08-09. Overconfidence

Question 1

What is overconfidence

Answer 1

Specification of how agents form expectation: one persistent finding: overconfidence.

Extensive evidence shows: people have the tendency to:
overestimate their skills and predictions for success
Perceive themselves more favorably than others perceive them
Perceive themselves more favorably than they perceive others.

People are overly optimistic = people believe that they are less likely to get hit by a bus or be robbed than their neighbors.

People are overconfident in their own abilities: new business owners believe their business has a 70M chance of success, but only 30% succeed.

Overconfidence and the stock market: overconfidence can lead to substantial losses when investors overestimate their ability to identify the next Amazon/Microsoft

Question 2

What are the most important facets of overconfidence?

Answer 2

Miscalibration
Positive illision = better-than-average effect
Illusion of control
Unrealistic optimism

Question 3

What is miscalibration?

Answer 3

Bias in which people are correct in their judgements far less often than they think they are.

Tendency of people to overestimate the precision of their knowledge.

Measurement (experimental) participants forecast: eg. The price of a stock and in addition are asked for 90% confidence intervals:
High and low forecasts such that there is a 5% chance that actual value of stock is above high forecast and a 5% chance that actual value is below low forecast.

Question 4

What is miscalibration in stock market forecasts?

Answer 4

Well calibrated belifs = on average the actual value will fall within the interval forecast 90% of the time.

Robust finding = average accuracy rate across studies well below predicted 90% = miscalibration.

When people are overconfident they set overly narrow confidence bands. High guesses too low and low guesses too high.

People are surprised more oftne than anticipated/

The intuition challenge: 90% of participants over-estimated the precision of their knowledge. Average surprise rate 4 out of 10. Miscalibration negatively correlated with points they earned.

Question 5

What did Ben-David discover?

Answer 5

2013: predictions made by senior finance executives (CFOs and financial vice presidents.
37 quarterly surveys between 2001-2010. 200-300 participants every quarter.

They have to make forecasting questions: over the next year, I expect the annual S&P500 return will be…

Important finding: miscalibtrated executives are also miscalibrated with regard to firm specific forecasts = eg. Internatl rate of return.

Question 6

What is positive illusion?

Answer 6

Unrealistically favourable attituted toward self.

Tendency of people to describe themselves as better-than-average.

Relative form of overconfidence.

Cognitive bias that causes people to overestimate their positive qualities and abilities and to underestimate their negative qualities, relative to others.

Bias has eg. Been observed among drivers, CEOs stock market analysts, college students, police officers and state education officials

Question 7

What is the illusion of control?

Answer 7

People tend to believe that they are able to influence events which in fact are governed mainly, or purely, by chance.

Example: experimental subjects have been induced to believe that they could affect the outcome of a purely random coin toss.

Interestinly: subjects who guessed a series of coin tosses more succesfully began to believe that they were actually better guessers than average.

Question 8

What is unrealistic optimism?

Answer 8

Unrealistic optimism towards the future or future life events.

Examples: persistent finding of unrealistic optimism in people’s estimates of the probabilities of (exogenous) future life events:
People judge the risk of positive events occuring to them as larger than for the average person, and the risk of negative events smaller

Question 9

Why is overconfidence so persistent?

Answer 9

1) Self attribution bias: tendency of people to attribute success to themselves and failure to others
2) Hindsight bias: tendency to think ‘I knew it all along’
3) Confirmation bias: tendency to search out evidence consistent with prior beliefs and to ignore conflicting data

Question 10

What is overconfidence in finance?

Answer 10

Some puzzles found in financial markets can be accounted for assuming overconfidence of investors.

Example: security misvaluations and excessive trading volumes

Furthermore, overconfidence has shown to have an effect on corporate decisions:

1) Merger and acquisition activities of corporations
2) Analyzes of internal corporate financing structures

Question 11

What is the overestimation of precision of private information

Answer 11

An investor is overconfident if he underestimates the error in his private information. If he interprets his signals overconfidently.

Overconfident investors treat their forecasts as more reliable information than they really are.

They underestimate the variance of the risky asset - overestimate the precision of their forecast. In extreme case c = 0 an investor believes that he knows the value of the risky asset with certainty.

Question 12

What did Odean show in 1998?

Answer 12

He showed that the more overconfident traders are (ie the lower c), the more different their belifs, which opens trading possibilities.

Trading volume increases as overconfidence increases.

Question 13

What is overconfidence and trading volume?

Answer 13

Individual demand for a security depends on investors private valuation –> eg. If an investor beliefs that the value of a security exceeds the market price, he will want to hold more of that security.

Many investors, assume all are price takers.

Investors use 2 pieces of information: vi=aivi + (1-ai)p, 0 ° captures overconfidence

The more overconfident an investor the more sensitive his demand to changes in the market price

Question 14

What is the evidence from surveys of overconfidence?

Answer 14

Often –> experimental and questionnaire studies.

For example: Glaser & Weber study a direct relation between investor overconfidence and trading volume.

