We examine how much of the market return is driven by trading that takes place outside of normal hours.
Officially, the US equity market is open six and a half hours per day from 9:30am – 4:00pm. However, the market now also has pre- and post- market trading from 7:00am until open and from close until 8:00pm. Meanwhile, equity index futures trade almost 24 hours per day from Sunday evening through Friday evening. Therefore, between pre- and post-market trading and futures, there are significantly more outside-of-market hours during which equities and derivatives trade than there are normal market hours. By-and-large your average investor doesn’t do much buying or selling outside of normal market hours, nor do they invest in futures. This means that, for the most part, these investors simply “take” the portion of the market return that is generated outside of market hours. This made us wonder – how much of the market return is driven by trading that takes place outside of normal hours? If there are more total hours of trading when the market is closed than when it’s open could the out-of-market-hours trading account for more of the market’s return than the hours the market is open? If we didn’t have to worry about taxes or transaction or costs would we theoretically be better off not leaving exposure open during the off-market hours?
To answer these questions, we compared the returns from for S&P 500 (SPX) from 1/7/2008 – 6/11/25, segregating the returns from market close to market open the next day and the returns from market open to market close on the same day. We examined the time period since 2008 because, although S&P futures started trading in 1982, the S&P price history on FactSet does not display a separate opening value before the beginning of 2008. The results of our study can be seen below. The figures labeled “Day” show the S&P’s return from open to close ((Close Value/Open Value)-1); the “Overnight” figures show the returns from the prior close to the open ((Open Value/Prior Close Value)-1). Overnight returns are grouped with the day of the opening value, i.e. the returns from Thursday close through Friday open are classified as “Friday” and Friday close through Monday open would fall under “Monday,” etc.
As you can see, the average day's return is roughly two times the average overnight return and the average move during the day is more than twice as large as overnight. Prior to looking at the data, we speculated that we might see an outsized overnight move from Friday close to Monday open as there is more time for market-moving news to break. However, while the absolute move was slightly larger than other days, the difference wasn’t dramatic. Finally, in the bottom section of the table we can see that the direction of overnight trading has predicted the direction of trading during market hours more than half the time – the change overnight and the change during market hours were either both positive or both negative about 58% of the time.
Over the last 30 years, the amount of equity trading done electronically has increased significantly, which has allowed trading hours to be extended to the point that there are now more pre- and post-market trading hours than there are hours when the exchanges are officially open. However, since 2008 the traditional trading hours have still tended to produce larger moves and more than half of the S&P’s return since 2008 has come between 9:30 and 4:00. It’s also clear that we have been better off taking the “overnight risk.”
Of course, just because the bulk of the market return came during regular trading hours in our lookback period, that doesn’t mean this is always the case. The table below shows the average day and overnight returns for every year since 2008. In most instances, the bulk of the returns came during the day, but there have been a few years when most of the positive returns were generated overnight.
