that while market
participants may be able to price correctly even initially, they tend to assume, wrongly, that they
can buy overpriced assets and sell them at even more inflated prices to others. Not commonly
addressed in the academic literature but prevalent among practitioners, the Greater Fool
explanation posits that bubbles are fueled by speculators who knowingly purchase overpriced
assets while hoping that they can sell those assets even more dearly to gullible investors, i.e.
greater fools (e.g. David Dreman, 1993; The Economist, 2003).

Electronic copy available at: http://ssrn.com/abstract=960178

For such a belief to have a cumulative effect on markets, a sufficiently large part of market
participant must be overconfident about their pricing acumen, believing that it is better than
others. Then, market participants will readily acquire overpriced assets, even if each individual
participant realizes that the assets are overpriced. Nevertheless, each individual participant
wrongly believes that enough others do not realize that those assets are overpriced and expects
that the others will be willing to buy them at a premium.
It is reasonable to expect the overconfidence necessary for the Greater Fool explanation.
A bias known as self-serving belief leads people assess themselves to be above average in various
positive characteristics (Ola Svenson, 1981; Linda Babcock and George Loewenstein, 1997).
People were shown to be generous in self-assessment versus the population average when asked
about their driving skills, managerial acumen, ethics, productivity, and other desirable
characteristics. Hence, if bubbles develop because participants subscribe collectively to a self-
serving belief, then we would expect to find that on average, each participant perceives her own
financial acumen to be superior to that of the others.
C. Institutionalization. It is also possible that bubbles stem not from individual biases,
but from the institutionalization of social norms. Economists argued that individuals observe
each other and base their decisions, at least partly, on imitation of others rather than on their
own cognition (Andrei Shleifer and Lawrence H. Summers, 1990). Some have predicted,
however, that such behavior would not occur in efficient markets, where sophisticated
participants can counter such herd behavior (Christopher Avery and Peter Zemsky, 1998). But
markets can also reward not behaviors that are intrinsically correct, but those that are correct in
matching what average opinion expects average opinion to be (John Maynard Keynes,
1936:156), thus amplifying pressures for institutional behavior. Further, economic sociologists
have shown how non-normative behavior, even if intrinsically correct, can have detrimental
financial and status consequences (Joel Podolny, 2005).
Institutionalization of social norms can lead to coordinated action through the
internalization of beliefs and interpretations of facts by each individual, even without formal
agreement or explicit discussion of coordination (e.g., folkways). While institutions can be
formal, such as the legal system, scholars have identified that institutions can also emerge to
provide order endogenously: Behavior becomes stable and patterned, or alternatively
institutionalized, not because it is imposed, but because it is elicited (Robert H. Bates et al.,
1998:8). Such processes have been shown to sustain practices across individuals and over time,
even when the practices were irrational. In an early laboratory experiment, Zucker (1977)
demonstrated how institutionalization facilitated the transmission of a practice that would have
clearly appeared counter-factual to an outside observer but not to the agents involved. Recently,
it has been shown that people learn to coordinate their price expectations without direct
communication and even when such coordination leads to erroneous outcomes (Cars Hommes
et al., 2005). Outside the laboratory, it was shown that US stock markets have systematically
responded favorably to (costless) announcements of stock buyback, although it was publicly
known that a significant proportion of the announcing firms did not act on their announcements
(James D. Westphal and Edward J. Zajac, 2001; Edward J. Zajac and James D. Westphal, 2004).
Economic sociologists have shown how traders of financial derivatives oscillated between
institutionalizing alternative approaches to asset pricing (Donald MacKenzie and Yuval Millo,
2003).
The three explanations should differ empirically in measurements of pricing skills and
self-serving bias ex ante and in the correlation of price discrepancy, i.e., the distance between the
market price of an asset and the intrinsic value of that asset. Pricing discrepancies can be
decomposed into dispersion and common components. If bubbles are caused by inapt pricing, then
one can expect indication of inadequate pricing skills ex ante and high dispersion component,
where market prices fall randomly above or below intrinsic value. If the Greater Fool
explanation holds, one can expect evidence of widespread self-serving bias. If bubbles are caused
by institutionalization, the common discrepancy component should be high and price
discrepancies should be correlated both among individuals and serially.

