Behavioral Problems

The word ‘behavior' very often invokes the concept of mis-behavior.  Perhaps it is because we more often hear about a person or organization's poor conduct rather than good conduct?  While this observation may be enlightening, you may by now be snickering ‘what does this have to do with finance?'  Before I answer, let us segue with two questions:  What are the last two digits of your Social Security number?  Next, what is the maximum dollar amount you would pay for an average bottle of wine?

When Ariely, Loewenstein and Prelec posed these questions (on a number ordinary objects, not just ‘average' wine), those with low Social Security digits answered about $15, while those with high digits answered $30.  Let's think about that- a completely unrelated and arbitrary piece of information significantly altered a person's estimate of value.  This phenomenon is known as anchoring.  (In case you are wondering, it is a statistically significant deviation and the typical response was more dramatic than the one above.  Depending on the item (wine, Belgian chocolates,etc.), the ratio of answers from people with high SS numbers to those with low SS numbers ranged from 2 times to 3.5 times and averaged 2.7 times!)  The more relevant the anchor, the more heavily the bias influences the outcome.

Imagine you are a Wall Street analyst and your estimate for the fair value of a particular stock is $20, but other analysts are all publishing price targets of $55.  All day long you watch the ticker and see prices greater than $20 or commercials advertising health club memberships for $49.  CNBC broadcasts an interview with the CEO and shows the stock chart with the price trending near $40.  Does this influence your estimate?  You bet.

With the help of anchoring, our analyst thinks the stock must be worth more than $20 and so does not discourage his clients from owning it.  The clients, meanwhile, suffer from a bias known as the endowment effect.  Simply put, this effect says that once you own something, you start to place a higher value on it than others would.  Kahneman, Knetsch & Thaler performed a study with university students and school logo mugs.  Mugs, which sold for $6 in the university store, were randomly distributed to half the class.  Students were allowed to sell the mugs to those who did not have them.  After adjusting for control a group, it turned out that sellers were asking 50% more than buyers were willing to pay.  All this implies that owners of stocks may be unconsciously swayed into thinking their positions are worth more than they are.

Following our hypothetical clients, let us assume they held their position as it fell from $40 to $20.  Strangely, they seem to be unwilling to sell.  Known as loss aversion, it has been found that people dislike losses far more than they like gains.  James Montier posed this question: you are offered a bet on the toss of a fair coin.  If you lose, you must pay me $100.  What is the minimum amount that you need to win in order to make this bet attractive?  The answer is subjective and has no right or wrong response.  However, Montier's data shows the average answer is $200.  That means people, on average, dislike losses twice as much as they like gains.  Other studies estimate the ratio between 2 and 2.5 times.

One reason for loss aversion is known as self-attribution bias, which is a belief that good outcomes are the result of skill, while bad outcomes are the result of sheer bad luck (and hence the impetus to ‘hang on' since luck causes a temporary setback).  This is one of my favorite biases because we have all made this rationalization at one time or another.

Of course, our analyst now suffers from hindsight bias insofar as he believes that he "knew it all along".  Rather proudly, he recounts the story of how he correctly anticipated the stock to be worth only $20 despite being the only one with that forecast.  Did he really "know it all along"?  I suggest the answer is no.  Only actions confirm "what you knew" (telling clients to sell, which he didn't do because he lacked confidence).  One can talk up both sides of the argument without making a decision- a valid dissemination of data and analysis, but not a recommendation.  Another way to negate this bias is to record your predictions- in essence, a form of action.  This is one reason I write letters to my clients- so my predictions and observations are recorded for future assessment.

Recognition of these biases are the result of a relatively new and exciting area of finance termed behavioral finance.  There are many more biases and topics to cover, so for now we'll conclude by asking ‘how do we combat such biases'?  One anti-bias method is to admit when you are wrong and examine your mistakes post-mortem.  Having the capability to admit mistakes increases your confidence by learning when you are right and wrong.  It helps to tune your analytical ability.  This intellectual honesty also enables you to act when your confidence in a market anomaly is reasonably strong.  Another important task is to remember that these biases apply to me, you and everyone else.  To do so, I keep a list of 15 biases and rules in front of me, between my keyboard and screen.  Every time I examine portfolios, markets and strategy, I am reminded of potential analytical mistakes.  I will still make them, but hopefully far less often.