This conventional wisdom makes sense, however, only if stock prices are not irrationally high. Those who believe in the efficient markets hypothesis claim that stock prices are always correct. The conventional wisdom is based on the assumption that markets are rational, thus stock prices cannot be too high. If markets are crazy, however, the best investments might be radically different from those suggested by the conventional wisdom.
Thus, in order to decide where to invest our money, we must first evaluate the idea that markets are rational. We do this in two parts. In Chapter 2, we ask whether people are rational, and in Chapter 3, we ask whether groups of individuals interacting in financial markets are rational. We will conclude that people are not rational and markets are often crazy.
In this section we meet the lizard brain—that part of our financial decision-making machinery that costs us money. We will find that we are built with a backward-looking, pattern-seeking brain that tends to make us want to buy when prices are irrationally high and sell when prices are irrationally low. We are built to be exactly out of sync with financial opportunity.
chapter two
CRAZY PEOPLE Lizard Brains and the New Science of Irrationality
Do Not Be Afraid to Meet the Lizard Brain
“‘Boy, the food at this place is really terrible’ . . . ‘Yeah, I know, and such small portions.’ Well, that’s essentially how I feel about life. Full of loneliness and misery and suffering and unhappiness, and it’s all over much too quickly.” So says Woody Allen in the role of Alvy in the opening scene of Annie Hall.
Similarly, our rational skills for finance are simply terrible, filled with systematic errors and biases. As with Woody Allen’s punch line about not getting enough bad food, we use our limited analytic skills far too rarely when we make financial decisions. As bad as we can be at making financial decisions with the more rational parts of our brains, we get in even more trouble when the lizard brain starts calling the shots.
In this book, I divide the human brain into two parts: the prefrontal cortex and the lizard brain. This is a dramatic simplification of an extremely complicated reality. Most, but not all, of what we think of as abstract cognition occurs in the human brain’s prefrontal cortex. The term “lizard brain” includes many important human brain regions that have nothing to do with reptiles.1
Thus, “lizard brain” is shorthand for an important idea. It is used in the spirit advocated by Sir Peter Medewar, a scientific expert in the study of aging, in his famous article, “An Unsolved Problem of Biology”:
Being in some degree crippled by the handicap of trying to be intelligible, I am bound to make statements which, if not baldly wrong, are true only with qualifications which I shall have not time to give them. This disability is not to be avoided; one gets nowhere if every sentence is to be qualified and refined.2
Similarly, the lizard brain is a term that I grew to use while conducting research with my Harvard Business School colleague Professor George Baker. I continue to find it productive even in discussions with experts in behavior and cognition. Because the reality is complicated, however, we must remember that “lizard brain” is verbal shorthand for the less cognitive, less abstract mental forces that influence human behavior, most of which have nothing to do with lizards.
The lizard brain is great for finding food and shelter, but terrible at navigating financial markets. Many financial problems occur when we use the lizard brain to make monetary decisions. Instead of using the analytical part of our brain, we often default to older parts of our brain that helped our human ancestors survive for tens of thousands of years before financial markets were created. The lizard brain is not stupid, but when confronted with problems never experienced by our ancestors it can make us look crazy and cost us money.
Before we investigate how the lizard brain leads us astray in financial matters, we must first deal with another human universal: Criticism is unpleasant. Being told that we are bad at something is, for most people, about as enjoyable as a mild electric shock. As a professor, I see this with my MBA students on a daily basis. At the Harvard Business School we follow the Socratic method, and an integral part of that technique is getting students to reveal their own logical errors. This approach is an effective way to teach, but one that can be painful for the student as they learn the limits of their knowledge.
As we embark on learning about the science of irrationality, the unpleasant message is that all humans are built to make certain sorts of mistakes. It’s all fun and games until the irrationality comes home to roost in our own brains. Then rather than learn, our instincts direct us to close the eyes, cover the ears, and deny the truth that we, too, are irrational. In all the oral stories of Homer, the only known reference to writing comes in the form of a secret message. It is in the Iliad, when Queen Antea falls in love with handsome Bellerophon who spurns her love. Enraged, Queen Antea convinces her husband, King Proteus, to kill Bellerophon (Antea does not reveal her secret and adulterous love).
Proteus wants to kill Bellerophon, but shies away from doing the dirty work himself. Instead he has Bellerophon travel to another kingdom, bearing a secret message for the ruler of the neighboring land. The content of the secret note is “kill the messenger.” So, one of the first mentions of writing reveals a human tendency to kill the messenger.
The reward, however, for not killing the messenger and critiquing one’s own behavior can be large. After the 1997 Masters golf championship, Tiger Woods reevaluated his game. In the Masters, he had dominated the field and won by a record 12 strokes. Furthermore, in less than one year on the professional tour, Tiger won four events, earned over a million dollars, and became a worldwide celebrity.