I. Method

We constructed an experimental double auction market (Vernon L. Smith, 1962), which

is known to possess extremely competitive characteristics (Charles A. Holt, 1995). In such
markets, each participant is endowed with experimental cash and assets, and he is free to post
bid and ask prices to buy and sell assets at will. The experimental market was programmed and
conducted in z-Tree (Urs Fischbacher, Forthcoming), based on the seminal design of Smith,
Suchanek and Williams (1988). To maintain consistency with prior work, we replicated a recently
published design (Martin Dufwenberg, Tobias Lindqvist and Evan Moore, 2005).
We recruited 62 undergraduate students with no prior experience in such experiments for
ten experimental sessions in what was described to them as a study of economic decision
making. Upon arrival to the experimental laboratory, participants received an instruction sheet
(Appendix 1a), which described the experiment and provided sufficient information for
calculation of intrinsic values. The participants then received a Price Questionnaire that probed
their knowledge of a standard asset-pricing model by asking, for instance, In the fourth period,
someone wants to sell you his stock. Write the maximum price you will be willing to pay for it.
The price questionnaire included 10 questions, one for each trading session, in random order
(Appendix 1b). The participants were told that they may consult the instruction sheet to answer
the questions, and were given 10 minutes to complete the questionnaire. All of them finished
within the time allocated.
After the Price Questionnaires were collected, the participants were asked to complete an
Assessment Questionnaire, which included questions designed to assess overconfidence. Each
participant was asked to provide 1) an assessment of the accuracy of his or her own responses, 2)
an assessment of the accuracy of the other participants responses, and 3) an assessment of the
other participants assessment of his or her own responses (Appendix 1c). This was followed by
a brief demographics questionnaire (Appendix 1d).
After completing the pre-trade questionnaires, the participants moved to a behavioral
laboratory, where they sat in separate cubicles in front of networked personal computers. Each
participant was randomly assigned to receive either two shares and 600 cents or six shares and
200 cents. They knew that their earnings would be paid to them in cash at the end of the
experiment. Once trading began, the participants could enter their minimum selling prices (ask)
and their maximum offers to buy (bid). Conditions resembled a highly efficient stock market: bid
and ask figures as well as eventual prices were visible to all of the participants and each one could
initiate a transaction by accepting a buying or selling offer. All transactions were anonymous and
participants could not communicate with each other.
Each experimental session consisted of six participants who traded for 10 periods lasting
120 seconds each. 1 At the end of each period, a dividend of 20 cents per share was payable with
a probability of 0.5. A summary screen appeared at the end of each period and presented
individualized trading and divided results. At the conclusion of the trading periods, the
participants received a $5 show-up fee and earnings in cash. Excluding the fee, earnings were
$13.38 on average (s.d.=2.87; range=$5.30-$18.30).

II. Results

In general, our results were similar to those obtained in prior work, i.e., we observed
bubbles in most of the experimental sessions (interestingly, we find this even though the Price
and Assessment Questionnaires could have affected behavior through psychological priming
effect). More significantly, we found that discrepancies between market prices and intrinsic
values were correlated among participants and became even more so over time. We found no indication
1 With the exception of the first session, which had eight participants.
that bubbles were caused by lack of knowledge. Quite the contrary; participants had a better
understanding of the theoretical pricing model ex ante, but somewhat astonishingly -- seem to
have partly abandoned that knowledge during trading. We also find no evidence of self-serving
bias.

