Sell in May Is Dead: Why It Fails in 2026

The investing world is full of adages and stories about some period of the year being favorable or unfavorable to investors. One of the most famous ones is “Sell in May and go away”, also called the Halloween indicator or the Halloween effect.

It is the idea that the period from November to April sees better price action in stocks, and that the rest of the year, from May to Halloween, is better spent away from markets and assets held in cash. Another, more moderate version is that it is best to sell in May and avoid the summer in general, even if the time to re-enter the market might be closer to September.

This is one of these theories with a kernel of truth, but also hardly a good guide for investors. And in 2026, this seems more obsolete than ever.

Where the Halloween Effect Comes From

From Brokers’ Holidays to Academic Research

Initially, it seems that a part of this trend was born almost two centuries ago, when stock markets were still in their infancy.

The wealthy class, rich enough for such investment, would leave London and head to their country estates for the summer, largely ignoring their investment portfolios, only to return in September. A similar effect occurred in other major European nations, and in the USA, finance specialists would often leave New York for the Hamptons or Nantucket in Massachusetts.

“The Stock Exchange is in a sort of twilight state at the moment. The potential buyers seems to have  ‘sold in May and go away’.”

Financial Times, May 30, 1964

The term Halloween Indicator was further popularized by Sven Bouman and Ben Jacobsen and published in a 2002 paper in the American Economic Review.

According to this paper, an investor who followed this strategy would theoretically reap the best part of an annual return, but with just half the exposure of someone who invests in stocks year-round.

And according to historical data, this has been at least true for a large segment of market history. But this is maybe not so simple.

Did Sell in May Ever Work?

Economic researchers are somewhat doubtful of this adage that mostly came from traders and brokers and not from academic research, at least until the Bouman and Jacobsen 2002 paper.

In part, this is because according to the efficient market hypothesis, as this phenomenon becomes better known, its effects will be priced into the market, making the strategy ineffective.

For example, over a long time set, events like the 1987 Black Monday crash can impact the overall average returns and show that any strategy that avoids it outperforms buy-and-hold strategies. But as such black swan events are unpredictable, this is unlikely to be predictive of future returns.

Still, it seems the effect is somewhat true, even if less powerful than often claimed.

What Causes the Halloween Effect?

However, behavioral science indicates that some of the effects might stem from human behaviors. For example, investment professionals’ summer vacations can impact market liquidity, or investors’ aversion to risk during the summer months can reduce returns. Potentially, many companies’ staff and management are also not going to take risks or make important announcements during the summer months.

Interestingly, it seems that the effect is more pronounced in European markets, with the work culture of these countries known to “take more seriously” the summer holiday than in America, so this might be a reason.

Alternatively, markets are known to be reflective and to sometimes depend as much on participants’ beliefs as actual facts. If many people believe it is best to sell in May, this can in itself create a stronger selling pressure, reducing performance this month. So it could be that the effect that started with holidays far away from physical stock exchanges is now a self-fulfilling prophecy.

Does Sell in May Still Matter in 2026?

The Halloween Effect Is Fading

While well established in historical data, the mostly unexplained Halloween effect has been fading in recent years.

JP Morgan’s trading desk has been pointing out that, “over the past 10 years, the S&P 500 has averaged a return of 1.5% in May and a 1.9% pop in June. Returns are even stronger in July at an average of 3.4%.”

Even in Europe, supposedly a bastion of the “Sell in May” strategy, the recent data are not encouraging, according to Deutsche Bank.

“in 25 of 39 years, the ‘sell in May’ strategy underperformed a straightforward Buy and Hold … offering no statistical edge.”

 Always-Connected Markets Changed the Pattern

A likely factor is that even in 2002, most market participants of importance were trading mostly from their computer desk, in an office on Wall Street or the City of London. Abroad or in the countryside, Internet connections were poor to non-existent, so the case for a decline in trading activity and risk taking place during the summer was a logical one.

In our hyper-connected world, where even the most remote off-grid mountain chalet can get high-speed Internet through Starlink and solar panels, this is hardly the case anymore.

 A News-Driven Market Cannot Take Summer Off

Another factor is that the 2020s are one very, very busy decade. Whether it was the Covid pandemic, the quick rise in importance of AI and its constant shuffling of the industry’s leader, or the explosion of geopolitical tensions with the war in Ukraine and now Iran, market participants can hardly take 6 months off and find the world back in the same state as they left it.

This seems especially true in May 2026, with a looming energy and commodities (metals, fertilizer, food) crisis from the closing of the Hormuz Strait.

If the Strait reopens and the war stops, reacting quickly to a likely boom in stock markets will be the right position to take, not avoiding markets throughout May 2026. And if the war escalates and global stability keeps worsening, a quick rotation into the rare sectors likely to benefit will be equally important.

The same can be said for booming sectors like AI. Missing the next six months of market activity can turn into a massive disruption of compounding returns.

And there are also other historical-scale potential events planned in the upcoming months, like the upcoming mega-IPO of SpaceX after its merger with xAI, for example.

Every Market Edge Ends Eventually

Because markets are controlled by the actions of investors, the more a specific strategy is known and popular, the more it is likely to soon stop working. Either because this specific strategy has become a crowded space, or because the underlying assumptions that made it work for a while are no longer valid.

With its declining predictive value in the past decade, it seems this is the latter that is happening to the Halloween effect.

A world where trading is done from a phone app and where connectivity is as omnipresent as power supply simply does not see a significant decline in trading activity in summer.

Today’s market activity is driven by always-connected professional and retail investors, as well as algorithms that never sleep, let alone take a break during the summer.

Similarly, AI companies are not going to slow down their intense competition with each other one bit over the next 6 months, nor will companies developing new semiconductors, SMRs, etc.

So while it might have been true, it seems that May 2026 is part of a trend of irrelevance for one of the well-established, but increasingly outdated, adages of Wall Street.