Anticipating Others' Anticipations
/ t was early 2000, the market was booming, and my investments in various index funds were doing well but not generating much excitement. Why investments should generate excitement is another issue, but it seemed that many people were genuinely enjoying the active management of their portfolios. So when I received a small and totally unexpected chunk of money, I placed it into what Richard Thaler, a behavioral economist I’ll return to later, calls a separate mental account. I considered it, in effect, “mad money.”
Nothing distinguished the money from other assets of mine except this private designation, but being so classified made my modest windfall more vulnerable to whim. In this case it entrained a series of ill-fated investment decisions that, even now, are excruciating to recall. The psychological ease with which such funds tend to be spent was no doubt a factor in my using the unexpected money to buy some shares of WorldCom (abbreviated WCOM), “the pre-eminent global communications company for the digital generation,” as its ads boasted, at $47 per share. (Hereafter I’ll generally use WCOM to refer to the stock and WorldCom to refer to the company.)
Today, of course, WorldCom is synonymous with business fraud, but in the halcyon late 1990s it seemed an irrepressibly
l
2 J o h n Allen Paulos
successful devourer of high-tech telecommunications companies. Bernie Ebbers, the founder and former CEO, is now viewed by many as a pirate, but then he was seen as a swashbuckler. I had read about the company, knew that high-tech guru George Gilder had been long and fervently singing its praises, and was aware that among its holdings were MCI, the huge long-distance telephone company, and UUNet, the “backbone” of the Internet. I spend a lot of time on the net (home is where you hang your @) so I found Gilder’s lyrical writings on the “telecosm” and the glories of unlimited bandwidth particularly seductive.
I also knew that, unlike most dot-com companies with no money coming in and few customers, WorldCom had more than $25 billion in revenues and almost 25 million customers, and so when several people I knew told me that WorldCom was a “strong buy,” I was receptive to their suggestion. Although the stock had recently fallen a little in price, it was, I was assured, likely to soon surpass its previous high of $64.
If this was all there was to it, there would have been no important financial consequences for me, and I wouldn’t be writing about the investment now. Alas, there was something else, or rather a whole series of “something elses.” After buying the shares, I found myself idly wondering, why not buy more? I don’t think of myself as a gambler, but I willed myself not to think, willed myself simply to act, willed myself to buy more shares of WCOM, shares that cost considerably more than the few I’d already bought. Nor were these the last shares I would buy. Usually a hardheaded fellow, I was nevertheless falling disastrously in love.
Although my particular heartthrob was WCOM, almost all of what I will say about my experience is unfortunately applicable to many other stocks and many other investors. Wherever WCOM appears, you may wish to substitute the symbols
A Mathematician Plays the Stock Market 3
for Lucent, Tyco, Intel, Yahoo, AOL-Time Warner, Global Crossing, Enron, Adelphia, or, perhaps, the generic symbols WOE or BANE. The time frame of the book—in the midst of a market collapse after a heady, nearly decade-long surge— may also appear rather more specific and constraining than it is. Almost all the points made herein are rather general or can be generalized with a little common sense.
Falling in Love with WorldCom
John Maynard Keynes, arguably the greatest economist of the twentieth century, likened the position of short-term investors in a stock market to that of readers in a newspaper beauty contest (popular in his day). The ostensible task of the readers is to pick the five prettiest out of, say, one hundred contestants, but their real job is more complicated. The reason is that the newspaper rewards them with small prizes only if they pick the five contestants who receive the most votes from readers. That is, they must pick the contestants that they think are most likely to be picked by the other readers, and the other readers must try to do the same. They’re not to become enamored of any of the contestants or otherwise give undue weight to their own taste. Rather they must, in Keynes’ words, anticipate “what average opinion expects the average opinion to be” (or, worse, anticipate what the average opinion expects the average opinion expects the average opinion to be).
Thus it may be that, as in politics, the golden touch derives oddly from being in tune with the brass masses. People might dismiss rumors, for example, about “Enronitis” or “World- Comism” affecting the companies in which they’ve invested, but if they believe others will believe the rumors, they can’t afford to ignore them.