[Figure 1 about here]

A. Ex-ante pricing skills. A comparison of average prices declared in the Price

Questionnaire with those obtained in actual trading revealed that, surprisingly, the prices
declared ex ante were better fit to intrinsic values. We used Haessel R2 (Walter Haessel, 1978) to
measure fit between the responses to the Price Questionnaire and intrinsic values and between
the trading prices and intrinsic values. We find better fit ex ante in nine out of 10 sessions. Figure
1 shows an illustration of the discrepancy between the prices declared ex ante and the actual
trading prices. Similarly, prices in trading also had higher normalized average price deviation in
nine out of 10 sessions. This statistic was calculated by summing the absolute deviation between
mean trading prices and intrinsic values for each period and dividing by the number of shares
outstanding. Prices in trading also had wider price amplitude in 10 out of 10 sessions. This
statistic was calculated by finding the highest and the lowest discrepancies between intrinsic
value and trading prices, subtracting the lowest from the highest and dividing by 100 (the initial
intrinsic value). These measures have been used frequently to assess the magnitude of bubbles
(Ronald R. King et al., 1993; Mark V. van Boening, Arlington W. Williams and Shawn LaMaster,
1993; David P. Porter and Vernon L. Smith, 2003; Martin Dufwenberg, Tobias Lindqvist and
Evan Moore, 2005).

[Table 1 about here]

B. Self-serving bias. We found no indication of self-serving bias. Results obtained from

the Assessment Questionnaire (Table 2) suggest that participants generally viewed their own
price assessment to be as precise as the others. The same is true about their assessment of
others perception of themselves they assumed that the others had a correct assessment of
their own capabilities. Measures of Cronbachs alpha (Lee J. Cronbach, 1951) show high
reliability for each group of questionnaire items.

[Table 2 about here]

 C. Decomposition of price discrepancy. The average discrepancy between a given
market price and the matching intrinsic value can be decomposed into two components:
dispersion and common (Cars Hommes et al., 2005). The former measures the deviation from a
common pricing method while the latter measures correlation between discrepancies in the
market. Formally, where t designates the trading period (110) and h is a counter of
transactions, the average price for period t is:

(1)
The average pricing discrepancy can be decomposed as:

1 1 1
10 10 10

(2)
The first term on the right-hand side of equation (2) measures the dispersion between
prices in market transactions. It is the squared distance between a price in a specific transaction
and the average price for that period, averaged across transactions and periods. The less similar
are prices in the market, the higher this component would be. The second term on the right-
hand side measures the common component. It is the squared distance between the average
price in a period and the intrinsic value for that period, averaged over periods. If price
discrepancies are serially uncorrelated and uncorrelated with the discrepancies of other
participants, this component will be zero. The higher the common component, the higher is the
similarity between the pricing discrepancies in the market.

[Table 3 about here]

Table 3 reveals that the common component plays a major role in the discrepancies
between market price and intrinsic values. It accounts for most of the variance in all sessions but
one. Moreover, as Figure 2 shows, the common component tends to increase during trading,
showing that errors became more correlated as trading progressed.

[Figure 2 about here]

