Image credit: “New York, New Haven and Hartford Locomotive No. 321 crash through roundhouse,” 1905, The DeGolyer Library, Southern Methodist University
By Jason Zweig | June 20, 2015 5:47 p.m. ET
Here’s another speech from my archives, in which I suggested that many people who try to apply psychological findings to the financial markets do it backwards: instead of studying their own biases and failings, they focus on those of other people.
The earlier conference I referred to was in the spring of 1996, when I first heard and met Daniel Kahneman. Behavioral finance in general, and Danny and his work in particular, have inspired many of my articles since. In 2007 and 2008, I had the honor of assisting Danny with his great book Thinking, Fast and Slow.
In the intervening years since this speech, I’ve misplaced the slides; I’ll try to recreate them one of these days. In the meantime, investors should always remember, as I said here, that “behavioral finance is not the study of how ‘other’ people behave. It is the study of how we all behave. It is not just a window onto the world; it is also a mirror onto ourselves.”
Behavioral Finance:
What Good Is It, Anyway?
Jason Zweig
The John F. Kennedy School of Government Program on Investment Decisions and Behavioral Finance
Harvard University
Cambridge, Mass.
October 25, 1999
I would be guilty of overconfidence if I thought even for a moment that I could fully convey what a great privilege it is to be speaking here this evening. It was at this very event, three and a half years ago, that behavioral finance first struck me like a baseball bat right on the bridge of my nose.
Once I recovered from the staggering force of that blow, I’ve never been the same since, either as a professional or as a person. For completely transforming how I do my job, and for making me drive my wife absolutely crazy with my newly optimal decision-making, I have this program to thank.
I hope that in this year’s sessions you’ll learn almost as much as I have in the past — but not quite as much, so you can do your day jobs better than before and yet somehow return each evening to a less stormy home life than I’ve created for myself.
But I’m not sure you can. You will do a great disservice to yourselves, to your clients, and to your businesses, if you view behavioral finance mainly as a window onto the world. In truth, it is also a mirror that you must hold up to yourselves. More worrisome, it is a mirror that magnifies and clarifies and highlights your own warts and imperfections.
After all, it takes no great bravery to look out a window onto the world below and watch the foolish masses aimlessly stumbling nowhere near where they really want to go–while you can see quite clearly, from your lofty vantage point, the simplest and safest path to follow.
But it takes a great deal of courage to stare into a mirror and to hold it steady for a long, long time while this image sinks in: Gazing right back at you is someone who relentlessly falls prey to the law of small numbers; to hindsight bias; to over-reaction; to narrow framing; to mental accounting; to status-quo bias; to the inability to evaluate your own future regret; and, most of all, to overconfidence.
This brings me to the core of the question I’ve been asked to try answering: What good is behavioral finance, anyway?
You have heard today, and will hear again tomorrow, a ton of evidence that the market can be beaten with strategies that are solidly grounded in behavioral finance: deeply depressed value stocks, momentum stocks, stocks with positive earnings surprises. Many of you in this room have come here to learn about behavioral finance in the hope that it will enable you to take the other side of the trade from the great majority of investors whose behavior is, shall we say, sub-optimal. Wouldn’t it be great to trade against these people all day long, knowing in advance what they’ll do? It would make clubbing baby seals seem like good sportsmanship.
But if the market can be beaten this way, why are so few managers doing it? If an understanding of behavioral finance can really enable you to take the other side of the trade from the millions of investors who don’t understand their own shortcomings, then why do so few professionals seem to use it? If behavioral finance is the ultimate weapon against market efficiency, the silver bullet to the heart of the index-fund vampire, then where is the real-world evidence of its success — not as theory, but in practice?
I don’t really need to remind this audience of the depressing fact that, year after year after year, most professionally managed money underperforms the S&P. But, precisely because it can make you uncomfortable, I am going to remind you of it.
Despite what you’ve heard and what many of you fervently believe, underperformance is not merely a temporary by-product of the narrow market of the past few years. Over the decade ended in mid-1974, 89% of all money managers lagged the S&P 500. Over the 20 years ended in 1964, the average fund underperformed by roughly 110 basis points. Even from 1929 through 1950, not a single major mutual fund beat the S&P. Take any period you like; the results are invariably discouraging.
What does this tell us, other than the obvious facts that the majority of a population can’t remain above the mean, and that money managers and their clients incur expenses while an index is a cost-free abstraction?