BWC (before WorldCom) such social calculations never interested me much. I didn’t find the market particularly inspiring or exalted and viewed it simply as a way to trade shares in businesses. Studying the market wasn’t nearly as engaging as doing mathematics or philosophy or watching the Comedy Network. Thus, taking Keynes literally and not having much confidence in my judgment of popular taste, I refrained from investing in individual stocks. In addition, I believed that stock movements were entirely random and that trying to outsmart dice was a fool’s errand. The bulk of my money therefore went into broad-gauge stock index funds.
AWC, however, I deviated from this generally wise course. Fathoming the market, to the extent possible, and predicting it, if at all possible, suddenly became live issues. Instead of snidely dismissing the business talk shows’ vapid talk, sports- caster-ish attitudes, and empty prognostication, I began to search for what of substance might underlie all the commentary about the market and slowly changed my mind about some matters. I also sought to account for my own sometimes foolish behavior, instances of which will appear throughout the book, and tried to reconcile it with my understanding of the mathematics underlying the market.
Lest you dread a cloyingly personal account of how I lost my shirt (or at least had my sleeves shortened), I should stress that my primary purpose here is to lay out, elucidate, and explore the basic conceptual mathematics of the market. I’ll examine—largely via vignettes and stories rather than formulas and equations—various approaches to investing as well as a number of problems, paradoxes, and puzzles, some old, some new, that encapsulate issues associated with the market. Is it efficient? Random? Is there anything to technical analysis, fundamental analysis? How can one quantify risk? What is the role of cognitive illusion? Of common knowledge? What are the most common scams? What are
4 J o h n Allen Paulos
A Mathematician Plays th e Stock Market 5
options, portfolio theory, short-selling, the efficient market hypothesis? Does the normal bell-shaped curve explain the market’s occasional extreme volatility? What about fractals, chaos, and other non-standard tools? There will be no explicit investment advice and certainly no segments devoted to the ten best stocks for the new millennium, the five smartest ways to jump-start your 401 (k), or the three savviest steps you can take right now. In short, there’ll be no financial pornography.
Often inseparable from these mathematical issues, however, is psychology, and so I’ll begin with a discussion of the no-man’s land between this discipline and mathematics.
Being Right Versus Being Right About the Market
There’s something very reductive about the stock market. You can be right for the wrong reasons or wrong for the right reasons, but to the market you’re just plain right or wrong. Compare this to the story of the teacher who asks if anyone in the class can name two pronouns. When no one volunteers, the teacher calls on Tommy who responds, “Who, me?” To the market, Tommy is right and therefore, despite being unlikely to get an A in English, he’s rich.
Guessing right about the market usually leads to chortling. While waiting to give a radio interview at a studio in Philadelphia in June 2002, I mentioned to the security guard that I was writing this book. This set him off on a long disquisition on the market and how a couple of years before he had received two consecutive statements from his 401(k) administrator indicating that his retirement funds had declined. (He took this to be what in chapter 3 is called a technical sell signal.) “The first one I might think was an accident, but two in
6 J o h n Allen Paulos
a row, no. Do you know I had to argue with that pension person there about getting out of stocks and into those treasury bills? She told me not to worry because I wasn’t going to retire for years, but I insisted ‘No, I want out now.’ And I’m sure glad I did get out.” He went on to tell me about “all the big shots at the station who cry like babies every day about how much money they lost. I warned them that two down statements and you get out, but they didn’t listen to me.”
I didn’t tell the guard about my ill-starred WorldCom experience, but later I did say to the producer and sound man that the guard had told me about his financial foresight in response to my mentioning my book on the stock market. They both assured me that he would have told me no matter what. “He tells everyone,” they said, with the glum humor of big shots who didn’t take his advice and now cry like babies.
Such anecdotes bring up the question: “If you’re so smart, why ain’t you rich?” Anyone with a modicum of intelligence and an unpaid bill or two is asked this question repeatedly. But just as there is a distinction between being smart and being rich, there is a parallel distinction between being right and being right about the market.
Consider a situation in which the individuals in a group must simultaneously choose a number between 0 and 100. They are further directed to pick the number that they think will be closest to 80 percent of the average number chosen by the group. The one who comes closest will receive $100 for his efforts. Stop for a bit and think what number you would pick.