 III. Discussion

The results suggest that price bubbles are not simply the result of individual cognitive
bias. We found that participants exhibited better understanding of intrinsic value pricing ex ante
than ex post. However, this understanding was apparently abandoned during trading, resulting in
worse fit to the fundamental model, higher price deviation, and higher amplitude ex post.
Because we have an indication of better knowledge ex ante, individual-level bounded rationality
alone cannot explain the deviations from the fundamental model. It is unlikely that time pressure
is responsible for the poorer performance during the trading session, because participants spent
much more time trading (20 minutes) than answering the Price Questionnaire (10 minutes).
Surveying the perception of the participants about their own pricing skills and those of
the others rules out the possibility of widespread self-serving bias. Taken together with prior
findings that documented bubbles even when speculation was not possible (Vivian Lei, Charles
N. Noussair and Charles R. Plott, 2001), these findings suggest little validity in the Greater Fool
explanation.
In contrast, a decomposition of price discrepancy shows how substantial the common
component is in such discrepancies and highlights the increasing similarity in discrepancy as
trading progressed. When this analysis is taken together with the evidence about better individual
pricing skills ex ante and the absence of self-serving bias, it seems to indicate that bubbles
originate from institutionalization of social norms about pricing.
As Douglas C. North suggested: We form mental models to explain and interpret the
environment[which] may be continually redefined with new experiences, including contacts with
others ideas (emphasis added, 1994:362-63). While prior work has acknowledged that people
observe and adopt others behavior, few have theorized or documented institutionalization in
highly efficient markets, as we do here (cf. Christopher Avery and Peter Zemsky, 1998). Further,
these results show that direct communication is not be necessary for institutionalization to
appear; the mere posting of bid and asks can be sufficient to spread beliefs and sway markets
away from intrinsic value. Hence, if institutionalization appears so rapidly and profoundly in
markets designed for high efficiency, it seems reasonable to presume an even greater
institutionalization in real-world markets, financial or others, with their inherent uncertainty,
lower efficiency, and direct communication between market participants.
The findings reported here allow us to rule out explanations that turn on individual
biases and focus on institutional processes instead. A potentially fruitful path would be to
investigate how social interaction leads to the spread and formation of shared beliefs in markets,
and whether small initial differences can lead to dramatically different outcomes, as was recently
shown in markets for cultural products (Matthew J. Salganik, Peter Sheridan Dodds and Duncan
J. Watts, 2006).
Avery, Christopher and Zemsky, Peter. "Multidimensional Uncertainty and Herd Behavior in
Financial Markets." American Economic Review, 1998, 88(4), pp. 724.
Babcock, Linda and Loewenstein, George. "Explaining Bargaining Impasse: The Role of
Self-Serving Biases." Journal of Economic Perspectives, 1997, 11(1), pp. 109-26.
Bates, Robert H.; Grief, Avner; Levi, Margaret; Rosenthal, Jean-Laurent and Weingast,
Barry W. "Introduction," R. H. Bates, A. Grief, M. Levi, J.-L. Rosenthal and B. W. Weingast,
Analytic Narratives. Princeton, New Jersey: Princeton University Press, 1998,
Camerer, Colin F. "Bubbles and Fads in Asset Prices." Journal of Economic Surveys, 1989, 3(1), pp.
3-41.
Cronbach, Lee J. "Coefficient Alpha and the Internal Structure of Tests." Psychometrika, 1951,
16, pp. 297-333.
Cyert, Richard and March, James. A Behavioral Theory of the Firm. Englewood Cliffs, NJ:
Prentice Hall, 1963.
De Long, J. Bradford; Shleifer, Andrei; Summers, Lawrence H. and Waldmann, Robert J.
"Noise Trader Risk in Financial Markets." Journal of Political Economy, 1990, 98(4), pp. 703-38.
Dreman, David. "One More for the Road?," Forbes. New York, 1993, 363.
Dufwenberg, Martin; Lindqvist, Tobias and Moore, Evan. "Bubbles and Experience: An
Experiment." American Economic Review, 2005, 95(5), pp. 1731-37.
Fischbacher, Urs. "Z-Tree: Zurich Toolbox for Ready-Made Economic Experiments."
Experimental Economics, Forthcoming.
Froot, Kenneth A. and Obstfeld, Maurice. "Intrinsic Bubbles: The Case of Stock Prices."
American Economic Review, 1991, 81(5), pp. 1189-214.
Garber, Peter M. "Famous First Bubbles." The Journal of Economic Perspectives, 1990, 4(2), pp. 35-
54.
Haessel, Walter. "Measuring Goodness of Fit in Linear and Nonlinear Models." Southern
Economic Journal, 1978, 44(3), pp. 648-52.
Holt, Charles A. "Industrial Organization: A Survey of Laboratory Research," J. H. Kagel and
A. E. Roth, Handbook of Experimental Economics. Princeton, New Jersey: Princeton University
Press, 1995, 349-443.
Hommes, Cars; Sonnemans, Joep; Tuinstra, Jan and Velden, Henk van de. "Coordination
of Expectations in Asset Pricing Experiments." Review of Financial Studies, 2005, 18(3), pp. 955-80.
Keynes, John Maynard. The General Theory of Employment, Interest and Money. London: Macmillan,
