Pragma Synesi – interesting bits

Compendium of interesting bits I come across, with an occasional IMHO

Rational financial decision making

We don’t make rational decisions, and that’s especially bad for us when it comes to our financial health.  The article below, from Benefits Canada, explores this:

Keeping feelings out of financial decision-making

September 23, 2010 | Sophie Mayrand

Behavioural finance, or the study of investor psychology, marries the psychology of decision-making with investors’ trading behaviour and asset pricing. Departures from rational decision-making emerge because individuals do not enjoy unlimited information-processing capabilities and instead rely on rules of thumb when making decisions under uncertainty. This leads to systematic cognitive errors and mispriced assets. But behavioural finance helps us to understand why certain market anomalies exist and persist.

Efficiencies and Inefficiencies
The efficient market hypothesis, introduced 30 years ago, argues that information is processed properly by investors and reflected in the price of assets instantaneously. By being rational, investors are expected to update their beliefs correctly upon receiving new information and, given those beliefs, make choices that are normatively acceptable to maximize their returns. Under this hypothesis, active management is unable to achieve superior returns to the market.

On the other hand, behavioural finance proponents acknowledge that investors are human beings who are subject to a range of emotions and information-processing errors that push them to make suboptimal trading decisions. This leads to mispriced assets and inefficient markets. Investors who understand and avoid these mistakes could exploit the resulting inefficiencies and end up earning excessive returns. With the discovery of market anomalies such as the value premium, the momentum effect and the post-earnings announcement drift, it is tempting to search for answers in human psychology to explain investors’ underreaction or overreaction.

Common Biases
David Hirshleifer (2001) believed most mistakes that individuals make can be traced to four common causes: self-deception (limits to learning), heuristic simplification (rules of thumb), emotions and social interaction.

Self-deception: Under this category, we observe overconfidence, the illusion of knowledge and the confirmation bias. How many of us are guilty of thinking we are better-than-average students, better-than-average drivers and better-than-average portfolio managers? While a majority of us think that way, the truth is, the math just does not work. We are simply overconfident.
The illusion of knowledge is the belief that the accuracy of our forecasts increases with more information. More information is not necessarily better information; it’s what you do with the information that matters. In fact, the dangerous thing about information overload is that it actually increases our confidence (overconfidence) but not necessarily our skills.

The confirmation bias is our habit of looking only for information that agrees with our pre-existing beliefs; it is our thirst for agreement. Because it feels good to have our opinions reflected back to us, we search for information that supports our views and ignore facts (or people) that conflict with our opinions.

Heuristics: With rules of thumb, individuals are subject to anchoring and representativeness. Anchoring is observed when someone’s initial expectations or prior probabilities are based on a starting point (an anchor), which is adjusted when new information arrives. While in theory, we should objectively consider the value of new information and arrive at a new predicted outcome, the fact is that we always start with an anchor first and then make adjustments. Investment professionals who entered the U.S. market in the mid-1990s could not believe the market was going anywhere but up before the turn of the century, whereas those who witnessed both the technology, media and telecom and the credit crunch crashes are a little more careful when assessing the relative valuation of the market nowadays.

Representativeness causes individuals to expect samples to be highly representative of the parent population from which they are drawn. People see trends and patterns in random events. They assign too much importance to extreme and strong new information and not enough to the information’s weight or statistical stability. For example, five spins of a roulette wheel ending on black does not mean red is due, just as five years of good earnings growth does not mean the company can sustain its competitive edge over the next five years. The sample is just too small to be highly representative.

Under the efficient market hypothesis, investors are rational decision-makers when faced with uncertainty. They weigh the value of each alternative separately, then choose the one with the highest expected return. On the other hand, in prospect theory—a positive model of judgment when faced with uncertainty—value is defined as a change in wealth around a reference point. For the same absolute amount, losses are felt with more pain than the corresponding joy of a gain. This is referred to as loss aversion. In addition, in the domain of gains, investors exhibit risk aversion. They prefer a certain gain over the possibility of a higher gain (e.g., they prefer a 100% chance of winning $3,000 over an 80% chance of winning $4,000).

In the domain of losses, however, investors become risk seekers. When faced with the certainty of a $3,000 loss or a higher but just as probable loss (80% probability of a $4,000 loss), they will go with the second option. The certainty effect gives more weight to the outcome, more joy for a certain gain and more pain for a certain loss. Interestingly enough, when probabilities are very low, the opposite occurs: humans become risk seekers for gains (what other explanation is there for purchasing lottery tickets?) and risk averse for losses (insurance premiums, anyone?).

