Older people among my readers will remember the time when there was – for a while – a discussion about how the US stock market had significantly higher returns between yesterday’s close and today’s open (when there were no trades at all) than during the day.
I used to wonder if some of it may be due to Europe's longer trading hours spreading things out a bit more (US is 09:30-16:00, UK 08:30-16:30, Europe 09:00-17:30.
Since I read your study charts as also incorporating the 246 trading days a year that *don't* include quarterly earnings, could it also reflect a US regional aversion to after-hours / ECN trading? I.e., are Europeans more trusting? I recall New York traders refering to after-hours trading as "the badlands" with super-low liquidity and spreads you could drive a truck through ... only the naïve or the truly desperate traded off-exchange because you'd get picked off.
I have another theory: US public companies and markets act much more like clockwork on corporate announcements, especially earnings. Announcements on the tape at 16:01 *after* the closing bell. Give people an hour to digest the numbers. Then have an exactly one-hour-long management conference call to discuss the results 17:00-18:00. Then a few hours of major broker analyst call-back conversations. That means that when the market opens, everyone's had 17 1/2 hours to calmly plug all the new numbers into their spreadsheets, write their reports, think everything over, and get a good night's sleep before make an informed trading decision for the open. The opening bell's as orderly as the start of a horse race. Then managements tend to stay in the office for a few days after earnings somthey're reachable for follow- up questions.
Europe's a mess by comparison. Earnings tend to be released right *before* the open, leaving still-groggy investors and analysts scrambling to pore over the numbers even as trading commences. Managements don't schedule their earnings calls until the afternoon, ostensibly for the convienience of US investors, but which leaves six-seven trading hours in which the stock can swing around wildly without management input. This is made even worse by companies who hold in-person press conferences in the late morning, forcing the investment community to read *reporters'* interpretations of verbal management comments on the broad tape whilst waiting until they can finally clarify things from managements themselves later that day. In addition, these eventual management investor conference calls can stretch on for hours and hours, because so much misinterpretive damage had been done earlier in the day. Then, after the calls end, managements head off on two-week investor road shows, rendering them unreachable just when they're needed most to answer follow-up questions.
When I first started working in Europe, I was also shocked to receive a fax (remember those?) with company results a few hours before the official release (while markets were still open) because I worked for "the house broker". Perhaps that's to mitigate the fact that the "house broker" is also expected to feed all your competitors lunch at corporate presentations on your premises. Unless this has been subsequently addressed, there's also a cohort who gets the numbers early.
I don't see in the study any mention of futures and options.
Option broker dealers, who are typically short puts or long calls, are long the lesser known option Greek, Vanna (how delta changes when implied volatility changes)
When implied volatility decreases overnight (a common scenario when volatility futures are in contango), Vanna causes the delta of these options to decrease, leading dealers to buy back futures contracts to remain delta-neutral. This buying pressure supports futures prices during overnight sessions, especially when liquidity is low.
The overnight effect of Vanna is usually supportive (bullish) in normal market conditions, but during periods of market stress or uncertainty, this effect can reverse, causing selling pressure instead. For example, before risky events, protective puts become expensive, dealers hedge by shorting futures, and volatility spikes, reversing the usual Vanna-driven overnight price support.
The interaction between volatility crush (falling implied volatility) and dealer hedging can create feedback loops where dealers reduce short hedges by buying futures, which further compresses volatility and encourages more buying.
There was a couple of chats about this on Andrew Gelman's blog a while back with the author and folks generally were skeptical of his methods and claims.
I used to wonder if some of it may be due to Europe's longer trading hours spreading things out a bit more (US is 09:30-16:00, UK 08:30-16:30, Europe 09:00-17:30.
Since I read your study charts as also incorporating the 246 trading days a year that *don't* include quarterly earnings, could it also reflect a US regional aversion to after-hours / ECN trading? I.e., are Europeans more trusting? I recall New York traders refering to after-hours trading as "the badlands" with super-low liquidity and spreads you could drive a truck through ... only the naïve or the truly desperate traded off-exchange because you'd get picked off.
I have another theory: US public companies and markets act much more like clockwork on corporate announcements, especially earnings. Announcements on the tape at 16:01 *after* the closing bell. Give people an hour to digest the numbers. Then have an exactly one-hour-long management conference call to discuss the results 17:00-18:00. Then a few hours of major broker analyst call-back conversations. That means that when the market opens, everyone's had 17 1/2 hours to calmly plug all the new numbers into their spreadsheets, write their reports, think everything over, and get a good night's sleep before make an informed trading decision for the open. The opening bell's as orderly as the start of a horse race. Then managements tend to stay in the office for a few days after earnings somthey're reachable for follow- up questions.
Europe's a mess by comparison. Earnings tend to be released right *before* the open, leaving still-groggy investors and analysts scrambling to pore over the numbers even as trading commences. Managements don't schedule their earnings calls until the afternoon, ostensibly for the convienience of US investors, but which leaves six-seven trading hours in which the stock can swing around wildly without management input. This is made even worse by companies who hold in-person press conferences in the late morning, forcing the investment community to read *reporters'* interpretations of verbal management comments on the broad tape whilst waiting until they can finally clarify things from managements themselves later that day. In addition, these eventual management investor conference calls can stretch on for hours and hours, because so much misinterpretive damage had been done earlier in the day. Then, after the calls end, managements head off on two-week investor road shows, rendering them unreachable just when they're needed most to answer follow-up questions.
When I first started working in Europe, I was also shocked to receive a fax (remember those?) with company results a few hours before the official release (while markets were still open) because I worked for "the house broker". Perhaps that's to mitigate the fact that the "house broker" is also expected to feed all your competitors lunch at corporate presentations on your premises. Unless this has been subsequently addressed, there's also a cohort who gets the numbers early.
I don't see in the study any mention of futures and options.
Option broker dealers, who are typically short puts or long calls, are long the lesser known option Greek, Vanna (how delta changes when implied volatility changes)
When implied volatility decreases overnight (a common scenario when volatility futures are in contango), Vanna causes the delta of these options to decrease, leading dealers to buy back futures contracts to remain delta-neutral. This buying pressure supports futures prices during overnight sessions, especially when liquidity is low.
The overnight effect of Vanna is usually supportive (bullish) in normal market conditions, but during periods of market stress or uncertainty, this effect can reverse, causing selling pressure instead. For example, before risky events, protective puts become expensive, dealers hedge by shorting futures, and volatility spikes, reversing the usual Vanna-driven overnight price support.
The interaction between volatility crush (falling implied volatility) and dealer hedging can create feedback loops where dealers reduce short hedges by buying futures, which further compresses volatility and encourages more buying.
https://www.linkedin.com/pulse/magic-overnight-stock-market-returns-david-steets/
https://systematicindividualinvestor.com/2021/03/12/tradable-effects-of-options-market-liquidity-flows/
https://arxiv.org/abs/2107.12516
https://arxiv.org/pdf/2010.01727
Seems these research papers -same author show the ON returns were a thing for the past 3 decades in ALL western global stock exchanges
Interesting, because I can't see any evidence for overnight returns in the UK and europe in my data...
There was a couple of chats about this on Andrew Gelman's blog a while back with the author and folks generally were skeptical of his methods and claims.
https://statmodeling.stat.columbia.edu/2023/02/06/stocks-go-down-during-the-day-and-up-at-night/
https://statmodeling.stat.columbia.edu/2025/04/19/for-15-years-tesla-stock-has-been-edging-down-during-the-day-and-shooting-up-overnight/