1936.
Kindleberger, Charles P. Manias, Panics, and Crashes: A History of Financial Crises New York:
Basic Books, 1978.
King, Ronald R.; Smith, Vernon L.; Williams, Arlington W. and van Boening, Mark V.
"The Robustness of Bubbles and Crashes in Experimental Stock Markets," R. H. Day and P.
Chen, Nonlinear Dynamics and Evolutionary Economics. Oxford, England: Oxford University Press,
1993,
Lei, Vivian; Noussair, Charles N. and Plott, Charles R. "Nonspeculative Bubbles in
Experimental Asset Markets: Lack of Common Knowledge of Rationality Vs. Actual
Irrationality." Econometrica, 2001, 69(4), pp. 831.
Lux, Thomas. "Herd Behaviour, Bubbles and Crashes." The Economic Journal, 1995, 105(431),
pp. 881-96.
MacKay, Charles. Memoirs of Extraordinary Popular Delusions and the Madness of Crowds. London R.
Bentley, 1841.
MacKenzie, Donald and Millo, Yuval. "Constructing a Market, Performing Theory: The
Historical Sociology of a Financial Derivatives Exchange." American Journal of Sociology, 2003,
109(1), pp. 107-45.
North, Douglass C. "Economic Performance through Time." American Economic Review, 1994,
84(3), pp. 359-68.
Podolny, Joel. Status Signals: A Sociological Study of Market Competition. Princeton, New Jersey:
Princeton University Press, 2005.
Porter, David P. and Smith, Vernon L. "Stock Market Bubbles in the Laboratory." Journal of
Behavioral Finance, 2003, 4(1), pp. 7-20.
Salganik, Matthew J.; Dodds, Peter Sheridan and Watts, Duncan J. "Experimental Study of
Inequality and Unpredictability in an Artificial Cultural Market." Science, 2006, 311(10 February
2006), pp. 854-56.
Shleifer, Andrei and Summers, Lawrence H. "The Noise Trader Approach to Finance." The
Journal of Economic Perspectives, 1990, 4(2), pp. 19-33.
Simon, Herbert A. "A Behavioral Model of Rational Choice." Quarterly Journal of Economics,
1955, LXIX, pp. 99-118.
Smith, Vernon L. "An Experimental Study of Competitive Market Behavior." Journal of Political
Economy, 1962, 70(2), pp. 111-37.
Smith, Vernon L.; Suchanek, Gerry L. and Williams, Arlington W. "Bubbles, Crashes, and
Endogenous Expectations in Experimental Spot Asset Markets." Econometrica, 1988, 56(5), pp.
1119-51.
Svenson, Ola. "Are We All Less Risky and More Skillful Than Our Fellow Drivers?" Acta
Psychologica, 1981, 47, pp. 143-48.
The Economist. "Of Manias, Panics and Crashes," 2003, 64.
Topol, Richard. "Bubbles and Volatility of Stock Prices: Effect of Mimetic Contagion." The
Economic Journal, 1991, 101(407), pp. 786-800.
van Boening, Mark V.; Williams, Arlington W. and LaMaster, Shawn. "Price Bubbles and
Crashes in Experimental Call Markets." Economics Letters, 1993, 41, pp. 179-85.
Westphal, James D. and Zajac, Edward J. "Decoupling Policy from Practice: The Case of
Stock Repurchase Programs." Administrative Science Quarterly, 2001, 46(2), pp. 202-28.
White, Eugene N. "The Stock Market Boom and Crash of 1929 Revisited." The Journal of
Economic Perspectives, 1990, 4(2), pp. 67-83.
Zajac, Edward J. and Westphal, James D. "The Social Construction of Market Value:
Institutionalization and Learning Perspectives on Stock Market Reactions." American Sociological
Review, 2004, 69(3), pp. 433-57.
Zucker, Lynne G. "The Role of Institutionalization in Cultural Persistence." American Sociological
Review, 1977, 42(5), pp. 726-43.
Appendix 1a: Participant Instructions
This is an experiment in economic decision making. The instructions are simple. If you follow
the instructions carefully and make good decisions, you will earn a considerable amount of
money. It will be paid to you in cash at the end of the experiment.
In this experiment, you will assume the role of a stock trader. During each period of the
experiment, you will be able to buy shares from the other participants or sell to them shares that
you own. You also might earn dividends on shares that you own. You will decide at which prices
you are willing to buy and sell, and how many shares to hold.
You will not know the identities of the other participants or how much money they earn or lose
during the experiment. Similarly, they will not know who you are or how much money you earn
or lose. Please do not speak with any other participants during this experiment.
The experiment will last for approximately an hour, including time for instructions and practice.
The Market
Six participants will trade in the experiment. Selected randomly at the beginning of the
experiment, half will receive 6 shares and 200 cents. The other half will receive 2 shares and 600
cents.
The experiment has 10 periods. In each period, you may buy or sell shares. The shares have a
lifespan of 10 periods, and your inventory of shares carries over from one trading period to the
next. Each period lasts for 120 seconds.
At the end of each trading period, each share may or may not pay a dividend. The dividend will
be randomly decided as 0 or 20 cents, with 50 percent chance for either outcome. That is, the
chance of receiving nothing is equal to the chance of receiving 20 cents. Thus, the average
dividend per period is 10 cents.
Your profits in the market will be equal to the total of the dividends that you receive on the
shares in your inventory at the end of each market period plus the cash you have at the end of
the experiment. Section 3 describes how to calculate your earning.