Some in the group might reason that the average number chosen is likely to be 50 and so these people would guess 40, which is 80 percent of this. Others might anticipate that people will guess 40 for this reason and so they would guess 32, which is 80 percent of 40. Still others might anticipate that people will guess 32 for this reason and so they would guess 25.6, which is 80 percent of 32.
A M athematician Plays th e Stock Market 7
If the group continues to play this game, they will gradually learn to engage in ever more iterations of this metareasoning about others’ reasoning until they all reach the optimal response, which is 0. Since they all want to choose a number equal to 80 percent of the average, the only way they can all do this is by choosing 0, the only number equal to 80 percent of itself. (Choosing 0 leads to what is called the Nash equilibrium of this game. It results when individuals modify their actions until they can no longer benefit from changing them given what the others’ actions are.)
The problem of guessing 80 percent of the average guess is a bit like Keynes’s description of the investors’ task. What makes it tricky is that anyone bright enough to cut to the heart of the problem and guess 0 right away is almost certain to be wrong, since different individuals will engage in different degrees of meta-reasoning about others’ reasoning. Some, to increase their chances, will choose numbers a little above or a little below the natural guesses of 40 or 32 or 25.6 or 20.48. There will be some random guesses as well and some guesses of 50 or more. Unless the group is very unusual, few will guess 0 initially.
If a group plays this game only once or twice, guessing the average of all the guesses is as much a matter of reading the others’ intelligence and psychology as it is of following an idea to its logical conclusion. By the same token, gauging investors is often as important as gauging investments. And it’s likely to be more difficult.
My Pedagogical Cruelty
Other situations, as well, require anticipating others’ actions and adapting yours to theirs. Recall, for example, the television show on which contestants had to guess how their spouses would guess they would answer a particular question.
8 J o h n Allen Paulos
There was also a show on which opposing teams had to guess the most common associations the studio audience had made with a collection of words. Or consider the game in which you have to pick the location in New York City (or simply the local shopping mall) that others would most likely look for you first. You win if the location you pick is chosen by most of the others. Instances of Keynes’s beauty contest metaphor are widespread.
As I’ve related elsewhere, a number of years ago I taught a summer probability course at Temple University. It met every day and the pace was rapid, so to induce my students to keep up with the material I gave a short quiz every day. Applying a perverse idea I’d experimented with in other classes, I placed a little box at the bottom of each exam sheet and a notation next to it stating that students who crossed the box (placed an X in it) would have ten extra points added to their exam scores. A further notation stated that the points would be added only if less than half the class crossed the box. If more than half crossed the box, those crossing it would lose ten points on their exam scores. This practice, I admit, bordered on pedagogical cruelty.
A few brave souls crossed the box on the first quiz and received ten extra points. As the summer wore on, more and more students did so. One day I announced that more than half the students had crossed the box and that those who did had therefore been penalized ten points. Very few students crossed the box on the next exam. Gradually, however, the number crossing it edged up to around 40 percent of the class and stayed there. But it was always a different 40 percent, and it struck me that the calculation a student had to perform to decide whether to cross the box was quite difficult. It was especially so since the class was composed largely of foreign students who, despite my best efforts (which included this little game), seemed to have developed little camaraderie. Without
any collusion that I could discern, the students had to anticipate other students’ anticipations of their anticipations in a convoluted and very skittish self-referential tangle. Dizzying.
I’ve since learned that W. Brian Arthur, an economist at the Santa Fe Institute and Stanford University, has long used an essentially identical scenario to describe the predicament of bar patrons deciding whether or not to go to a popular bar, the experience being pleasant only if the bar is not thronged. An equilibrium naturally develops whereby the bar rarely becomes too full. (This almost seems like a belated scientific justification for Yogi Berra’s quip about Toots Shor’s restaurant in New York: “Nobody goes there any more. It’s too crowded.”) Arthur proposed the model to clarify the behavior of market investors who, like my students and the bar patrons, must anticipate others’ anticipations of them (and so on). Whether one buys or sells, crosses the box or doesn’t cross, goes to the bar or doesn’t go, depends upon one’s beliefs about others’ possible actions and beliefs.