After this initial round of fame and success, what was Tiger’s view of his game? He decided that he needed to fundamentally change his swing. In an interview with Time magazine (August 14, 2000), looking back on the decision, he told writer Dan Goodgame:
I knew I wasn’t in the greatest positions in my swing at the [1997] Masters. But my timing was great, so I got away with it. And I made almost every putt. You can have a wonderful week like that even when your swing isn’t sound. But can you still contend in tournaments with that swing when your timing isn’t as good? Will it hold up over a long period of time? The answer to those questions, with the swing I had, was no. And I wanted to change that.
Tiger went back to the drawing board. He revamped his swing, suffered through some disappointments, but ultimately emerged as the dominant player in the game. At one point, Woods’s lead over the second-ranked player was larger than the gap between No. 2 and No. 100.3 He went from being a great player to perhaps the greatest player of all time. The lesson is clear: Winning requires critical self-examination. If Tiger’s game needed improvement and benefited from some objective review, the rest of us surely can profit from honing our investment skills.
The Science of Individual Irrationality
The debate about irrationality has two components. First, do individuals make good decisions? Second, are market prices correct? While there is still a debate about the efficiency of market prices (we’ll cover this topic in the next chapter), the first question has been answered. Over the last 30 years, a significant body of research has clearly illustrated our human shortcomings.
In the late 1970s, Professors Daniel Kahneman and Amos Tversky began the rigorous documentation of human decision-making problems. One of Kahneman and Tversky’s famous experiments concerns a hypothetical woman named Linda. Here’s what they asked in the experiment:4
Linda is 31 years old, single, outspoken, and very bright. She majored in philosophy. As a student she was deeply concerned with issues of discrimination and social justice, and also participated in antinuclear demonstrations.
Which of these two alternatives is more probable?1. Linda is a bank teller.
2. Linda is a bank teller and is active in the feminist movement.
Take a moment to answer the question (we’ll get to the correct answer shortly)
. First, know that most people provide the wrong answer, and there is an intellectual debate over how to interpret the errors. Old-school economists have said that the errors were caused by poor experimental design. Their first response was to deny the evidence that humans make the mistakes shown by behavioral economists.
Behavioral economists refined their techniques and provided proof that people make mistakes in many important areas, going far beyond the Linda problem. Mainstream economists no longer refute this evidence, but still insist that models of robotic, cool-headed decision making are appropriate. In contrast, behavioral economists believe that conventional theories about rational behavior need to be fundamentally revised.
Back to Linda. What was your answer? The correct answer is: Linda is a bank teller. Of all the bank tellers in the world, only some of them are active in the feminist movement. This is true for any two attributes. Consider 100 college athletes. How many of them are women? How many are women and over 6 feet tall? Without knowing anything about the group of 100, the number of tall women cannot exceed the number of women. Similarly, there have to be more bank tellers than there are bank tellers who are also feminists.
People who answer number two in the Linda-the-bank-teller problem suffer from what Kahneman and Tversky label the “conjunction fallacy”: The conjoined probability of two statements must be lower than for either of the individual statements. Of the people in Kahneman and Tversky’s experiments, 85% gave the wrong answer. Why do we do so poorly on such simple tests?
Rocket Scientists Who Can’t Figure
Part of the cause of our individual irrationality is that we aren’t very good at doing calculations.
In one of my Harvard Business School classes we investigate the causes behind corporate waste. We examine situations in which executives use corporate funds to pay for individual perks. One of the most famous and well documented of these is RJR Nabisco in the early 1980s. As chronicled in Barbarians at the Gate, the CEO of that time, Mr. Ross Johnson, used corporate money to host lavish parties, hang out with celebrities, and build an “air force” of expensive private jets.5
One cause of these excesses was the fact that neither Mr. Johnson nor his board of directors had much stake in the company stock. In fact, Mr. Johnson owned 0.05% of the company stock, and this represented a small part of his total wealth. At a key point in my class, I ask my students to calculate Mr. Johnson’s share of the $21 million purchase price of one jet in the RJR air force. So, what is 0.05% of $21 million?
In one such session I picked a student—who had not volunteered—to provide the figure. He reached over for his calculator. I said, “Excuse me, but don’t you have two degrees from MIT?” He said yes. “And aren’t those degrees in course six (electrical engineering and computer science), one of the toughest and most mathematical areas of MIT?” “Yes,” he answered. “And you still need a calculator for this simple calculation?” The student said that yes he did need a calculator.
Most people, even those with analytical abilities sufficient to excel at MIT, are not good at even basic calculations. The calculator can readily provide the figure for Ross Johnson’s $10,500 (0.05% of $21 million) contribution to RJR’s jet fleet. For other problems that our brains do not solve well, however, the solution is not so simple. Consider the following two problems taken from the book Mean Genes, which I coauthored with my friend, Professor Jay Phelan of UCLA.6
Puzzle 1. Chinese families place a high value on sons, yet the Chinese government exerts extreme pressure to limit family size. Let’s assume that the chance of having a girl is exactly 50%, but every couple stops having babies once they have a son. So some families have one son, some have an older daughter and a son, some two older daughters and a son, and so on. In this scenario, what percentage of Chinese babies will be female?
