Investors, in their panic, seemed determined to sell everything at once. Yet even on a day like this, there were winners as well as losers.

When War Picks Its Own Winners

US airlines, unsurprisingly, bore the brunt of the sell-off. Investors had already priced in fears of flying, canceled flights and tighter security measures. Shares in United Airlines fell by 43.2%. American Airlines lost 39.4%, Delta 44.6% and Southwest Airlines 24.1%.

Defense contractors told a different story. With spending on security and military operations clearly set to rise, their shares climbed even as the broader market sank. Lockheed Martin gained 14.7%. Northrop Grumman rose 15.7%, and Raytheon surged by as much as 26.8%. General Dynamics added approximately 9%.

In other words, the market behaved exactly as the textbooks predict it should in the event of a major terrorist attack or the outbreak of war. It is here, however, that a rather different story begins.

A closer look reveals that one large US investor had wagered on a fall in United Airlines stock through options just days before the attacks, and not modestly. On 6 September, a single US institutional investor accounted for 95% of all put option purchases on United Airlines.

The trade suggests that someone was betting heavily on a sharp fall in the airline's share price, and it quickly became grist for the mill of conspiracy theorists, who took it as a sign that someone somewhere had inside information. The information would have had to be reliable, since a failed bet of that size would have cost the investor dearly.

The reality, though, was rather more complicated. On 10 September, the same investor also bought 115,000 shares of American Airlines, a trade running in precisely the opposite direction. To a committed conspiracy theorist, this might look like an elaborate attempt to disguise the original bet from the US Securities and Exchange Commission (SEC). That is a plausible-sounding explanation, but plausibility is not proof of intent.

In fact, the two trades may not have been contradictory at all. The investor may simply have expected United Airlines to underperform American Airlines, betting against one and in favor of the other. Such relative bets are commonplace in financial markets. Rather than a cover-up, then, this could just as easily have been a fairly ordinary investment strategy.

Suspicion Is Not Proof

That question, precisely, was the one that needed answering. It fell to the SEC to answer it, and given the circumstances, the regulator's investigation was a thorough one.

The Commission began combing through the suspicious trades as early as 12 September 2001. By the time it was done, it had examined more than 9.5 million stock market transactions carried out in the weeks leading up to the attack. The results, however, did not bear out the conspiracy theory.

Neither the SEC nor the FBI turned up any evidence that anyone with advance knowledge of the attacks had traded on it. In the case of the United Airlines investor, no link to al-Qaeda could be established at all. The inquiry into the large volume of put options on American Airlines reached a similarly mundane conclusion: some of those purchases were traced to an American investment newsletter that had, as early as 9 September, urged its subscribers to bet against the airline's stock.

We could leave the story there, take the official record at face value and move on. That, though, would mean missing the more interesting part of it. For that, it is worth turning to the science.

In his study Unusual Option Market Activity and the Terrorist Attacks of September 11, 2001, the academic Allen Poteshman put the episode to a rigorous statistical test. His conclusion turned out to be rather less tidy than the one reached by the securities regulator.

Statistically speaking, trading in the options market really had been unusual, and Poteshman's own conclusion was accordingly a cautious one. The data, he found, were consistent with the possibility that someone had traded on foreknowledge of the attacks. That is a long way from proof that anyone actually did. His study does not identify who was behind the trades, nor does it establish any link between them and the terrorists themselves.

The Power of a Single Successful Bet

There is another possible explanation, one that requires no talk of terrorists or insiders: the so-called gray swan, a rare but entirely foreseeable event. The risk of a terrorist attack, after all, did not begin on 11 September. Throughout the summer of 2001, US security agencies had repeatedly warned of a heightened terrorist threat, and airlines themselves had been cautioned about the possibility of hijackings. None of this amounted to specific knowledge of what was being planned, but it hardly counted as a secret either.

Seen in that light, a bet against airline stocks looks considerably less like a shot in the dark. It remained, certainly, a risky piece of speculation. But the financial markets are never short of risky bets of exactly this kind, made by someone, somewhere, at almost any given moment.

Today's equivalent might be a bet against the artificial intelligence boom, on the assumption that the bubble will eventually burst. The overwhelming majority of such wagers go nowhere, and nobody pays them the slightest attention. Should one of them come off, though, it draws outsized media coverage and lodges itself firmly in the public imagination. That, in essence, is why bets on collapse hold such enduring appeal. The odds of success are slim, but for the rare investor who calls it right, the reward is not just profit. It is something closer to lasting fame.