The Consumer Confidence Index, which measures consumers’ propensity to consume and their confidence in their own economic future, is likewise subject to a flighty, reflexive sort of consensus. Since people’s evaluation of their own economic prospects is so dependent on what they perceive others’ prospects to be, the CCI indirectly surveys people’s beliefs about other people’s beliefs. (“Consume” and “consumer” are, in this context, common but unfortunate terms. “Buy,” “purchase,” “citizen,” and “household” are, I think, preferable.)
A M athematician Plays the Stock Market 9
Common Knowledge, Jealousy, and Market Sell-Offs
Sizing up other investors is more than a matter of psychology. New logical notions are needed as well. One of them,
10 J o h n Allen Paulos
“common knowledge,” due originally to the economist Robert Aumann, is crucial to understanding the complexity of the stock market and the importance of transparency. A bit of information is common knowledge among a group of people if all parties know it, know that the others know it, know that the others know they know it, and so on. It is much more than “mutual knowledge,” which requires only that the parties know the particular bit of information, not that they be aware of the others’ knowledge.
As I’ll discuss later, this notion of common knowledge is essential to seeing how “subterranean information processing” often underlies sudden bubbles or crashes in the markets, changes that seem to be precipitated by nothing at all and therefore are almost impossible to foresee. It is also relevant to the recent market sell-offs and accounting scandals, but before we get to more realistic accounts of the market, consider the following parable from my book Once Upon a Number, which illustrates the power of common knowledge. The story takes place in a benightedly sexist village of uncertain location. In this village there are many married couples and each woman immediately knows when another woman’s husband has been unfaithful but not when her own has. The very strict feminist statutes of the village require that if a woman can prove her husband has been unfaithful, she must kill him that very day. Assume that the women are statute-abiding, intelligent, aware of the intelligence of the other women, and, mercifully, that they never inform other women of their philandering husbands. As it happens, twenty of the men have been unfaithful, but since no woman can prove her husband has been so, village life proceeds merrily and warily along. Then one morning the tribal matriarch comes to visit from the far side of the forest. Her honesty is acknowledged by all and her word is taken as truth. She warns the assembled villagers that there is at least one philandering husband
A M athematician Plays th e Stock Market 11
among them. Once this fact, already known to everyone, becomes common knowledge, what happens?
The answer is that the matriarch’s warning will be followed by nineteen peaceful days and then, on the twentieth day, by a massive slaughter in which twenty women kill their husbands. To see this, assume there is only one unfaithful husband, Mr.
A. Everyone except Mrs. A already knows about him, so when the matriarch makes her announcement, only she learns something new from it. Being intelligent, she realizes that she would know if any other husband were unfaithful. She thus infers that Mr. A is the philanderer and kills him that very day.
Now assume there are two unfaithful men, Mr. A and Mr.
B. Every woman except Mrs. A and Mrs. B knows about both these cases of infidelity. Mrs. A knows only of Mr. B’s, and Mrs. B knows only of Mr. A’s. Mrs. A thus learns nothing from the matriarch’s announcement, but when Mrs. B fails to kill Mr. B the first day, she infers that there must be a second philandering husband, who can only be Mr. A. The same holds for Mrs. B who infers from the fact that Mrs. A has not killed her husband on the first day that Mr. B is also guilty. The next day Mrs. A and Mrs. B both kill their husbands.
If there are exactly three guilty husbands, Mr. A, Mr. B, and Mr. C, then the matriarch’s announcement would have no visible effect the first day or the second, but by a reasoning process similar to the one above, Mrs. A, Mrs. B, and Mrs. C would each infer from the inaction of the other two of them on the first two days that their husbands were also guilty and kill them on the third day. By a process of mathematical induction we can conclude that if twenty husbands are unfaithful, their intelligent wives would finally be able to prove it on the twentieth day, the day of the righteous bloodbath.
Now if you replace the warning of the matriarch with that provided by, say, an announcement by the Securities and Exchange Commission, the nervousness of the wives with the
12 J o h n Allen Paulos
nervousness of investors, the wives’ contentment as long as their own husbands weren’t straying with the investors’ contentment as long their own companies weren’t cooking the books, killing husbands with selling stocks, and the gap between the warning and the killings with the delay between announcement of an investigation and big sell-offs, you can understand how this parable of common knowledge applies to the market.
Note that in order to change the logical status of a bit of information from mutually known to commonly known, there must be an independent arbiter. In the parable it was the matriarch; in the market analogue it was the SEC. If there is no one who is universally respected and believed, the motivating and cleansing effect of warnings is lost.
Happily, unlike the poor husbands, the market is capable of rebirth.
Nothing distinguished the money from other assets of mine except this private designation, but being so classified made my modest windfall more vulnerable to whim. In this case it entrained a series of ill-fated investment decisions that, even now, are excruciating to recall. The psychological ease with which such funds tend to be spent was no doubt a factor in my using the unexpected money to buy some shares of WorldCom (abbreviated WCOM), “the pre-eminent global communications company for the digital generation,” as its ads boasted, at $47 per share. (Hereafter I’ll generally use WCOM to refer to the stock and WorldCom to refer to the company.)
Today, of course, WorldCom is synonymous with business fraud, but in the halcyon late 1990s it seemed an irrepressibly
l
2 J o h n Allen Paulos
successful devourer of high-tech telecommunications companies. Bernie Ebbers, the founder and former CEO, is now viewed by many as a pirate, but then he was seen as a swashbuckler. I had read about the company, knew that high-tech guru George Gilder had been long and fervently singing its praises, and was aware that among its holdings were MCI, the huge long-distance telephone company, and UUNet, the “backbone” of the Internet. I spend a lot of time on the net (home is where you hang your @) so I found Gilder’s lyrical writings on the “telecosm” and the glories of unlimited bandwidth particularly seductive.
I also knew that, unlike most dot-com companies with no money coming in and few customers, WorldCom had more than $25 billion in revenues and almost 25 million customers, and so when several people I knew told me that WorldCom was a “strong buy,” I was receptive to their suggestion. Although the stock had recently fallen a little in price, it was, I was assured, likely to soon surpass its previous high of $64.
If this was all there was to it, there would have been no important financial consequences for me, and I wouldn’t be writing about the investment now. Alas, there was something else, or rather a whole series of “something elses.” After buying the shares, I found myself idly wondering, why not buy more? I don’t think of myself as a gambler, but I willed myself not to think, willed myself simply to act, willed myself to buy more shares of WCOM, shares that cost considerably more than the few I’d already bought. Nor were these the last shares I would buy. Usually a hardheaded fellow, I was nevertheless falling disastrously in love.
Although my particular heartthrob was WCOM, almost all of what I will say about my experience is unfortunately applicable to many other stocks and many other investors. Wherever WCOM appears, you may wish to substitute the symbols
A Mathematician Plays the Stock Market 3
for Lucent, Tyco, Intel, Yahoo, AOL-Time Warner, Global Crossing, Enron, Adelphia, or, perhaps, the generic symbols WOE or BANE. The time frame of the book—in the midst of a market collapse after a heady, nearly decade-long surge— may also appear rather more specific and constraining than it is. Almost all the points made herein are rather general or can be generalized with a little common sense.
Falling in Love with WorldCom
John Maynard Keynes, arguably the greatest economist of the twentieth century, likened the position of short-term investors in a stock market to that of readers in a newspaper beauty contest (popular in his day). The ostensible task of the readers is to pick the five prettiest out of, say, one hundred contestants, but their real job is more complicated. The reason is that the newspaper rewards them with small prizes only if they pick the five contestants who receive the most votes from readers. That is, they must pick the contestants that they think are most likely to be picked by the other readers, and the other readers must try to do the same. They’re not to become enamored of any of the contestants or otherwise give undue weight to their own taste. Rather they must, in Keynes’ words, anticipate “what average opinion expects the average opinion to be” (or, worse, anticipate what the average opinion expects the average opinion expects the average opinion to be).
Thus it may be that, as in politics, the golden touch derives oddly from being in tune with the brass masses. People might dismiss rumors, for example, about “Enronitis” or “World- Comism” affecting the companies in which they’ve invested, but if they believe others will believe the rumors, they can’t afford to ignore them.
BWC (before WorldCom) such social calculations never interested me much. I didn’t find the market particularly inspiring or exalted and viewed it simply as a way to trade shares in businesses. Studying the market wasn’t nearly as engaging as doing mathematics or philosophy or watching the Comedy Network. Thus, taking Keynes literally and not having much confidence in my judgment of popular taste, I refrained from investing in individual stocks. In addition, I believed that stock movements were entirely random and that trying to outsmart dice was a fool’s errand. The bulk of my money therefore went into broad-gauge stock index funds.
