Lamont, who is a portfolio manager at the $195 billion quantitative hedge fund and has been a faculty member at Harvard University, Yale School of Management, University of Chicago Graduate School of Business, and Princeton University, looked back at financial history and found a startling pattern.

“Even if systematic equities aren’t your thing,” he wrote in July 2025 on his Acadian blog, Owenomics, “you need to be mentally prepared for an epic financial disaster over the coming three months.”

His research draws a direct line between the timing of many of the most devastating financial crises and a centuries-old pattern: Market crashes tend to cluster during the so-called harvest time, spanning August to October.

The historical pattern

“For grizzled practitioners of systematic equity strategies,” Lamont writes, “August is the cruelest month.” He cast his mind back to the “quant quake” of August 2007, writing that analysts ever since have spent August “compulsively checking our phones and having nightmares about screens full of glowing red numbers.”

When reached for comment in August 2025, Lamont said every year around this time, panic is “certainly on my mind,” as it is for any quant equities managers who is over 50 years old.

Although overshadowed by the onset of the Great Financial Crisis in September 2008, the 2007 quant crash was a classic fit, Lamont writes, occurring during a sleepy time in markets when liquidity is thin because so many traders are away from their desks. Lamont cites modern research showing that August and September are periods of unusually low trading liquidity, as investors and market makers take summer vacations in the Northern Hemisphere. Lower market liquidity means less capacity to absorb big, sudden trades—a recipe for outsize volatility if a crisis does erupt.

Looking at the past 50 years, Lamont underscored the fact most major U.S. market crises have struck between August and October, when thinner markets amplified shocks. Among the historic market meltdowns during these months were two in September: 1998’s collapse of Long-Term Capital Management and 2008’s Lehman Brothers bankruptcy, and two in October: 1987’s Black Monday stock market crash and 1997’s Asian financial crisis. But going back to the founding of the United States itself, he sees a similar pattern.

The deep roots of harvest time

Lamont wrote that America’s first bubble, “Scriptomania,” occurred in July/August 1791, and the Panics of 1857 and 1873 occurred in August and September, respectively. Then the Panic of 1907 followed in October.

The culprit is clear to Lamont: summer vacation. But, in a chicken-or-the-egg discussion, he argues America’s agricultural economy created the need for time off in the summer, as that was when harvests occurred and money needed to flow from the big East Coast cities and into the Western agricultural regions.

Lamont cited Oliver Mitchell Wentworth Sprague’s diagnoses of “panic season” in 1910’s History of Crises Under the National Banking System: “With few exceptions all our crises, panics, and periods of less severe monetary stringency have occurred in the autumn, when the western banks, through the sale of the cereal crops, were in a position to withdraw large sums of money from the East.” The pattern was spotted as far back as 1884 by English economist William Stanley Jevons. The creation of the U.S. Federal Reserve system itself was in part a reaction to such panics, Lamont adds, citing a 1986 American Economic Review article by Jeffrey Miron.

“If you do the rough math, there’s a 10% chance of an epic disaster between August and October this year, and just a 2% chance from November through the following July,” Lamont writes, cautioning investors to “be mentally prepared” for outsize risk in the coming quarter.

Lamont told Fortune that a market crash is still a “rare event,” and he wasn’t aware of any particularly levered players in the market that could spark a crash. But then again, he added, he wasn’t aware of any in August 2007 when the quant crash happened.

Lamont’s summer of 2026: panic season, live

A year on from that original conversation, Lamont has spent the summer of 2026 documenting a strange incarnation of panic season—not a crash, but a market that looks calm on the surface while churning wildly underneath it. As of late August, the S&P 500 hasn’t moved more than 1% in either direction on a single day since hitting a record high on Aug 13.

In a column titled “Crazy days in the stock market,” he catalogued single-day swings that wouldn’t have been out of place in his original panic-season essay: Microsoft’s market cap rose $450 billion on July 30 (by Lamont’s own measure, “1.04 Houstons,” using the Texas city’s entire taxable property base as a yardstick), while Apple lost $360 billion the very next day. Daily dispersion that week ranked third-highest since 2015, trailing only “vaccine Monday” in November 2020 and the DeepSeek shock of January 2025. Lamont told Fortune it was a “crisis-like mechanism on a small scale,” noting that some levered hedge funds got wiped out to the tune of tens of billions of dollars.

Lamont has also used the summer to flag two other symptoms he associates with late-stage market euphoria. In “Hynix Hijinks,” he pointed to SK Hynix’s Nasdaq ADR listing—the largest foreign equity sale in U.S. history—trading at a 49% premium to its Korean shares within days, calling it a “law of one price” violation of the kind that shows up “during stock market bubbles.” In June, Lamont warned “the whirlwind is upon us,” calculating that April and May 2026 ranked as the fourth- and third-highest dispersion months for global stocks since 1995, trailing only December 1999 and February 2000, the peak of the dot-com bubble.

“The chamber of dispersion has been opened,” he wrote, “the beast of volatility has awakened, and the season of chaos is at hand.”

None of this is precisely the calendar-seasonality argument Lamont made in 2025; it’s a companion diagnosis, built on dispersion and correlation rather than the calendar. But the throughline is the same instinct: Markets that look serene are often anything but, and the surface calm itself may be the thing to distrust. With harvest time now officially underway and Lamont’s own dispersion data flashing dot-com-era readings, his 2025 warning to “be mentally prepared” reads less like a seasonal reminder and more like a live diagnosis.

As of August 2025, Lamont told Fortune he hadn’t changed his mind—his “back-of-the-envelope math” indicated a 10% chance of a huge disaster between August and October.

“So we’ve seen no disasters in 2025 and none yet in 2026 (knock on wood),” about what you’d expect for two years. Low summer liquidity is visible in lower trading volume for August 2026, with one recent day having the third-lowest trading volume for the entire year.

Remote harvest time?

Fortune asked Lamont in 2025 if the harvest/panic season thesis has something in common with “flash crashes,” which often occur overnight, after trading in America ends and before it starts in Asia. He said that’s a bit of an extreme vacation of an illiquid market, “like what would happen if everyone went asleep.” He reiterated his belief “weird stuff happens” in illiquid markets. Then he got philosophical about how economics requires all of us to have some kind of appetite for weirdness.

What about Europe, which traditionally takes much longer vacations in August, sometimes the whole month, compared to Americans and their much more reserved time-off policy? Lamont agreed, but noted that with America as the world’s global financial center, with a much larger market, the impact of thinner liquidity is felt more strongly. He noted other academics have covered seasonalities in other countries, such as Australia, where it seems to be the opposite case, or the impact of seasonal affective disorder on trading in Northern countries.

Ultimately, he told Fortune, the benefits of the current system outweigh the risks. The “traditional, heavy-handed approach,” he said, would be to shut the market down, calling off trading in August altogether.

Lamont told Fortune of his upbringing in the two schools of economics that revolve around heavy regulation and libertarianism, with the East Coast “saltwater” tradition he learned at MIT a major influence on him before he spent eight years on faculty at the libertarian “freshwater” school, the University of Chicago.

“A basic principle of economics is you should let people trade,” he said, before adding that he also believes in behavioral finance, which holds that “people mess up and markets make mistakes.” He believes governments make mistakes, too, he added.

The whole issue may be resolved over time by the rise of remote work, he added.

“One theory would be that because nowadays we can all work remotely, vacations are less impactful on [trading] volume,” he said. In fact, in both August 2025 and 2026, Lamont added, he worked remotely from his summer house in Maine.

For now, he added, we are trapped in the paradox of tradition that began with our agricultural economy. People take vacation in August because that’s when people take vacation.

“Especially with family gatherings,” he said, “you want to be on vacation the same time your relatives are on vacation.”

How’s that for behavioral finance?

For this story, Fortune journalists used generative AI as a research tool. An editor verified the accuracy of the information before publishing.

A version of this story was originally published on Fortune.com on August 10, 2025.