Confidence Breaker

Information overload is not only overwhelming, it’s also dangerous. Too much information actually increases people’s confidence (overconfidence) but not necessarily their skills. This has been proven in Stuart Oskamp’s classic experiment (1965), showing how confidence increases as people gain more knowledge.

Thirty-two judges (psychologists and graduate students) were asked to read background information about a case of adolescent maladjustment, divided into four stages. After each stage, they had to answer 25 multiple choice questions involving personal judgments. The results confirmed the hypothesis that as more information was received by the judges, their confidence level would rise markedly and steadily. However, the accuracy of their conclusions would quickly reach a ceiling. (See pdf for figure)

Common Market Anomalies
Investors’ overreaction in extrapolating a firm’s recent earnings growth and pushing the stock’s price too high or too low has been linked to the representativeness bias. Based on relatively small samples of earnings releases, investors will tend to become too bearish on a stock that has been performing poorly, pushing its price down until it becomes relatively cheap and has a better chance, going forward, to outperform its peers (the value premium).

The momentum and earnings surprise anomalies are both seen as underreaction to news. Momentum is the serial correlation of returns that has been observed in various parts of the world, whereby past winners/losers outperform/underperform their peers in the following one to 12 months.

Earnings surprise is the post-earnings announcement drift in the stock price that occurs when a company delivers earnings that exceed/miss the consensus (the surprise) and continues to outperform/underperform (the drift) for subsequent periods as long as six months into the future.

Conservatism, whereby we overweight our prior views and underweight new evidence, and anchoring (on past prices) can help us explain why the release of new information is reflected only in prices over time instead of immediately, as the proponents of the efficient market hypothesis believe.

The main take-away from prospect theory is that investors tend to hold on to losing positions for too long (loss aversion, risk seeking and confirmation bias). Investors simply hate to be proven wrong and love to be proven right. This is exactly why investors also sell their winners too quickly; the certainty effect causes them to be risk-averse in the domain of gains. (Remember the old proverb: a bird in the hand is better than two in the bush.)

Integrating Behavioural Finance Into Your Investment Process
So how can investors avoid behavioural biases while attempting to exploit the biases of others? A quantitative process can help. First, by focusing on a few crucial elements for stock selection, you can avoid falling into the illusion that more information is better. What are the most important factors that drive a stock’s returns? Use this insight on as many stocks as you can, instead of trying to know everything about a few.

Second, by integrating an optimizer (an investment software) into your portfolio construction process, the memory-free machine will not have an attachment to existing views when new information comes along. It will therefore be less subject to a confirmation bias. Based on new evidence, the optimizer will objectively evaluate the new information’s value and arrive at a new price estimate without the past price being as relevant an anchor. The optimizer does not care about a firm’s past earnings pattern. If fundamentals and sentiment are good today and value is appealing, the optimizer will consider it for inclusion in the portfolio. And because most optimizers on the market have been developed using an expected utility framework, they will not fall prey to risk aversion or gambling in the domain of gains or losses. If a better opportunity comes along—meaning investors will have to sell a relatively worse-ranked stock at a loss—the machine will not have second thoughts about it.

The Five Financial Ws
The psychology of investing has been embraced by more and more portfolio managers over the past 15 years and has produced a different approach to stock selection and portfolio construction. It does not replace sound financial analysis but complements it by helping investors to remove emotional elements from the investment process. It brings investors’ attention to the five Ws they should always ask themselves before adopting a ranking process to select stocks:

1. What works?
2. Where does it work (in which part of the investable universe)?
3. When does it work (up/down markets)?

And by exploring the field of behavioural finance, investors can get answers to the last two Ws:

4. Why does it work?
5. Will it continue to work?

But is it really crucial to understand why these anomalies exist? Is it not enough to know that these biases exist and lead to mispriced assets that can be exploited? Behavioural finance is both fascinating and challenging, given the fact that investors are all subject to cognitive and emotional limitations. It also puts a human face to the black-box reputation of quantitative investing. But while the understanding of market anomalies can be insightful, investors should avoid the overconfidence and the illusion of control that comes with additional information. By combining the fundamentals derived from the analysis of financial statements and the study of investor psychology that creates mispriced assets, investors hope to generate returns in excess of the market over the long term. BC

Sources available on request.

Sophie Mayrand is vice-president and senior portfolio manager at State Street Global Advisors in Montreal.

Advertisements

September 28, 2010 - Posted by | behaviour, decision making, emotions, investing, psychology | , , , ,

No comments yet.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: