THE PSYCHOLOGY OF THE STOCK MARKET

1
THE PSYCHOLOGY OF THE STOCK MARKET

• Werner De Bondt
• April 2003

2
OVERVIEW

• The challenge
• What is behavioral finance?
• Case study
• The quality of analysts forecasts

3
VALUATION

• Value is about the future.
• The future is uncertain.
•  
• The quality of judgment is critical.
•   

4
FORECASTING?

• Economic fundamentals
• Cash flows the opportunity cost of capital.
•  
• The state of the market
• The dynamics of crowd behavior.

5

• "A full understanding of human limitations will
  ultimately benefit the decision-maker more than
  will naïve faith in the infallibility of his
  intellect."
•  
• Paul Slovic
• Journal of Finance
• 1972

6

• What was forecasted
• did not happen.
• What happened
• was not forecasted.

7
MYTHS OF THE TECHNOLOGY BOOM

• Tech companies generate breathtaking gains in
  economic productivity. The exponential growth
  will accelerate over time. Therefore, these firms
  are good investments. There is no price too high
  for a good tech company."
• Prospects are more important than immediate
  earnings.
• Monopolies create unbeatable advantages.
• Tech companies are not subject to the business
  cycle.

8

• ..It is still fair to say that, on the whole,
  our eco-nometric models are at best very crude
  approxi-mations of the true economy. .. these
  models require large doses of judgment if they
  are to be useful to decision-makers. .. the
  strengths of intuitive forecasting complement the
  weaknesses of model-based prediction.
• Alan Greenspan
• 1990

9
WHAT IS BEHAVIORAL FINANCE?

• "The study of financial decision-making with the
  help of concepts borrowed from psychology."
•  
• Modern finance was conceived as a logical,
  normative model of an idealized decision-maker.
  It assumes self-interested utility maximization,
  rational expectations, and optimal information
  processing.
• Herbert Simon (1983) calls this approach the
  Olympic model.

10
PEOPLE ARE HUMAN

• The logic of choice is not an adequate foundation
  for a descriptive theory of decision making.
• The deviations from the normative model are
  widespread, systematic, and fundamental.
• gt We see many anomalies and we need to develop
  new, better, descriptive theories.

11
XENOPHON, OECONOMICUS, 390 B.C.

• ISCHOMACHUS
• ..to a careful man, no business gives quicker
  returns than farming..
• My father taught me that .. For he never allowed
  me to buy a piece of land that was well farmed
  but pressed me to buy any that was uncultivated
  and unplanted owing to the owner's neglect or
  incapacity. Well farmed land, he would say,
  costs a large sum and can't be improved.
• .. Nothing improves more than a farm that is
  being transformed from a wilderness into fruitful
  fields. I assure you, Socrates, that we have
  often added a hundredfold to the value of a farm
  ..

12

• SOCRATES
• .. Did you father keep all the farms that he
  culti-vated, Ischomachus, or did he sell when he
  could get a good price?
• ISCHOMACHUS
• He sold, of course, but, you see, due to his
  indus-trious habits, he would promptly buy
  another that was out of cultivation ..
• SOCRATES
• Of course, and I declare, Ischomachus, on my oath
  that I believe you, that all men naturally love
  whatever they think will bring them profit.

13
SINCE 1975

• David Dreman
•  Paul Slovic, Amos Tversky, Daniel Kahneman
• Robert Shiller
• Werner De Bondt and Richard Thaler
• Hersh Shefrin and Meir Statman
•  Robert Haugen
• Reinhart Selten
•  After 1990, gt25 other researchers

14
HOW DO PEOPLE THINK?

• Bounded rationality.
•  
• The sources of knowledge
• Experience, logic, authority.
•  
• 1. Mental frames.
• (Simple conceptual models)
•   
• 2. Heuristics.

15
MENTAL FRAMES

• Affect choice
•  
• Can be managed
•  
• Are socially/professionally shared
•  
• Differ in sophistication
•  
• Do not change easily

16
HEURISTICS

•  Anchoring and adjustment.
• (First impressions matter.)
•  
• Availability.
• (Vivid information has greater impact.)
•  
• Representativeness.
• (Intuitive logic is distorted by stereotypes.)

17
LINDA

• Linda is 31 years old, single, outspoken, and
  very bright. She majored in philosophy. As a
  student, she was deeply concerned with issues of
  discri-mination and social justice, and also
  participated in anti-nuclear demonstrations.

18

• What is most likely? What is least likely?
• 1. Linda works in a bookstore and takes Yoga
  classes.
• 2. Linda is a bank teller.
• 3. Linda is a bank teller and is active in the
  feminist movement.