Earnings from Dividends

You can use the table to help you make decisions. There are five columns in the table. The first
column, labeled Ending Period, indicates the last trading period of the market. The second
column, labeled Current Period, indicates the period during which the average holding value is
being calculated. The third column gives the number of holding periods from the period in the
second column until the end of the market. The fourth column, labeled Average Dividend Value
per Period, gives the average amount that the dividend will be in each period for each unit held
in your inventory. The fifth column, labeled Average Holding Value per Unit of Inventory, gives
the expected total dividend for the remainder of the experiment for each unit held in your
inventory for the rest of the market. That is, for each unit you hold for the remainder of the
experiment, you can expect to receive, on average, the amount listed in column 5. The amount in
column 5 is calculated by multiplying the numbers in columns 3 and 4.
For example, suppose that there are 4 periods remaining. Since there is an equal chance of
receiving nothing or 20 cents, the dividend is on average 10 cents per share in each period. If you
hold a share for 4 periods, the total dividend paid on the share (over the 4 periods) can be
expected to be 410 = 40.
Ending Current Number of Average Dividend Average Earnings
Period Period Holding Periods Value per Period per Share

10 1 10 10 100
10 2 9 10 90
10 3 8 10 80
10 4 7 10 70
10 5 6 10 60
10 6 5 10 50
10 7 4 10 40
10 8 3 10 30
10 9 2 10 20
10 10 1 10 10

Calculate Your Earnings

Your earnings in each period equal the value of the dividends you receive each period for the
shares you hold at the end of the period. That is,
Your earnings for a period = Dividend per share number of shares held at the end of period
When you buy shares, the cost of purchase is deduced from your cash. When you sell shares, the
income is added to your cash. Your total earnings are the earnings in each one of the periods,
plus the amount of cash that you have at the end of period 10. That is
Your total earnings in the market =
Earnings for period 1 + Earnings for period 2 + Earnings for period 3 + Earnings for period 4
+ Earnings for period 5 + Earnings for period 6 + Earnings for period 7 + Earnings for period
8 + Earnings for period 9 + Earnings for period 10 + cash at the end of period 10.
The computer will display your earnings after each period.
There will also be a show up fee of $5 to all participants

5. The Trading Screen

Top
Remaining time (sec): number of seconds remaining until the current period ends.
Periodout of...: The number of the current period out of the total number of periods in the
experiment.
Cash: Your current cash amount (in cents)
Shares: The current number of shares that you hold.

