Against the Gods: The Remarkable Story of Risk Read online

  The optimism of the Victorians was snuffed out by the senseless destruction of human life on the battlefields, the uneasy peace that followed, and the goblins let loose by the Russian revolution. Never again would people accept Robert Browning's assurance that "God's in his heaven:/All's right with the world." Never again would economists insist that fluctuations in the economy were a theoretical impossibility. Never again would science appear so unreservedly benign, nor would religion and family institutions be so unthinkingly accepted in the western world.

  World War I put an end to all that. Radical transformations in art, literature, and music produced abstract and often shocking forms that stood in disturbing contrast to the comfortable modes of the nineteenth century. When Albert Einstein demonstrated that an imperfection lurked below the surface of Euclidean geometry, and when Sigmund Freud declared that irrationality is the natural condition of humanity, both men became celebrities overnight.

  Up to this point, the classical economists had defined economics as a riskless system that always produced optimal results. Stability, they promised, was guaranteed. If people decided to save more and spend less, the interest rate would fall, thereby encouraging investment or discouraging saving enough to bring matters back into balance. If business managers decided to expand their firms rapidly but households failed to save enough for them to borrow what they needed for expansion, the interest rate would rise to set matters right. Such an economy would never suffer involuntary unemployment or disappointing profits, except perhaps during brief periods of adjustment. Although individual firms and investors took risks, the economy as a whole was risk-free.

  Such convictions died hard, even in the face of the economic problems that emerged in the wake of the war. But a few voices were raised proclaiming that the world was no longer what once it had seemed. Writing in 1921, the University of Chicago economist Frank Knight uttered strange words for a man of his profession: "There is much question as to how far the world is intelligible at all.... It is only in the very special and crucial cases that anything like a mathematical study can be made."' During the depths of the Great Depression, John Maynard Keynes echoed Knight's pessimism:

  We are faced at every turn with the problems of Organic Unity, of Discreteness, of Discontinuity-the whole is not equal to the sum of the parts, comparisons of quantity fail us, small changes produce large effects, and the assumptions of a uniform and homogeneous continuum are not satisfied.'

  In 1936, in his masterwork, The General Theory of Employment, Interest and Money, Keynes flatly rejected Devon's faith in the universal applicability of measurement: "[Most of our decisions] to do something posi tive ... can only be taken as a result of animal spirits ... and not as the outcome of a weighted average of quantitative benefits multiplied by quantitative probabilities. "3

  Faced with the tensions of the postwar years, only the most naive theorist could pretend that all problems could be solved through the rational application of differential calculus and the laws of probability with well-ordered preferences. Mathematicians and philosophers had to admit that reality encompassed entire sets of circumstances that people had never contemplated before. The distribution of odds no longer followed the distribution Pascal had defined. It violated the symmetry of the bell curve and was regressing to means that were far more unstable than what Galton had specified.

  Researchers sought for ways of conducting a systematic analysis of the unexpected. Before the war they had concentrated on the inputs that went into decision-making. Now they recognized that the decision is only the beginning. The devil is in the consequences of our decisions, not in the decisions themselves. As Robert Dixon, an Australian economist, has remarked, "Uncertainty is present in the decision-making process, not so much because there is a future as that there is, and will be, a past.... We are prisoners of the future because we will be ensnared by our past."4 That ultimate realist, Omar Khayyam, had had the same thought nearly a thousand years before:

  What do you do when a decision leads to a result that was not even contemplated in your set of probabilities? Or when low-probability outcomes seem to occur more frequently than they should? Don't the patterns of the past always reveal the path to the future?

  Knight and Keynes, the first two to confront such questions in a serious fashion, were both noisy nonconformists, but, together, they defined risk as it has come to be understood today.

  Frank Knight was born on a farm in White Oak Township, Illinois, in 1885, the oldest of eleven children.' Though he lacked a high-school diploma, he attended two tiny colleges, perhaps the best he could afford in view of his family's poverty. The first was American University (which had no connection to the university with the same name in Washington, D.C.); this college emphasized temperance above all else and even taught "the principles of political economy in regard to the use of intoxicating liquors." In its national advertising it urged "parents to send their hard-to-handle boys to American University for disciplining." The second college was Milligan. On Knight's graduation, the president of the college described him as "the best student I have had ... best read student ... [with] practical business capacity as well as technical knowledge."

  Knight claimed that the reason he became an economist was that plowing was too hard on his feet. Before turning to economics he did graduate work in philosophy at Cornell; he switched to economics after a professor declared, "Stop talking so much, or leave the philosophy department!" But it was not just the overuse of his high, squeaky voice that got him into trouble; one of his philosophy professors predicted, "He will destroy the true philosophic spirit wherever he touches it." Knight was an incurable cynic about human nature. A more sympathetic professor once told him, "You came out of a malodorous environment where every man with a mind doubts everything."

  Knight began teaching economics at the University of Iowa in 1919 and moved to the University of Chicago in 1928. He was still teaching there when he died in 1972 at the age of 87; "It beats working for a living," he once remarked. His lectures were often ill prepared, delivered in a rambling, country-boy manner, and larded with heavy-handed humor.

