Why do stocks gains come at night?

The tendency of stocks to produce all their gains at night, when markets are closed, and systematically lose money during the daylight hours, has baffled researchers for four decades and potentially put retail investors at a disadvantage.

The debate over what drives this behavior will surely be revived by the award of the prestigious Harry Markowitz prize for the best paper of the year in investment management to Victor Haghani, Vladimir Ragulin and Richard Dewey for “Night Moves: Is the Overnight Drift the Grandmother of All Market Anomalies?”

In 22 of 24 countries, investors would have lost money since 1990 holding the major stock market index only from market open to market close - selling at the close and buying back at the open - even without transaction costs. The losses were not small; in seven countries they were over 90% of capital. All the substantial gains in global stock markets came from holding stocks overnight, from market close to the next day’s open. If you want to check your own portfolio, the paper’s authors have made an online calculator available.

The anomaly was first identified in 1986 and has since resisted explanation. It is statistically much stronger than most other well-known anomalies, but harder to take advantage of, since the transaction costs of buying at the close every day and selling at the open exceed the profits for most investors. Moreover, it does not apply to all individual stocks, only to large diversified portfolios and indices.

The main contribution of the Haghani paper is to offer a “meme stocks” explanation. The authors demonstrate that the effect is strongest for stocks that generate retail excitement, including things such as Bitcoin exchange-traded funds and Cathie Wood’s ARK Innovation ETF, and weaker or missing for duller stocks.

The simple story is that retail investors pick stocks mostly outside of their working hours and put in orders to be executed at the open. This pushes up opening prices, inflating overnight returns and reducing intraday returns. While retail investors also sell at the open, buying tends to concentrate in a few stocks with meme news, while selling of stocks whose meme status has cooled is more diverse and more often done with limit rather than market orders.

The authors do not claim this is the only factor contributing to overnight drift, nor that it explains everything about the phenomenon. Physicist and former DE Shaw quant Bruce Knuteson has argued since 2016 for a quite different and more ominous explanation. He has referred to the academic and regulatory literature (but not the Haghani paper specifically) as “whack-a-mole,” trotting out multiple explanations to explain individual aspects of the behavior, and putting up new ones when old ones are refuted.

Knuteson is a conspiracy theorist. I use that term in a neutral sense - a simple explanation for a host of puzzles that is logical, but that requires coordinated malfeasance by many nominally responsible people without any leaks. Most, but by no means all, conspiracy theories turn out to be false.

Knuteson thinks large quantitative market-neutral trading firms systematically inflate their portfolios before, at and shortly after opening; and reduce them back down around the close. Because markets are much less liquid around the opening, inflating positions has a large effect on prices. The reductions during more liquid times, therefore, do not remove the entire price effect. This causes the firms to make systematic mark-to-market profits. There are many possible explanations for why firms might do this for their long positions more aggressively than their shorts, including some of the costs and difficulty of borrowing securities to place short bets.

No conspiracy is needed for this part. Firms could stumble onto this strategy in a number of ways without setting out to manipulate prices. The conspiracy claim is that the trading firms have figured out the manipulation, even if it was not their original intention, and that other market participants and regulators are either aware of the scam or are content to see no evil. So long as the strategy tends to push prices up, everyone is content. If the strategy starts pushing index prices down, firms will quickly fix it to prevent that, for example, by ramping up the strategy on long positions and dampening it for short positions.

Rising stock markets make investors happy and mean more earnings for financial firms. Those things make regulators and politicians happy. The extra trading firms do to accomplish this strategy puts money in the pockets of dealers and market makers. According to Knuteson, all these people realize that it’s in their interests to avoid asking questions. The only losers are investors when the price inflation reaches a point at which economic fundamentals require a crash.

One of Knuteson’s strongest pieces of evidence is that China, the only country that outlaws buying a stock and selling it on the same day, is the only country where the overnight drift is reversed.

The “whack-a-mole” literature has turned up many smaller and more complex partial explanations that don’t fit into Knuteson’s theory. The meme-stock explanation, for example, requires no conspiracy, just some of the more foolish and aggressive retail traders being sloppy about transaction costs. It does not imply that equity values are inflated from fundamental values.

Most proposed market anomalies evaporate soon after being discovered. A few survive long enough to be named “risk factors,” no longer considered anomalous. Overnight drift does not fit either pattern. It’s as persistent as established risk factors such as value, momentum, quality and size, but not sufficiently well understood to be incorporated into factor portfolios. Perhaps the Haghani paper will help build that understanding. But if Knuteson is correct, it is a toxic factor that cheats investors and builds market instability. Either way, it deserves the attention that the Markowitz award may bring.

This column does not necessarily reflect the opinion of the editorial board or Bloomberg LP and its owners. Aaron Brown is a former head of financial market research at AQR Capital Management. He is also an active crypto investor, and has venture capital investments and advisory ties with crypto firms.