Bottom

Make an Offer to Sell (in cents): Enter the amount that you demand for selling a single share.
Click the button to publish the offer. It will immediately appear in the Offers to Sell column and
will be visible to the other participants. The offer is not visible until you publish it by pressing
the button.

Buy: To buy a share, click on one of the offers in the Offers to Sell column, and click the Buy
button. Unless someone else was quicker in responding, you will receive one share and the cost
will be deducted from your cash.

Sell: To sell a share, click on one of the requests in the Requests to Buy column, and click the
Sell button. Unless someone else was quicker in responding, you will receive the amount
specified in the request and one share will be deducted from your holdings.
Make a Request to Buy (in cents): Enter the amount that you are willing to pay for a single
share. Click the button to publish the request. It will immediately appear in the Requests to Buy
column and will be visible to the other participants. The request is not visible until you publish it
by pressing the button.

Center

Offers to Sell Shows the available offers in descending price order, so that the lowest price is at
the bottom.
Transaction price column Shows all of the prices at which a share has been bought or sold in
the current period.
Requests to Buy Shows the available requests in ascending order, so that the highest price is at
the bottom.
Earnings Report
The earnings report appears at the end of each period. After seeing your earnings, press the
Next button to go to the next period. The next period will begin after all of the participants
press the Next button.
Appendix 1b: Price Questionnaire
Please write your participant number

Please take about 10 minutes to fill in this questionnaire. You may use the instruction sheet to
assist you.

1. In the 4th period, someone wants to sell you his stock. Write the maximum price you will
be willing to pay for it

2. In the 8th period, someone wants to sell you her stock. Write the maximum price you will
be willing to pay for it.

3. In the 1st period, write the minimum price you will be willing to sell a single stock for.

4. In the 3rd period, write the minimum price you will be willing to sell a single stock for.

5. In the 6th period, you are offered a single stock. Write the maximum price you will be
willing to pay for it.

6. In the 7th period, write the minimum price you will be willing to sell a single stock for.

7. In the 2nd period, you are offered a single stock. Write the maximum price you will be
willing to pay for it.

8. In the 5th period, someone offers to buy your stock. Write the minimum price you will be
willing to sell the stock for.

9. In the 9th period, someone offers to buy your stock. Write the minimum price you will be
willing to sell the stock for.

10. In the 10th period, write the maximum price you will be willing to buy a stock.
Appendix 1c: Self Assessment Questionnaire
Please write your participant number Please take about 5 minutes to fill in this questionnaire

1. How confident are you in your 1 2 3 4 5

answers to these questions? I am not I am very
confident confident
at all

2. How sure are you that you wrote 1 2 3 4 5

the right answers? I am not I am very
sure at all sure

3. In your assessment, how well do 1 2 3 4 5

you normally answer questions I normally I normally
do not answer
such as the ones on the previous answer very well
page? well at all

4. Compared to the other people in 1 2 3 4 5

the room, how precise are your My My
answers answers
answers? are much are much
less more
precise precise

5. If we were to rank the answers of 1 2 3 4 5

the people in this room, where do Worse Better
than all than all
you think your answers would other other
rank? answers answers

6. In your opinion, how confident are 1 2 3 4 5

the other people in the room in The others The others
are not are very
their answers to these questions? confident confident
at all

7. In your opinion, how sure are the 1 2 3 4 5

other people in the room that they The others The others
are not are very
wrote the right answers? sure at all sure

8. In your assessment, how well did 1 2 3 4 5

the other people in the room in The others The others
did not answered
answering the questions on the answer very well
previous page? well at all