  Despite his early exposure to religion and his continuing study of religion throughout his life, Knight was an implacable enemy of everything to do with organized forms of religion. In his presidential address to the American Economic Association in 1950, he likened the pope to Hitler and Stalin. He once said that religion was responsible for his bad sleeping habits: "It's that damned religion. I just can't get it out of my mind."

  An irascible, dedicated, honest man, he took a dim view of people who took themselves too seriously. He claimed that economic theory was not at all obscure or complicated, but that most people had a vested interest in refusing to recognize what was "insultingly obvious." Noting a quotation by Lord Kelvin chiseled in stone on the social science building at Chicago-"[W]hen you cannot measure it ... your knowledge is of a meager and unsatisfactory kind"-Knight sarcastically interpreted it to mean, "Oh, well, if you cannot measure, measure anyhow."6

  Knight's cynicism and concern for moral values made it hard for him to come to terms with the selfishness, and frequently the violence, of capitalism. He despised the self-interest that motivates both buyers and sellers in the marketplace, even though he believed that only selfinterest explains how the system works. Yet he stuck with capitalism, because he considered the alternatives unacceptable.

  Knight had no interest in working up empirical proofs of his theories. He harbored too many doubts about the rationality and consistency of human beings to believe that measuring their behavior would produce anything of value. His bitterest sarcasm was reserved for what he saw as "the near pre-emption of [economics] by people who take a point of view which seems to me untenable, and in fact shallow, namely the transfer into the human sciences of the concepts and products of the sciences of nature."

  The attitude reflected in this remark is evident in Knight's doctoral dissertation, which was completed at Cornell in 1916 an
d published as a book in 1921. Risk, Uncertainty and Profit is the first work of any importance, and in any field of study, that deals explicitly with decision-making under conditions of uncertainty.

  Knight builds his analysis on the distinction between risk and uncertainty:

  Uncertainty must be taken in a sense radically distinct from the familiar notion of Risk, from which it has never been properly separated. ... It will appear that a measurable uncertainty, or "risk" proper ... is so far different from an unmeasurable one that it is not in effect an uncertainty at all.'

  Knight's emphasis on uncertainty decoupled him from the predominant economic theory of the time, which emphasized decision making under conditions of perfect certainty or under the established laws of probability-an emphasis that lingers on in certain areas of economic theory today. Knight spoke of the failure of the probability calculus to, in Arrow's words, "reflect the tentative, creative nature of the human mind in the face of the unknown." 8 Clearly Knight was a creature of the twentieth century.

  The element of surprise, Knight argued, is common in a system where so many decisions depend on forecasts of the future. His main complaint against classical economics with its emphasis on so-called perfect competition arose from its simplifying assumption of "practical omniscience on the part of every member of the competitive system."9 In classical economics, buyers and sellers, and workers and capitalists, always have all the information they need. In instances where the future is unknown, the laws of probability will determine the outcome. Even Karl Marx, in his dynamic version of classical economics, never makes reference to forecasting. In that version, workers and capitalists are locked in a drama whose plot is clear to everyone and whose denouement they are powerless to change.

  Knight argued that the difficulty of the forecasting process extends far beyond the impossibility of applying mathematical propositions to forecasting the future. Although he makes no explicit reference to Bayes, he was dubious that we can learn much from an empirical evaluation of the frequency of past occurrences. A priori reasoning, he insisted, cannot eliminate indeterminateness from the future. In the end, he considered reliance on the frequency of past occurrences extremely hazardous.

  Why? Extrapolation of past frequencies is the favored method for arriving at judgments about what lies ahead. The ability to extrapolate from experience is what differentiates adults from children. Experienced people come to recognize that inflation is somehow associated with high interest rates, that moral character is desirable in the choice of whom we play poker with and whom we marry, that cloudy skies frequently presage bad weather, and that driving at high speed along city streets is dangerous.

  Business managers regularly extrapolate from the past to the future but often fail to recognize when conditions are beginning to change from poor to better or from better to worse. They tend to identify turning points only after the fact. If they were better at sensing imminent changes, the abrupt shifts in profitability that happen so often would never occur. The prevalence of surprise in the world of business is evidence that uncertainty is more likely to prevail than mathematical probability.

  The reason, Knight explains, is this:

  [Any given] "instance" . . . is so entirely unique that there are no others or not a sufficient number to make it possible to tabulate enough like it to form a basis for any inference of value about any real probability in the case we are interested in. The same obviously applies to the most of conduct and not to business decisions alone.10 (Italics are mine.)

  Mathematical probabilities relate to large numbers of independent observations of homogeneous events, such as rolls of the dice-in what Knight describes as the "apodeictic certainty" of games of chance.' But no event is ever identical to an earlier event-or to an event yet to happen. In any case, life is too short for us to assemble the large samples that such analysis requires. We may make statements like "We are 60% certain that profits will be up next year," or "Sixty percent of our products will do better next year." But Knight insisted that the errors in such forecasts "must be radically distinguished from probability or chance.... [I]t is meaningless and fatally misleading to speak of the probability, in an objective sense, that a judgment is correct."12 Knight, like Arrow, had no liking for clouds of vagueness.