This study is based on the combination of several datasets –> combination of naturally occuring data with information elicited from a survey.

Main dataset consists of 563,104 buy and sell transactions of 3079 individual investors from a German online broker in period January 1997-April 2001.

Second dataset consists of several demographic and other self-reported information collexted by the online broker at the time each investor opened account.

Third dataset consists of answers to an online questionnaire designed to elicit several measures of overconfidence.

Question 15

What was the online questionnaire with overconfidence?

Answer 15

All 3079 investors received an email from the online broker on Aug, 2, 2001, 129 investors answered in following week

Remaining group of investors received a second email on sept 20th, 2001, 86 investors answered in following week

Response rate of 7% = comparable to similar questionnaire studies.

Question 16

What were the different forms of overconfidence measured?

Answer 16

1) Better than average effect
Investors were asked questions concerning their skills and performance relative to others

2) Stock market forecasts:
Investors were asked to provide median as well as upper and lower bounds of 90% confidence intervals to 5 questions concerning stock market forecasts

Remember: statistically the percentage of surprises of well calibrated investors should be 10%. The mean percentage of surprises in their study was 75%.

They calculate the volatility forecasts of investors implied by their subjective confidence intervals.

Finding: subjective volatility forecasts are significantly lower than objective historical benchmarks.

Prices are more random than implied by the subjective forecasts: too n arrow confidence intervals.

In addition: they combine information on the trading behaviour of investors with their answers to the skills questions.

Question 17

What did they found out about better-than-average effect?

Answer 17

Finding = significant better than average effect and investors who think that they are above average do trade significantly more.

Question 18

What was the evidence from Field?

Answer 18

Barber & Odean (BO) analyse the trading history of more than 60.000 US discount brokerage investors between 1991-1996.

BO (2000) trading is hazardous to your wealth: the common stock investment performance of individual inveestors.

Starting point: they wanted to see whether trades were justified in the sense that they led to improvements in portfolio performance.

Suppose you sell one stock and you want to use the procees to buy another and in doing so incure 200 DKK in transaction costs.

Trade only justified if you expect to generate a higher portfolio return - high enough to offset the transaction costs.

In their data investors do a lot of trading: turn over 75% of their portfolios annually.

For a typical investors who holds a 100.000 portfolio, in a given year, he trades 75.000 worth of stocks. Those who were trading the least only turned over less than 3% of their portfolio per year, whereas those who traded the most turned over more than 300% per year.

So it was not worth while to trade so much.

Question 19

What did Barber & Odean test for overconfidence?

Answer 19

Barber & Odean: 2001.

Starting point of their approach: psychological research showing that in areas such as finance, men are more overconfident than women. Theory would predict that men trade more excessively than women

And excessive trading hurts performance.

Question 20

What were the hypotheses in Barber & Odean?

Answer 20

H1: men trade more than women

H2: by trading more, men hurt their performance more than women

Question 21

What was the data in Barber and Odean?

Answer 21

Primary data set fromo a large discount brokerage firm - investments of 78,000 households for 6 years ending in december 1996.

For this period: (i) end-of month position statements and (ii) trades that allow to estimate monthly returns.

Second data is demographic information compiled by Infobase and provided by the brokerage house.

These data identiefies the gender of the person who opened a household’s account for 37,664 households.

Question 22

What is the effect of overconfidence on return and turnover?

Answer 22

To evaluate the investment performance of men and women, the gross and net return performance of each houehold is calculated.

Net return performance is calculated after a reasonable accounting for costs = eg. Commissions

Monthly portfolio turnover for each household calculated taking into account sales and purchases during a month.

Portfolio turnover: share of the portfolio that is turned over each month.

Question 23

What is the effect of overconfidence on security selection?

Answer 23

Hypothesis: men will underperform women because men trade more and trading is costly.

Alternative cause of underperformance: inferior security selection.

Two investors with similar initial portfolios and similar turnover will differ in performance if one consistently makes poor security selections.

To measure security selection ability, they also compare the returns of stocks bought with those of stocks sold

Question 24

What are the conclusions on the 2 hypotheses?

Answer 24

Analysis shows, men trade more than women, and this difference is greatest between single men and women.

Furthermore, men lower their returns more through excessive trading than do women, and this difference is greatest between single men and comen.

The two ‘overconfidence’ hypothesis cannot be refuted.

Question 25

What are 2 alternative explanations instead of overconfidence?

Answer 25

1) Risk aversion
Since men ad women differ in both overconfidence and risk aversion, it is natural to ask whether differences in risk aversion alone explain findings.

While rational informed investors will trade more if they are less risk averse, they will also imporve their performance by trading.

But both groups hurt their performance by trading. And men do so more than women. Cannot be explained by differences in risk aversion alone.

2) Gambling
To what extent may gender differences in the propensity to gamble explain the differences in turnover and returns that we observe.

Trade for entertainment?

The average annual cost of trading for the quintile of most active traders is 2849, expensive entertainment