19
CENTRAL INSIGHTSOF BEHAVIORAL FINANCE

• 1. Intuition is fragile. Basic investment and
  statistical principles are not learned from
  everyday experience.
•   
• 2. Some decision errors may be avoided by
  education and by changing the decision process.
  Valuation is most successful if it is based on a
  structured, disciplined framework.
•  
• 3. Other people's errors may be our opportunity
  for profit, even if they create risk.

20
ISSUES

• Error vs. bias.
•   
• Experts vs. amateurs.
•  
•  Decision anomalies in the lab
• vs. financial anomalies in the real world
• vs. price anomalies in markets

21
ANALYST BIAS

• 1. Excessive optimism
•  2. Excessive use of stereotypes
• (Naive extrapolation. Overreaction)
• 3. Excessive confidence
• (Large errors)

22

• 4. Excessive rationalization
• (Missed turning points. Underreaction)
• 5. Excessive agreement
• (Conformism. Groupthink)

23
DO ANALYSTS OVERREACT?

• U.S. companies with December fiscal years, with
  EPS-data on Compustat, with returns on CRSP, with
  IBES consensus forecasts of EPS.
• 1- and 2-year forecasts, 1976-1984 (April)
  5-year forecasts of EPS-growth, 1982-1984
  (April).
• No survivorship bias of any kind.
• Source  De Bondt and Thaler (1990), De Bondt
  (1992).

24
DO ANALYSTS OVERREACT?

• 1. The forecast errors are large.
•  
• 2. The forecasts are too optimistic.
•  
• 3. The forecasts are too extreme. This problem is
  worse for long-term forecasts.
•  
• 4. Betting against the forecasts (especially,
  against long-term forecasts) is profitable.

25
INTERPRETATION

• The results are consistent with cognitive bias
  caused by representativeness.
•  
• The profitability of the investment strategy
  agrees with prior work on market overreaction.
  Here, the overreaction is to earnings
  information.

26
HERDING IN ANALYST FORECASTS

• U.K. companies, 1986-1997.
•  
• gt400,000 forecasts made by individual analysts, 0
  to 26 months before the month of the earnings
  announcement, provided by IBES.
•  
• Matching actual EPS data.
•  
• Source De Bondt and Forbes (1999).

27
HERDING IN ANALYST FORECASTS

• 1. The consensus forecast errors are large. On
  average, the forecasts are too extreme and too
  optimistic. Indeed, the "low" forecasts are more
  accurate than the consensus forecasts.
• 2. The range of forecasts is surprisingly narrow.
  Most actual EPS-numbers fall outside the range.

28

• 3. Disagreement among analysts does not increase
  a great deal as we get further away from the
  date of the earnings announcement.
• 4. Disagreement does not increase with the number
  of analysts (A), once A exceeds 7 or 8.

29
INTERPRETATION

• Herding depends on anticipated regret, i.e.,
  error potential. This potential increases with
  task difficulty (forecast horizon) and with the
  number of analysts following the stock.

30
THE GENIUS OF BENJAMIN GRAHAM(The Intelligent
Investor, 1959)

• "The market is always making mountains out of
  molehills and exaggerating ordinary vicissitudes
  into major setbacks." (p. 110)
•  
• "Our own records indicate that the interval
  re-quired for a substantial undervaluation to
  correct itself averages 1.5 to 2.5 years." (p.
  37)
•  
• "No one really knows anything about what will
  happen in the distant future but analysts and
  investors have strong views on the subject just
  the same." (p. 133)
•  

31
GOOD STORIES

• What are plausible stories that inspire
  confidence?
• Certainty of information
• Certainty of inference patterns
•  
• Do different patterns lead to the same or
  opposite conclusions?
•  
• There are no long chains in people's plausible
  reasoning. Knowledge about the world usually
  prevents absurd inferences.

32

• BEHAVIORAL PROBLEMS
•  
• 1. Shortcuts. Analogies. Categorization errors.
  Extrapolation.
• Value an ideamonopology renta business plan?
•  
• 2. Confirmation bias. Troubling facts are easily
  absorbed into existing mental frames.

33
LESSONS

• 1. The main value of an analyst's report lies in
  the point of view that it develops.
•  
• 2. Beware of story-stocks. If large changes in
  earnings are predicted, the actual changes will
  probably be less.
•  
• 3. "Low" forecasts of earnings tend to be more
  accurate than the consensus forecast.

34

• 4. Sell recommendations precede weak return
  performance. Buy recommendations are almost
  meaningless.
•  
• 5. The range of analyst EPS forecasts is too
  narrow relative to the range of outcomes.
•  
• 6. Invest contrary to analyst forecasts of
  long-run EPS growth.

35

• As a general rule, it is foolish to do just what
  other people are doing, because there are almost
  sure to be too many people doing the same thing.
•  
• William Stanley Jevons,
• Primer on Political Economy