I think what it tells us is that it is remarkably difficult for money managers to profit from the errors of uninformed investors. This is the limited arbitrage efficiency that Richard Zeckhauser spoke about this morning. But even so, that’s a bit puzzling. You have MBAs and Ph.Ds and CFAs; my readers don’t. You go eyeball-to-eyeball with CEOs and CFOs; my readers don’t. You have Bloomberg machines; my readers don’t. You even get to trade stocks at wholesale instead of retail; my readers don’t (although they think they do).
So why can the most sheepish of my readers, who parks his money in the no-brainer choice of an S&P index fund, make so many of you look so clueless so much of the time?
I think the answers are quite interesting. First of all, there is a substantial body of evidence that overconfidence grows worse as people become more expert in a given field. This is called the “inverted expertise” effect. In short, the more you know, the more you think you more know than you really do.
The mere fact that I’m standing at this podium is proof that I suffer from the inverted expertise effect; but you’d be taking the cheap way out if you concluded that it therefore doesn’t apply to each and every one of you.
It does. I believe that’s mainly because the more intensively you research a stock, the more of yourself you’ve invested in it. Your own superior knowledge of the stock becomes a sunk cost. Once you invest in a stock, you’ve invested something of yourself, too. It has become “your” stock, and you want it to go up to validate the superiority of your own judgment. Remember Arnie Wood’s discussion this morning of how people think they “know” something about a playing card after they touch it?
What’s more, we know that people generally regret errors of commission more than errors of omission. Hanging on to a stock while it decays is less painful than selling it and buying another stock that decays. We’re haunted by the fear of doing something that will later make us say, “If only I hadn’t done that!” As the poet John Greenleaf Whittier wrote, “For of all sad words of tongue and pen, The saddest are these: It might have been.” That’s bad poetry, but it’s good psychology. Our tendency to engage in counterfactual thinking — to ask “what might have been” if we had taken a different course of action (or inaction) — can have a paralytic effect on decision making. Maya Bar-Hillel of the Hebrew University and Gretchen Chapman of Rutgers have recently found that even when you give people something for nothing, they are reluctant to trade it away for something else you give them for nothing.
Thus a big holder of a stock is more likely, not less, to ignore the first faint whiffs of trouble that will make other investors’ noses twitch — because he’s afraid that if he sells his favorite stock that he knows so well, and replaces it with a new idea, the new one might do even worse.
This is often called the “disposition effect” or “status-quo bias.” When it reverses, it reverses violently — because it’s so finely interwoven with the fabric of your own self-esteem. After all, to sell a stock is to concede that you were fundamentally wrong. You made a mistake, and you’re kicking yourself for being such an idiot, and you want out of the damn thing just as fast as possible so you can go back to believing that you really do know what you’re doing. The sooner you can hide any evidence to the contrary, the better. Believe me, whoever gets to take the other side of that trade will be delighted that he found you.
[As you can see in this slide,] Wayne Wagner, whose Plexus Group researches transaction costs, has found that it costs more to sell a stock than to buy one, and I am quite sure that this is why. Note, too, that the larger the trade (and the more of your self-esteem you’ve invested in a stock), the more it costs you to unload it. It’s worth pointing out that these data are taken from a calendar quarter in which the S&P rose 21.3%, so selling should have been cheaper than buying.
Given normal rates of portfolio turnover and market volatility, I’d estimate that this kind of sub-optimal selling behavior easily costs the average fund manager 100 basis points a year.
So how can you get this runaway cost under control? It’s vital to have what Danny Kahneman calls a well-calibrated sense of your future regret. The good news is that in your business, unlike in mine and most people’s, every mistake, every regret, has a precisely quantifiable cost. But the bad news is that if you don’t anticipate how keenly you will feel your regrets, you’ll never be able to get those costs under control.
I’m always amazed how inarticulate most fund managers become when I ask them about their sell discipline. I’d think you should be extremely worried if you can’t explain your sell discipline in, say, 50 words or less. If I were you, I would give this some serious, serious thought.
Also, you must continually ask yourselves the most basic possible questions before you buy a stock: How confident can I realistically be that my analysis will turn out to be right? How have similar analyses worked out for me in the past? What’s the probability that I could be wrong? And how prepared am I for the consequences of surprise if I turn out to be wrong? You’ll be hearing a lot about overconfidence at this conference; take it to heart. You just don’t know as much as you think you do; none of us do.
That brings us to my next point. I want to rub your noses in the notion that behavioral finance explodes the “myth” of market efficiency — not on the theoretical level, but on the practical level.
Let’s think about information for a minute. In the old days, the investor who got the earliest grasp on the best information earned the highest return. The classic example, of course, is Nathan Rothschild and his flock of carrier pigeons, which almost 200 years ago gave him the finest early warning system in Europe and enabled him to dominate the foreign currency and bond markets for decades. In this kind of environment, the commodity that could be arbitraged most profitably was time itself.