  Knight's ideas are particularly relevant to financial markets, where all decisions reflect a forecast of the future and where surprise occurs regularly. Louis Bachelier long ago remarked, "Clearly the price considered most likely by the market is the true current price: if the market judged otherwise, it would quote not this price, but another price higher or lower." The consensus forecasts embedded in security prices mean that those prices will not change if the expected happens. The volatility of stock and bond prices is evidence of the frequency with which the expected fails to happen and investors turn out to be wrong. Volatility is a proxy for uncertainty and must be accommodated in measuring investment risk.

  Galton, a Victorian, would have expected prices to be volatile around a stable mean. Knight and Bachelier, neither of them a Victorian, are silent on precisely what central tendency would prevail, if any. We will have more to say about volatility later on.

  Knight disliked John Maynard Keynes intensely, as he revealed when, in 1940, the University of Chicago decided to award Keynes an honorary degree. The occasion prompted Knight to write a rambling letter of protest to Jacob Viner, a distinguished member of the Department of Economics at Chicago. Viner, Knight declared, was the person reported to be responsible "more than anyone else" for the decision to honor Keynes and therefore was "the appropriate party to whom to express something of the shock I received from this news."13

  Knight grumbled that Keynes's work, and the enthusiasm with which it had been greeted by academics and policymakers, had created "one of my most important ... sources of difficulty in recent years." After crediting Keynes with "a very unusual intelligence, in the sense of ingenuity and dialectical skill," he went on to complain:

  I have come to consider such capacities, directed to false and subversive ends, as one of the most serious dangers in the whole project of education.... I regard Mr. Keynes's [views] with respect to money and monetary theory in particular ... as, figuratively speaking, passing the keys of the citadel out of the window to the Philistines hammering at the gates.

  Although most of the free-market economists at Chicago disagreed with Keynes's conviction that the capitalist system needed a frequent dose of government intervention if it was to survive, they did not share Knight's disdain. They deemed it fit to honor Keynes as a brilliant innovator of economic theory.

  Knight may simply have been jealous, for he and Keynes shared the same philosophical approach. For example, they both distrusted classi cal theories based on the laws of mathematical probability or assumptions of certainty as guides to decision-making. And they both despised the "the mean statistical view of life."14 In an essay written in 1938, titled "My Early Beliefs," Keynes condemns as "flimsily based [and] disastrously mistaken" the assumption of classical economists that human nature is reasonable.15 He alludes to "deeper and blinder passions" and to the "insane and irrational springs of wickedness in most men." These were hardly the views of a man who was passing the keys of the citadel to the Philistines hammering at the gates.

  Knight may have been annoyed that Keynes had carried the distinction between risk and uncertainty much further than he himself had carried it. And he must surely have been angered when he discovered that the sole reference Keynes made to him in The General Theory of Employment, Interest and Money was in a footnote that disparages one of his papers on the interest rate as "precisely in the traditional, classical mould," though Keynes also conceded that the paper "contains many interesting and profound observations on the nature of capital."16 Only this, after Knight's pioneering explorations into risk and uncertainty fifteen years before.

  Keynes was from the opposite end of the intellectual and social spectrum from Knight. He was born in 1883 to an aff
luent, wellknown British family, one of whose ancestors had landed with William the Conqueror. As Robert Skidelsky, his most recent biographer, describes him, Keynes was "not just a man of establishments, but part of the elite of each establishment of which he was a member. There was scarcely a time when he did not look down at England, and much of the world, from a great height."17 Among Keynes's close friends were prime ministers, financiers, philosophers Bertrand Russell and Ludwig Wittgenstein, and artists and writers such as Lytton Strachey, Roger Fry, Duncan Grant, and Virginia Woolf.

  Keynes was educated at Eton and Cambridge, where he studied economics, mathematics, and philosophy under leading scholars. He was a superb essayist, as he demonstrated in presenting his controversial ideas and proposals.

  Keynes's professional career began with an extended stint at the Treasury, including service in India and intense involvement in Treasury activities during the First World War. He then participated as chief Treasury representative at the Versailles peace negotiations after the war. Finding the treaty so vindictive that he was convinced it would lead to economic turmoil and political instability, he resigned his post to write a book titled The Economic Consequences of the Peace. The book soon became a best seller and established Keynes's international reputation.

  Keynes subsequently returned to his beloved King's College at Cambridge to teach, write, and serve as the college's bursar and investment officer, all this while serving as chairman-and investment manager-of a major insurance company. He was an active player in the stock market, where his own fortunes fluctuated wildly. (Like many of his most famous contemporaries, he failed to predict the Great Crash of 1929). He also enriched King College's wealth by risktaking on the Exchange. By 1936, Keynes had built a personal fortune from a modest inheritance into the equivalent of £10,000,000 in today's money.18 He designed Britain's war financing during the Second World War, negotiated a large loan by the United States to Britain immediately after the war, and wrote much of the Bretton Woods agreements that established the postwar international monetary system.