Puzzle 2. Imagine that you are a doctor and one of your patients asks to take an HIV test. You assure her that the test is unnecessary as only one woman out of a thousand with her age and sexual history is infected. She insists, and sadly the test result indicates viral infection. If the HIV test is 95% accurate, what is the chance that your patient is actually sick?
As with Linda the bank teller, almost everyone gets these two problems wrong, and I could pose many other brainteasers that would also trip up most people.
In fact, when doctors and staff at the Harvard Medical School were asked the question about the HIV test, the most common answer they gave was a 95% chance that the patient was sick.7 The correct answer is under 2%. Similarly, as long as the chance of having a baby girl in each pregnancy is exactly 50%, the population will also have 50% girls. This is true regardless of any rule on when to stop having babies. If you are interested in detailed analysis of these sorts of problems, I suggest that you read the risk chapter of Mean Genes. The key message for this book is that most people have trouble doing mathematical calculations.
Sound investing is based on mathematical analysis that is far more complicated than the problems we just discussed. At the core of every investment is a set of costs and benefits that need to be predicted over many years and in many scenarios. Coming up with the correct price for IBM stock or for our own house involves some serious math!
All of us who get even simple problems wrong are in good company. Not only do Harvard doctors make huge mistakes on these problems, so do the most sophisticated people in the world. One of my buddies, Chris, has both undergraduate and doctoral degrees from MIT in physics. His research on lasers is so secretive that he cannot reveal the sponsor of his work. In other words, he is a twenty-first century rocket scientist (for Val Kilmer fans, watch Real Genius to understand this brainy culture). In spite of all his ability and training, Chris admitted that he got the HIV problem wrong.
So we aren’t built to do mathematical calculations, and relatively simple problems trip up MIT rocket scientists. The news gets even worse. The second big problem we face in investing is that we are systematically overconfident. We are bad at doing the calculations required to analyze investments, and simultaneously we are unaware of our shortcomings.
Our overconfidence comes in many flavors. When people are asked to rank themselves compared to others, the average rating is always above average. For example, far more than 50% of people rank themselves in the top half of driving ability, although that is a statistical impossibility.8 When couples were asked to estimate their contribution to household work, the combined total routinely exceeded 100%.9
Myriad studies have documented this bias in our self-analysis, but my favorite remains an old study that asked men to rank themselves according to athletic ability. How many men do you think put themselves in the bottom half of male athletic ability? I suspect that you know the answer—not a single man who was surveyed reported that he had below-average athletic ability.10
Our overconfidence extends beyond self-analysis to our views of the world. Let’s take a simple test: How many people were employed by Wal-Mart in January 2004, around the world? Without looking up any information, write down a specific estimate. That may not seem fair, as different people know more or less about Wal-Mart.
To make the question fair, in addition to your guess, write down an upper-bound and a lower-bound number. Pick these bounds so that you are 90% sure that the actual number of employees is between your extreme high and your extreme low guesses.
If you answer 10 questions of this sort, nine of the answers should fall between your upper and lower bound. Do you have your three numbers for Wal-Mart? Your best estimate of the correct number, and lower- and upper-bound numbers?
We’ll get to the correct answer in a moment. Under exactly these sorts of conditions, when people are asked 10 such questions, they usually get between two and four questions wrong.11 This poor performance comes even after they have been told to give estimates wide enough to get only one of the 10 questions wrong.
People fail in this guessing game because they place too much confidence in their own estimates. Actually, I ought
to say that “we” fail, as I have been tested in this manner and also came up overconfident. Before I viewed my 10 questions, I resolved to make my lower and upper guesses extremely wide. Even with that preparation, only 8 of my 10 upper and lower bounds contained the correct answer. Back to Wal-Mart: In January 2004, the firm had 1.5 million employees.
The summary is that we come to the investing game with an analytic tool kit that lacks some of the key tools required for investment analysis. To add further insult to injury, our overconfidence makes us believe we have the required skills for investing.
Split-Brain Investing
Even though our analytic investing tool kit is not complete, it is our best hope to make good choices. An amazing fact is how rarely we use analysis to make our decisions.
During the early 1990s my biggest investment was in Microsoft. One evening I was standing outside my Harvard graduate dormitory chatting with my buddy Matt. I said, “Matt, I have a puzzle that I want to discuss with you. The puzzle is that Microsoft’s business is doing great, yet the stock has not gone up in months.” Matt allowed me to blather on about the fantastic business of selling software to the world, and then he asked, “What is the price-to-earnings ratio of Microsoft?”
Mean Markets and Lizard Brains Page 3