9. In your assessment, how skillful 1 2 3 4 5

are the other people in the room in The others The others
are not are very
answering questions such as the skilful at skilful
ones on the previous page? all
10. Try to imagine what the other 1 2 3 4 5
people in the room think of you. They think They think
that my that my
In their opinion, how precise are answers answers
your answers? are not are very
precise at precise
all
11. Try to imagine what the other 1 2 3 4 5
people in the room think of you. They think They think
that my that my
In their opinion, how skillful are answers answers
you in answering questions such as are not are very
precise at precise
the ones on the previous page? all

12. In your opinion, how confident are 1 2 3 4 5

the others in the room about their They are They are
not very
impression of you? confident confident
at all in in their
their impression
impression of me
of me
Appendix 1d: Demographics Questionnaire

Please write your participant number

Please take about 3 minutes to fill in this questionnaire

a. Gender (circle): Male Female

b. Year of Birth:

c. Race:

d. Nationality:

e. Highest Level of Education:

If you have academic education or currently a university student, please answer the following:

a. Currently a student: (circle): Yes No

b. Year in School (circle): 1 2 3 4 Graduate Student

c. What was (is) your academic major (if undecided, write undecided):

d. Have you taken any classes in Finance or Economics? Please write their full titles below:
 Session
1a 2a 3a 4a 5a 6a 7a 8a 9a 10a
Haessel R2
Declared (D) 0.927 0.903 0.859 0.916 0.888 0.906 0.969 0.004 0.927 0.758
Trading (T) 0.552 0.962 0.691 0.698 0.069 0.747 0.896 0.000 0.810 0.525
T>D No Yes No No No No No No No No
Normalized Absolute Price Deviation
Declared (D) 0.356 0.279 0.779 0.675 0.454 0.392 0.288 0.793 0.452 0.456
Trading (T) 0.594 0.176 0.567 0.757 1.342 0.394 0.167 1.289 0.203 1.856
T>D Yes No No Yes Yes Yes No Yes No Yes
Normalized Average Price Deviation
Declared (D) 0.016 0.036 0.083 0.067 0.044 0.041 0.083 0.113 0.057 0.052
Trading (T) 0.142 0.040 0.120 0.131 0.162 0.092 0.034 0.204 0.039 0.295
T>D Yes Yes Yes Yes Yes Yes No Yes No Yes
Price Amplitude
Declared (D) 0.119 0.123 0.317 0.367 0.195 0.133 0.142 0.602 0.200 0.398
Trading (T) 0.678 0.149 0.560 0.443 0.786 0.479 0.186 0.785 0.395 1.329
T>D Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes

Table 1: Measures of Goodness of Fit, Price Deviation, and Price Amplitude for Average
Declared versus Trading Prices
Construct No. of Mean Standard Cronbachs
Items Deviation Alpha
Participants self-assessment 3 2.90 1.04 0.79
Participants assessment of the other participants 6 3.00 0.79 0.69
Participants assessment of the other 3 2.95 0.77 0.83
participants perception of him/her

Table 2: Items measured in Assessment Questionnaire

 Session Average Individual Error Average Dispersion Error Average Common Error

1 1 1
10 10 10

A1 1809.811 30.990 (2%) 1778.821(98%)

A2 128.689 17.068 (13%) 111.620 (87%)

A3 1324.818 129.751 (10%) 1195.067 (90%)

A4 1231.000 155.975 (13%) 1075.025 (87%)

A5 3160.076 981.072 (31%) 2179.003 (69%)

A6 590.917 19.452 (3%) 571.464 (97%)

A7 254.326 164.033 (64.5%) 90.292 (35.5%)

A8 3358.302 143.153 (4%) 3215.148 (96%)

A9 281.455 59.723 (21%) 221.731 (79%)

A10 11839.706 2569.031 (22%) 9270.669 (78%)

Table 3: Decomposition of the average individual pricing discrepancy into dispersion and
common components
Figure 1: Illustration of declared, intrinsic, and actual prices over trading periods
Legend (left to right): average trading price; average present value (intrinsic value); maximum present value possible
(intrinsic value); average declared value in Pricing Questionnaire.
Figure 2: Illustration of common component in price discrepancy over trading periods