12 Conservatism Flashcards
What is conservatism bias?
Conservatism = individuals are slow to change their beliefs in the face of new evidence. Psychological phenomena.
People do not take full advantage of the new information.
They do revise in the correct direction, but the change is not as large as Bayes rule suggest.
What is the consequence of conservatism?
Underreaction = individuals subject to conservatism might disregard the full information content of eg. An earnings (or some other public) announcemend and hence underreact.
Why:
(1) Perhaps because they believe that this number contains a large temporary component, and still cling at least partially to their prior estimates of earnings.
(2) Overconfidence about their prior information
Interpretation: tendency to underweight useful evidence relative to the less useful evidence used to form their priors.
What is the experimental evidence of conservatism bias?
Their updating is insufficient –> do not fully take into account all the infromation contained in the sample: 8 blue and 4 red.
Apparently –> people overweight the prior information 0.5 relative to the new information contained in the sample.
Explanations for insufficient updating –> anchoring, overconfidence and costly information processing.
A consequence: the stock market has the tendency to underreact to fundamental information eg. Dividend omissions, earnings reports.
What is the other experimental evidence of conservatism bias?
Estimate 8x7x6x5x4x3x2x1 vs. 1x2x3x4x5x6x7x8
Interestingly this initial value from which peole adjust can be:
1) Prior information that people have, like the ratio of red and blue balls in the urn example.
2) The result of a partial computation, as in our sequence example.
3) A value given by the formulation of the problem, however unrelated to the answers
What is underreaction?
Underreaction to good news announcement.
Underreaction = expected return on the company’s stock in the period following an announcement of good news is higher than the average return in the periuod following bad news.
Standard: good or bad news immediately reflected in price. NO implications for prices nor returns in next period.
Stocks underreact to good news, a mistake which is corrected in the following period, giving a higher return at that time.
The impact of the announcement is spread over time.
Because the stock underreacts to the actual announcement, profits can potentially be earned by trading in the stock after the news announcement.
What did Cutler look at?
Cutler et al. (1991): Speculative dynamics
They examine autocorrelations in returns on various indexes over different horizons.
Autocorrelation: each observation is statistically dependent on the previous one: why does the conservatism bias imply autocorrelation in returns.
The general picture: postive autocorrelations in excess index returns over hirzons of between one month and one year.
This autocorrelation evidence is consistent with the underreaction hypothesis.
Stock prices incorporate information slowly, leading to trends in returns over short horizons.
What did Jegadeesh and Titman (1994) examine?
They examined a cross-section of US stock returns and document evidence that over a 6 month horizon, stock returns are postively autocorrelated.
“Returns to buying winners and selling losers: implications for stock market efficiency”.
What is momentum?
Momentum in stock returns points to underreaction to information and slow incorporation of information into prices.
Momentum = empirically observed tendency for rising asset prices to raise further.
Momentum investing: a system of buying stocks or other securities that have had high returns over the past three to twelve months, and selling those that have had poor returns over the same period.
Jagedeesh and Titman (1993): trading strategies which buy 6-12 month winners and short sell past losers earn significant profits over the subsequent 6-12 month period
What is overreaction?
In addition, there is a lot of evidence suggesting that over longer horizons (3-5years), stock markets ‘overreact’ to consistent news pointing in the same direction.
After a series of good news announcements investors become overly optimistic sending the stock prices to unduly high levels.
Reference: Barberis et al (1998): a model of investor sentiment.
Overreaction associated with overconfidence, optimism bias and representativeness.
What are the empirical studies of overreaction?
Empirical studies of predictability of aggregate index returns over longer horizons are numerous: Fama & French (1988), Poterba & Summers (1988) –> over horizons from 3-5 years there seems to be negative autocorrelation in stock returns in many markets.
De Bondt and Thaler: 1985: portfolios of stocks with extremely low returns in the previous three years dramatically outperform portfolios of stocks with extreme high returns.
Chopra et al (1992): firms that had a sequence of bad earnings realizations subsequently outperformed (in terms of stock returns) firms with a sequence of good earnings.
How can short run underreaction and longer run overreaction be reconciled?
A model of investor sentiment that can accommodate both: over- and underreaction.
Single risk neutral investors that discounts future cash flow at delta.
Security pays out all earnings as dividends. True model of earnings is a random walk, yt is a random show to earnings that can take on two values +y and -y with equal probability.
Investor mistakenly believees that the earnings are generated by one of two models: model 1 and model 2.
In both models the value taken by yt only depends on the value taken by yt-1. Only difference between the two models lies in the transition probabilities.
What is the model for conservatism bias?
The expectations in this expression are the expectations of the investor who does not realize that the true process for earnigns is a random walk.
The price deviates from the correct value beaucse the investor does not use the random walk model to forecast earnings, but rather some combination of models 1 and 2, neither of which is a random walk.
Model of how investors form expectations of future earnings.
Claim: paper makes resonable, and empirically supportable, assumptions and derives testable empirical implications from these assumptions.
What conclusion can we draw from conservatism bias?
Conservatism bias: individuals are slow to change beliefs in the face of new information.
People do not takef ull advantage of all information.
Overweigh their priors and hence underreact to news, earnings announcement over short periods and overreact over longer periods.
This gives rise to profitable momentum strategies.
Momentum: buy stocks from past winners and sell stocks from past losers.