But today, virtually every bit and byte of market information is transmitted instantaneously to every investor everywhere on earth. A great deal of information, in fact, is old before it even exists. Once, buy-side analysts spent weeks painstakingly calculating their own earnings estimates; today, no one is interested in actual earnings at all. What counts is “whisper numbers” and even “pre-whispers.” In other words, what matters is not what a company earns, but what Wall Street thinks Wall Street thinks a company will earn. Weeks in advance of any actual earnings release, the future has already been decided—and any differences of opinion disappear meaninglessly into the consensus.
Meanwhile, fund performance — which used to be measured annually, then quarterly, then monthly — is now measured daily. And millions of investors, retail and professional alike, measure stock performance in real time, tick by tick — and soon they’ll be going 24 hours a day. What’s more, a recent article in the Journal of Financial Economics found that day-trading is most profitable for holding periods of 80 seconds or less.
Thus, the long term has shrunk down to one-and-a-half minutes. Time is no longer arbitrageable. The velocity of learning has hit warp speed. The informational efficiency of the stock market has never been higher. As professionals, you have lost the edge that time arbitrage once gave you; scarier still, you have become the victims of it.
This graph
is based on experiments conducted on many species ranging from rodents to birds to insects to humans. The bars plotted on the Y axis show the actual value of future rewards. The X axis shows the length of time remaining until the rewards can be obtained. The curves plot the perceived or subjective value of each reward as time passes. (Those of you who heard David Laibson speak earlier today should get the point here.)
This “preference reversal” shows that when the time to receive a reward (in the region of t2 on our graph) is in the distant future, the larger, more remote reward is more attractive. But when delays are shorter (in the region of t1), then the smaller, closer payoff becomes far more preferable.
If I’ve lost you, just think of this: When you’re hungry, would you rather eat a large meal hours from now, or a small meal right now? The answer is obvious: When time is compressed, short-term partial gratification becomes more satisfying than long-term fuller gratification. This makes the pursuit of any long-horizon investment strategy — like, say, a deep-value approach — psychologically painful, both for you and your clients. They want to eat now, not later — and so do you. A strategy that will not pay off for years has become an almost unbearably expensive luxury in this increasingly short-term world. And that makes practical implementation of the theoretical implications of behavioral finance extremely difficult.
I also think this long bull market has changed the way prospect theory affects investors. Prospect theory holds that most individuals perceive the pain of loss at least twice as keenly as they feel the pleasure of gains.
But, in this kind of raging bull market, what is a “loss”? It rarely means losing money. It might, however, be a diminished gain. And when risk is coded not as a true loss, but merely as a foregone gain, it becomes much easier for investors to dump an investment. Instead of kicking themselves, they just kick you instead. That makes bailing out of an investment — like, say, one of your funds — much easier. For your clients today, firing you is no longer a damaging admission of their own fallibility; after all, they made some money instead of losing it. In short, they’re playing with the house money, as Dick Thaler alluded to earlier. And that makes firing you easier than ever. That’s why avoiding tracking error has become the prime directive; that’s why relative performance has assumed absolute importance.
In this climate, the agency costs are huge. If you pick stocks according to the principles of behavioral finance, you’re going to end up with a portfolio that looks nothing like any index. It might be full of dead and dying value stocks, or it might be full of scorching momentum stocks, or it might be full of stocks that keep shocking the daylights out of the analysts. Whichever approach you choose, in the long run it should make your clients very wealthy. But in the short run, these behavioral strategies may well make you poor — by sending your tracking error shooting straight through the roof, and your clients shooting straight out the door.
Let’s say that in Keokuk, Iowa, lives a sanitation worker named Henry Smith. Your fund was selected for Henry’s 401(k) plan and is continually monitored for relative performance by a pension consultant, who is continually monitored by Henry’s plan sponsor. Meanwhile, Morningstar plops your fund into one of nine style boxes, and financial planners all around the country stand ready to pull the plug on you the instant your fund migrates into another box.
Henry himself goes online several times a month, perhaps even several times a day, logging on to money.com, morningstar.com, thestreet.com to compare your fund’s performance with the S&P. And back at the truck depot where Henry works, the TV hasn’t been tuned to ESPN since the Chicago Bears last played in the Super Bowl; instead, CNBC blares away 14 hours a day.
If your fund falls too far behind the S&P, Henry will dump it at the drop of a hat. Forget the old days, when mutual fund investors used to need a signature guarantee before they could even write a letter to the transfer agent requesting a redemption, which took seven business days to settle. These days, three mouse-clicks and Henry’s gone.