AWC, however, I deviated from this generally wise course. Fathoming the market, to the extent possible, and predicting it, if at all possible, suddenly became live issues. Instead of snidely dismissing the business talk shows’ vapid talk, sports- caster-ish attitudes, and empty prognostication, I began to search for what of substance might underlie all the commentary about the market and slowly changed my mind about some matters. I also sought to account for my own sometimes foolish behavior, instances of which will appear throughout the book, and tried to reconcile it with my understanding of the mathematics underlying the market.
Lest you dread a cloyingly personal account of how I lost my shirt (or at least had my sleeves shortened), I should stress that my primary purpose here is to lay out, elucidate, and explore the basic conceptual mathematics of the market. I’ll examine—largely via vignettes and stories rather than formulas and equations—various approaches to investing as well as a number of problems, paradoxes, and puzzles, some old, some new, that encapsulate issues associated with the market. Is it efficient? Random? Is there anything to technical analysis, fundamental analysis? How can one quantify risk? What is the role of cognitive illusion? Of common knowledge? What are the most common scams? What are
4 J o h n Allen Paulos
A Mathematician Plays th e Stock Market 5
options, portfolio theory, short-selling, the efficient market hypothesis? Does the normal bell-shaped curve explain the market’s occasional extreme volatility? What about fractals, chaos, and other non-standard tools? There will be no explicit investment advice and certainly no segments devoted to the ten best stocks for the new millennium, the five smartest ways to jump-start your 401 (k), or the three savviest steps you can take right now. In short, there’ll be no financial pornography.
Often inseparable from these mathematical issues, however, is psychology, and so I’ll begin with a discussion of the no-man’s land between this discipline and mathematics.
Being Right Versus Being Right About the Market
There’s something very reductive about the stock market. You can be right for the wrong reasons or wrong for the right reasons, but to the market you’re just plain right or wrong. Compare this to the story of the teacher who asks if anyone in the class can name two pronouns. When no one volunteers, the teacher calls on Tommy who responds, “Who, me?” To the market, Tommy is right and therefore, despite being unlikely to get an A in English, he’s rich.
Guessing right about the market usually leads to chortling. While waiting to give a radio interview at a studio in Philadelphia in June 2002, I mentioned to the security guard that I was writing this book. This set him off on a long disquisition on the market and how a couple of years before he had received two consecutive statements from his 401(k) administrator indicating that his retirement funds had declined. (He took this to be what in chapter 3 is called a technical sell signal.) “The first one I might think was an accident, but two in
6 J o h n Allen Paulos
a row, no. Do you know I had to argue with that pension person there about getting out of stocks and into those treasury bills? She told me not to worry because I wasn’t going to retire for years, but I insisted ‘No, I want out now.’ And I’m sure glad I did get out.” He went on to tell me about “all the big shots at the station who cry like babies every day about how much money they lost. I warned them that two down statements and you get out, but they didn’t listen to me.”
I didn’t tell the guard about my ill-starred WorldCom experience, but later I did say to the producer and sound man that the guard had told me about his financial foresight in response to my mentioning my book on the stock market. They both assured me that he would have told me no matter what. “He tells everyone,” they said, with the glum humor of big shots who didn’t take his advice and now cry like babies.
Such anecdotes bring up the question: “If you’re so smart, why ain’t you rich?” Anyone with a modicum of intelligence and an unpaid bill or two is asked this question repeatedly. But just as there is a distinction between being smart and being rich, there is a parallel distinction between being right and being right about the market.
Consider a situation in which the individuals in a group must simultaneously choose a number between 0 and 100. They are further directed to pick the number that they think will be closest to 80 percent of the average number chosen by the group. The one who comes closest will receive $100 for his efforts. Stop for a bit and think what number you would pick.
Some in the group might reason that the average number chosen is likely to be 50 and so these people would guess 40, which is 80 percent of this. Others might anticipate that people will guess 40 for this reason and so they would guess 32, which is 80 percent of 40. Still others might anticipate that people will guess 32 for this reason and so they would guess 25.6, which is 80 percent of 32.
