New York Stock Exchange opens at 9:30 am and closes at 4:00 pm. But after its close, institutional investors can still trade. Other stock markets, including Tokyo Stock Exchange, Shanghai Stock Exchange, etc. work in a similar manner.
A basic pattern in modern stock markets is that stocks go up when the market is closed and stay flat when it is open. I mean, not always, obviously, but in general the market’s moves during the day add up to around zero, while the market’s moves overnight are positive. On average, stocks open higher in the morning than they closed the previous afternoon, but they close in the afternoon at roughly the same price as they opened that morning.
This is a weird pattern, and the explanations of it are not entirely satisfactory. One explanation that is fun and has gotten a lot of attention is that it is the result of a vast conspiracy by quantitative trading firms who manipulate markets by buying at the open every day and selling at the close. I would not say that I believe this explanation, precisely, but it is enjoyable. Any story of stock-market manipulation has to have the rough form “you buy stocks to make them go up, and then you sell them without making them go down,” and in general that is a dubious story: If your buying will make stocks go up, then your selling should make them go down, so your manipulation won’t work. But it is often the case that trading at the open is less liquid than trading at the close: Big index funds, etc., like to trade at the end of the day (to match the official trading price), so there is more liquidity, so your trading will move prices more in the morning and less at night. So if you own $1 billion worth of stock, and you buy $100 million worth every morning and sell $100 million worth every afternoon, you will tend to push prices up more in the morning and than you will push them down in the afternoon, which will make your portfolio more valuable. Obviously this is not investing advice.
Here is a fun paper from Victor Haghani, Vladimir Ragulin and Richard Dewey about the phenomenon, titled “Night Moves: Is the Overnight Drift the Grandmother of All Market Anomalies?” From the abstract:
We then take a closer look at the behavior of individual US stocks for clues about aggregate stock market behavior. We found that not only did the effect exist at the index level as previously reported, but it also shows up in a suggestively clustered pattern in individual stocks returns, and is particularly strong in “Meme” stocks. We find that a simple long-short portfolio that only takes exposure when the market is closed would have earned a return of 38% per annum (importantly, ignoring transactions costs) with an annualized Sharpe Ratio of about 3.
A day-trader who bought AMC Entertainment at the open and sold it at the close every day from the start of 2019 to late May 2022 would have suffered a 99.6% loss of capital – but, during the night, would have made a return of 30,000% over the same period (both ignoring transaction costs).
They suggest that the overnight effect might come, not from quantitative funds buying at the open to push up prices, but from retail traders buying at the open because they’ve had time to think about their orders:
Stock market liquidity is deeper at the close of the trading day, and shallower at the open. A given size trade executed at the open has a bigger price impact than at the close.
Retail investors place their orders more at the open, and institutional investors more at the close. This is seen from studies of brokerage trading records and analysis of the timing of small and large trades over the course of the day. Small trade sizes occur more towards the beginning of the day, and large trades later in the day. It seems reasonable that retail investors tend to make their single stock investment selections at leisure in the evenings or over the weekends, and then place their orders before going to work, which will often be executed at the open.
That is, the basic pattern might be that motivated traders buy mainly in the opening auction (when liquidity is bad), pushing prices up, so stocks open higher each day. Perhaps those motivated traders are quant traders trying to manipulate the market by consciously taking advantage of liquidity differentials, or perhaps they are just regular people who work 9-to-5 jobs and have to put their orders in before work.
Meanwhile why isn’t this arbitraged away? Why aren’t people buying at the close, selling at the open, and collecting all the excess returns? Haghani et al. have some theories, with the main one being transaction costs; another is that “the overnight-versus-intraday drift may be one of a number of anomalies caused by retail flows, making it a less attractive opportunity as part of a portfolio that already has a lot of exposure to strong retail flows.” I also like this one:
Risk tolerance: HFT and other market makers exhibit a strong preference to end the day with flat positions. They like to be able to manage their exposure minute to minute, and are averse to being locked in for hours or days (i.e. weekends and holidays). Similarly, mid-frequency statistical arbitrage firms like to end the day without significant factor exposures and are willing to pay to close positions.
You get paid more for holding stocks overnight because holding stocks overnight is less pleasant than holding them during the day: If things go wrong, you can’t sell. The people in the business of arbitraging stock prices are mostly in that business from 9:30 to 4; an arbitrage that requires buying all the stocks at 4 to hold overnight is not their business.