What does all this really mean? It means that it doesn’t matter a bit whether the market is efficient or not. It just doesn’t matter. For you, the real risk is not systematic market risk, it’s not unsystematic stock risk — it’s business risk. It’s the hazard you subject your business to whenever you do anything that raises your tracking error. If you are not prepared to take that business risk head-on, then the market could be as inefficient as a three-fingered violinist, and that still wouldn’t do you one bit of good. After all, if you want to produce superior returns, it doesn’t suffice for the market to be inefficient; you must be efficient enough to take advantage of the market’s inefficiency. That’s why the best use for behavioral finance is not the study of the markets outside you, but the study of yourself, of your own firm, and of how you are prepared to do business.
You need to realize that the twin heuristics of representativeness and availability are what drive cash flows in and out of investments. Representativeness (sometimes called “the law of small numbers”) is the human tendency to consider short series of data to be typical of long-term trends. Thus, a money manager who has beaten the market for all of, say, three years in a row is considered a genius—even though a primitive calculation of probability would show that 12% of all chimpanzees should be able to beat the market in a given three-year period. Availability, meanwhile, is the natural inclination to let recent and vivid events overwhelm our judgment about normal expected outcomes. Thus a manager who generates a 100% return in a 9% market is also regarded as a genius, instead of the more likely assumption that he is either incredibly lucky or dangerously insane. Representativeness and availability are the main reasons why the public buys high and sells low—and why pension consultants hire high and fire low.
This chart
shows the relationship between performance and cash flow at a leading small-cap mutual fund over the five years ending May 31, 1997. The bars plot monthly cash flows, measured on the left scale in millions of dollars. The line plots the fund’s cumulative return, with May 1992 indexed to 100. As you can see, from mid-1992 through the end of 1995, monthly cash inflows ranged from zero to not much more than zero. The fund more than doubled in value — while it had virtually no customers. But then this fund earned Lipper’s No. 1 ranking for capital appreciation over the trailing three, five and ten years–and the manager yodeled that No. 1 ranking at the top of its lungs in advertisements far and wide. Advertising its No. 1 ranking was like rubbing raw meat across a lion’s nose; the public didn’t just invest in this fund, they attacked it. At the end of 1992, the fund had had total net assets of just $3 million. In the first six months of 1996, it took in $2.5 billion. These towering black bars are the availability heuristic turned into pure dollars.
And then what happened? Small-caps corrected, the fund’s returns tumbled, the manager had to panic-sell into a dropping market, and the public yanked out its money. The end result? The white bar
shows the portfolio’s time-weighted compound return for the period: a stunning 28%. The black bar shows the dollar-weighted return, telling us what the average shareholder in the fund earned: less than 4%. Because the manager’s own behavior aided and abetted their own worst behavior, the public investors earned an average of less than half the return of a CD in a fund that nearly tripled in value. Today, the fund’s assets — which peaked at $6 billion in the height of the public feeding frenzy — languish below $3 billion. And its returns have gone from the top decile to the tenth.
I hope you’ll remember I began this talk by pointing out that you can view the world through a window, or you can view it through a mirror. By heavily promoting its performance when it was hottest — and precisely when regression to the mean had the highest potential to destroy investors’ wealth — this fund’s management looked at the world through a window, instead of through a mirror. These folks clearly understood the weaknesses of the investing public. It’s far from clear that they understood their own. In the fullness of time, they ended up not only devastating their customers but nearly destroying their own business.
The lesson here is inescapable. The cash flow from your clients now rivals your investment process itself as the main determinant of total return. Unlike the profitability of your stock picks, you can control the rate at which cash flows into your funds. How you market your funds has become an inextricable part of how you invest. Your own investment behavior — your very ability to execute a distinctive strategy — will be determined largely by how you encourage your clients to behave.
Whether you are able to turn the theory of behavioral finance into the practice of superior investment management does not depend on how diligently you master the research of the great minds you’ll hear at this conference. It depends rather on the strength of your own character, on whether your firm can show the resolve to stand out from the herd and to stick to your strategy when your customers’ dollars are bleeding away by the millions and the billions. Learning how to invest smarter will do you no good whatsoever unless you are willing to endure the short-term pain of being right in the long run.
That’s why it’s so crucial to realize that behavioral finance is not the study of how “other” people behave. It is the study of how we all behave. It is not just a window onto the world; it is also a mirror onto ourselves. Only if you are willing to use behavioral finance as a mirror — and only if you are willing to pay the necessary price, which is assessed in precious units of self-esteem and foregone management fees — can you really get anything good out of it. Ultimately, that’s the only way you can pass through the gate. Thank you.