A M athematician Plays th e Stock Market 7
If the group continues to play this game, they will gradually learn to engage in ever more iterations of this metareasoning about others’ reasoning until they all reach the optimal response, which is 0. Since they all want to choose a number equal to 80 percent of the average, the only way they can all do this is by choosing 0, the only number equal to 80 percent of itself. (Choosing 0 leads to what is called the Nash equilibrium of this game. It results when individuals modify their actions until they can no longer benefit from changing them given what the others’ actions are.)
The problem of guessing 80 percent of the average guess is a bit like Keynes’s description of the investors’ task. What makes it tricky is that anyone bright enough to cut to the heart of the problem and guess 0 right away is almost certain to be wrong, since different individuals will engage in different degrees of meta-reasoning about others’ reasoning. Some, to increase their chances, will choose numbers a little above or a little below the natural guesses of 40 or 32 or 25.6 or 20.48. There will be some random guesses as well and some guesses of 50 or more. Unless the group is very unusual, few will guess 0 initially.
If a group plays this game only once or twice, guessing the average of all the guesses is as much a matter of reading the others’ intelligence and psychology as it is of following an idea to its logical conclusion. By the same token, gauging investors is often as important as gauging investments. And it’s likely to be more difficult.
My Pedagogical Cruelty
Other situations, as well, require anticipating others’ actions and adapting yours to theirs. Recall, for example, the television show on which contestants had to guess how their spouses would guess they would answer a particular question.
8 J o h n Allen Paulos
There was also a show on which opposing teams had to guess the most common associations the studio audience had made with a collection of words. Or consider the game in which you have to pick the location in New York City (or simply the local shopping mall) that others would most likely look for you first. You win if the location you pick is chosen by most of the others. Instances of Keynes’s beauty contest metaphor are widespread.
As I’ve related elsewhere, a number of years ago I taught a summer probability course at Temple University. It met every day and the pace was rapid, so to induce my students to keep up with the material I gave a short quiz every day. Applying a perverse idea I’d experimented with in other classes, I placed a little box at the bottom of each exam sheet and a notation next to it stating that students who crossed the box (placed an X in it) would have ten extra points added to their exam scores. A further notation stated that the points would be added only if less than half the class crossed the box. If more than half crossed the box, those crossing it would lose ten points on their exam scores. This practice, I admit, bordered on pedagogical cruelty.
A few brave souls crossed the box on the first quiz and received ten extra points. As the summer wore on, more and more students did so. One day I announced that more than half the students had crossed the box and that those who did had therefore been penalized ten points. Very few students crossed the box on the next exam. Gradually, however, the number crossing it edged up to around 40 percent of the class and stayed there. But it was always a different 40 percent, and it struck me that the calculation a student had to perform to decide whether to cross the box was quite difficult. It was especially so since the class was composed largely of foreign students who, despite my best efforts (which included this little game), seemed to have developed little camaraderie. Without
any collusion that I could discern, the students had to anticipate other students’ anticipations of their anticipations in a convoluted and very skittish self-referential tangle. Dizzying.
I’ve since learned that W. Brian Arthur, an economist at the Santa Fe Institute and Stanford University, has long used an essentially identical scenario to describe the predicament of bar patrons deciding whether or not to go to a popular bar, the experience being pleasant only if the bar is not thronged. An equilibrium naturally develops whereby the bar rarely becomes too full. (This almost seems like a belated scientific justification for Yogi Berra’s quip about Toots Shor’s restaurant in New York: “Nobody goes there any more. It’s too crowded.”) Arthur proposed the model to clarify the behavior of market investors who, like my students and the bar patrons, must anticipate others’ anticipations of them (and so on). Whether one buys or sells, crosses the box or doesn’t cross, goes to the bar or doesn’t go, depends upon one’s beliefs about others’ possible actions and beliefs.
The Consumer Confidence Index, which measures consumers’ propensity to consume and their confidence in their own economic future, is likewise subject to a flighty, reflexive sort of consensus. Since people’s evaluation of their own economic prospects is so dependent on what they perceive others’ prospects to be, the CCI indirectly surveys people’s beliefs about other people’s beliefs. (“Consume” and “consumer” are, in this context, common but unfortunate terms. “Buy,” “purchase,” “citizen,” and “household” are, I think, preferable.)
