The Survivor

Every plate so far has reasoned about the premium looking forward — in expectation, before the dice are cast. But the premium is a statement about averages, and the fortunes and ruins that risk-bearing actually produces are not averages: they are single outcomes, each the result of one throw, run once and never repeated. When we look at the world that risk has made — at the great fortunes above all — we are looking backward, after the throw, and after the throw the price of risk wears a disguise the elegant theory does not prepare us for. For after the throw, we can see only the survivors.

Consider what the premium promises and what it delivers. Beforehand, a risky venture offers a higher expected return than a safe one, in compensation for its risk; this is the premium, and it is real and fair. But "higher on average" is a statement about the whole distribution of possible outcomes, most of which never come to pass. When the venture is run, a single outcome is drawn: the founder is rich, or ruined. The realized return — what actually befell this one person — is not the premium; it is one sample from the distribution the premium summarized, and a single sample may lie anywhere, from total loss to vast fortune. The premium describes the cloud; the realized fortune is one raindrop, and from a single raindrop the cloud cannot be read.¹

Now add the fact that changes everything: we do not see the whole distribution of outcomes, because the losers vanish. The firms that failed are wound up and forgotten; the investors who were wiped out leave the market and the conversation; the ventures that came to nothing leave no memoir. What remains in view — the firms we study, the fortunes we admire, the track records we pore over — are the survivors, the favorable draws, and they are not a representative sample but a censored one, selected precisely for having won. This is survivorship bias, and it is no minor statistical caveat but the central fact about how the realized price of risk presents itself: a graveyard of the ruined that we never visit, and a gallery of the victorious that we mistake for the whole.²

From survivorship the volume's central inversion follows. Confronted with a survivor — the founder worth billions, the fund that beat the market for a decade, the speculator who called the crash — we reach instinctively for an explanation in skill, judgment, foresight, nerve. And some of it may be those things. But survivorship guarantees that, out of a great enough multitude bearing risk, some will win again and again by chance alone, and these chance winners are indistinguishable, from the outside and after the fact, from the genuinely gifted, for both display the identical record of success. The great fortune that looks like the wage of extraordinary skill is, in unknowable part, the wage of having drawn the favorable outcome from a distribution that ruined others who bore the very same risk no less ably. What presents itself as a return to skill is, in large measure, a return to risk-bearing whose realized form is luck — the luck, simply, of having survived. The figure is genius; the ground is the cemetery.³

And the disentangling cannot, in the single case, be done. To know whether a track record reflects skill or luck, one would need many independent repetitions — but a life affords few, a great fortune is usually made once, and by the time enough trials have accumulated to separate the able from the fortunate, the game has changed and the verdict is moot. Statisticians can sometimes prise skill from luck in the aggregate, across thousands of fund managers or firms, and when they do, the finding is sobering: the share attributable to persistent skill is small, and most of what looked like skill does not persist — as it would not, if it were luck. But for the individual fortune — this founder, this investor — the question the volume's subject poses, how much was courage and how much was luck, has no answer that can be read off the outcome, because the outcome is exactly what both skill and luck produce.⁴

There is a second, quieter strangeness, and it meets the first. Even setting individual fortunes aside and looking only at the average realized premium — the long-run excess return of risky stocks over safe bonds — the number is hard to explain. It is far too large, the puzzle runs, to be the compensation any plausible degree of risk aversion would demand for the risk that stocks actually carry: to justify the historical premium by risk aversion alone requires people implausibly, almost pathologically, afraid of risk. So either investors are far more averse than they appear, or the risk being priced is not the modest variability the model measures but something larger and rarer — the small chance of catastrophe, of a disaster that wipes out wealth, which history's relatively calm sample happens not to have realized; or the aversion is not to risk at all but to ambiguity, to the unmeasurable; or it is, in part, a fear the theory does not name. The average premium, like the individual fortune, seems to price more than risk, properly reckoned, can account for.⁵

None of this is to deny that risk has a real price, or that skill is real, and the careful statement matters. Beforehand, the premium is genuine: risk-averse people truly must be paid to bear undiversifiable risk, and that compensation is fair and necessary, the engine that funds enterprise and rewards those who back it. Skill, too, is real: some entrepreneurs and investors are genuinely better, and over enough trials it tells. The claim is narrower and sharper. It is that the realized rewards of risk-bearing — the fortunes we actually observe — are a sample censored by survivorship, in which luck and skill are fused beyond separation in any single case; that the premium we infer from history is inflated by the disasters that did not happen to strike; and that the universal human instinct to read a great fortune as a monument to its holder's virtue is, in large and unknowable part, the survivor's privilege of writing the history after the fact. The reward for bearing uncertainty is real; the story that it was deserved is mostly told by those who lived to tell it.⁶

The volume's subject asked how much of a great fortune is courage and how much is luck, and the economics of risk returns an answer of studied discomfort: that the question, asked of any single fortune, cannot be answered from the fortune itself, because courage rewarded and luck rewarded leave the identical trace — a survivor, holding the winnings. The premium is the price paid, in advance, for bearing what cannot be foreseen; this was Knight's insight, that profit beyond the wage of skill is the reward for bearing true, unmeasurable uncertainty, the residue that falls to whoever stakes himself against an unknowable future and is left standing. But "left standing" is the operative phrase, doing more work than it admits, for among those who staked themselves equally bravely, who is left standing is settled, in unmeasurable part, by luck — and the standing survivor, surveying a field cleared of those who fell, may be forgiven, but should not be believed, when he calls the clearing his own doing.⁷

So the fourth of the prices completes the lesson of the first three, and turns it inward, upon ourselves. The wage, we found, is read as worth but is partly power; the rent is read as contribution but is the toll of position; the rate is read as a fact of time but is in part a decision. The premium is read as the reward for skill and courage in facing the unknown — and is, in its realized form, the reward for having survived a gamble that ruined others as able, dressed by the survivors in the language of desert. In each, a payment that flatters its recipient as the just return to a virtue turns out to be, in large part, something cooler and less earned: a position held, a decision taken, a power across the table, a favorable draw. The price of risk is the most personal of the four, for it touches the stories we tell about success itself — and it suggests that the deepest premium the world pays is not for bearing risk, but for surviving it, and being there, afterward, to say the bearing was brave.⁸