A M athematician Plays the Stock Market 9
Common Knowledge, Jealousy, and Market Sell-Offs
Sizing up other investors is more than a matter of psychology. New logical notions are needed as well. One of them,
10 J o h n Allen Paulos
“common knowledge,” due originally to the economist Robert Aumann, is crucial to understanding the complexity of the stock market and the importance of transparency. A bit of information is common knowledge among a group of people if all parties know it, know that the others know it, know that the others know they know it, and so on. It is much more than “mutual knowledge,” which requires only that the parties know the particular bit of information, not that they be aware of the others’ knowledge.
As I’ll discuss later, this notion of common knowledge is essential to seeing how “subterranean information processing” often underlies sudden bubbles or crashes in the markets, changes that seem to be precipitated by nothing at all and therefore are almost impossible to foresee. It is also relevant to the recent market sell-offs and accounting scandals, but before we get to more realistic accounts of the market, consider the following parable from my book Once Upon a Number, which illustrates the power of common knowledge. The story takes place in a benightedly sexist village of uncertain location. In this village there are many married couples and each woman immediately knows when another woman’s husband has been unfaithful but not when her own has. The very strict feminist statutes of the village require that if a woman can prove her husband has been unfaithful, she must kill him that very day. Assume that the women are statute-abiding, intelligent, aware of the intelligence of the other women, and, mercifully, that they never inform other women of their philandering husbands. As it happens, twenty of the men have been unfaithful, but since no woman can prove her husband has been so, village life proceeds merrily and warily along. Then one morning the tribal matriarch comes to visit from the far side of the forest. Her honesty is acknowledged by all and her word is taken as truth. She warns the assembled villagers that there is at least one philandering husband
A M athematician Plays th e Stock Market 11
among them. Once this fact, already known to everyone, becomes common knowledge, what happens?
The answer is that the matriarch’s warning will be followed by nineteen peaceful days and then, on the twentieth day, by a massive slaughter in which twenty women kill their husbands. To see this, assume there is only one unfaithful husband, Mr.
A. Everyone except Mrs. A already knows about him, so when the matriarch makes her announcement, only she learns something new from it. Being intelligent, she realizes that she would know if any other husband were unfaithful. She thus infers that Mr. A is the philanderer and kills him that very day.
Now assume there are two unfaithful men, Mr. A and Mr.
B. Every woman except Mrs. A and Mrs. B knows about both these cases of infidelity. Mrs. A knows only of Mr. B’s, and Mrs. B knows only of Mr. A’s. Mrs. A thus learns nothing from the matriarch’s announcement, but when Mrs. B fails to kill Mr. B the first day, she infers that there must be a second philandering husband, who can only be Mr. A. The same holds for Mrs. B who infers from the fact that Mrs. A has not killed her husband on the first day that Mr. B is also guilty. The next day Mrs. A and Mrs. B both kill their husbands.
If there are exactly three guilty husbands, Mr. A, Mr. B, and Mr. C, then the matriarch’s announcement would have no visible effect the first day or the second, but by a reasoning process similar to the one above, Mrs. A, Mrs. B, and Mrs. C would each infer from the inaction of the other two of them on the first two days that their husbands were also guilty and kill them on the third day. By a process of mathematical induction we can conclude that if twenty husbands are unfaithful, their intelligent wives would finally be able to prove it on the twentieth day, the day of the righteous bloodbath.
Now if you replace the warning of the matriarch with that provided by, say, an announcement by the Securities and Exchange Commission, the nervousness of the wives with the
12 J o h n Allen Paulos
nervousness of investors, the wives’ contentment as long as their own husbands weren’t straying with the investors’ contentment as long their own companies weren’t cooking the books, killing husbands with selling stocks, and the gap between the warning and the killings with the delay between announcement of an investigation and big sell-offs, you can understand how this parable of common knowledge applies to the market.
Note that in order to change the logical status of a bit of information from mutually known to commonly known, there must be an independent arbiter. In the parable it was the matriarch; in the market analogue it was the SEC. If there is no one who is universally respected and believed, the motivating and cleansing effect of warnings is lost.
Happily, unlike the poor husbands, the market is capable of rebirth.
Comentarii
Trimiteți un comentariu