Overnight vs Intraday Returns: When Does the Nasdaq Really Rise?

If you trade NQ, this is more than an academic curiosity. The ETF's overnight half is the futures market's Globex session, and the gap you see at 9:30 ET is that overnight return, already priced in.

Splitting the day in two

Take any trading day. The overnight return runs from yesterday's 16:00 ET close to today's 9:30 ET open. The intraday return runs from 9:30 to 16:00 ET. Compound the two and you get the full close-to-close return. Most charts and most performance numbers only show that total. The split shows when the money was actually made.

1. 01When did the Nasdaq-100 ETF earn its returns?

   QQQ, 1999–2006 (Kelly and Clark, 2011) · Close-to-open risk premium of about +23.7%, open-to-close about −23.3%, as reported by the authors

A visual summary of the article's table — not a performance chart or a new backtest.

Quick example, with round numbers, not real data: say the index closes at 100 on Monday. Overnight, futures rally and the cash index opens Tuesday at 101.5, an overnight return of +1.5%. During the day it drifts down and closes at 101, an intraday return of −0.49%. The day's total is +1.0%. A chart of daily closes shows a good day, but a trader who was only in from 9:30 to 16:00 lost money on it.

Repeat that shape over hundreds of days and the two halves can tell very different stories, even when the total looks totally normal.

The first finding: the Nasdaq-100 made its money at night

In a 2011 article in the Journal of Asset Management (vol. 12, no. 2), Returns in trading versus non-trading hours: The difference is day and night, Michael A. Kelly and Steven P. Clark compared risk-adjusted returns in and out of trading hours for broad index ETFs.

For the Nasdaq-100 ETF from 1999 to 2006, they report a geometric average close-to-open risk premium of about +23.7%, against about −23.3% from open to close. The overnight return was also the less volatile of the two. Over that period, an investor who was only in during market hours would have lost money on average, while one who was only in overnight would have made money.

The authors point out that this doesn't sit well with a simple view of efficient markets, where risk-adjusted returns shouldn't depend on the time of day.

The second finding: a tug of war

Dong Lou, Christopher Polk and Spyros Skouras pushed the question further in A tug of war: Overnight versus intraday expected returns, published in the Journal of Financial Economics in 2019 (vol. 134, no. 1). Looking at individual U.S. stocks, they found that:

• overnight returns and intraday returns each tend to persist, for years, within their own half of the day;
• across 14 well-known trading strategies, profits came almost entirely from one half of the day, either overnight or intraday;
• the two halves often had opposite signs, which the authors call a tug of war between different groups of investors.

Their explanation comes down to who trades when. Some investors tend to trade at the open, others during the day and at the close. When those groups favor different stocks or different directions, the overnight and intraday halves of the return pull against each other.

Why overnight returns might be different

The research and practitioners offer several explanations. None of them is settled.

• Payment for risk. Holding through hours when the cash market is shut is risky, since news can hit anytime, and investors may want a premium for that.
• Order flow at the open. Orders placed overnight by investors reacting to news, and filled at the open, can push opening prices one way.
• Different crowds. As Lou, Polk and Skouras argue, the investors who dominate the open aren't the same ones who dominate the day.
• Global information. Asian and European trading, and their news, happen during the U.S. night, and that shows up in the opening price.

What this means if you trade NQ

Your "overnight" is a real session. For the ETF, the night is a gap. For NQ, it's the Globex session, with its own highs, lows and volume. The cash index's opening gap is the net result of hours of futures trading.

The open is a handoff, not a starting point. At 9:30 ET, cash-market traders show up and either confirm or push back on the overnight move. That handoff is exactly the window our Daily Bias covers, from 9:30 to 12:00 ET.

Don't assume the 1999–2006 split still holds. It comes from one period that included a bubble and its collapse, and it should be measured again on recent data before anyone builds on it.

Know which half your strategy lives in. A strategy that only trades during the cash session is only exposed to the intraday half. If that half has been weak for long stretches, your strategy is swimming against the current.

How we plan to measure it

We'll publish our own split of the Nasdaq-100's returns into overnight and intraday parts, on QQQ and on NQ, once we've added daily open and close history. Here's the method, stated now so it can't be tweaked later:

1. Overnight return = previous close to today's open; intraday return = open to close.
2. Dividends go to the overnight return of the ex-dividend date.
3. Results are shown by year and by regime, over the full period, including the years when the pattern fails.
4. Costs: a strategy that only holds overnight trades twice a day, so we'll show results before and after realistic costs.

What the opening gap tells you each morning

Since the overnight return is known by 9:30 ET, it's one of the few pieces of information a morning read can use directly. Three things are worth noting before the open:

1. The size of the gap from yesterday's close to the current futures price, in points and in percent.
2. Where price sits in the overnight range. A gap holding near the overnight high is different from one that's already faded.
3. What caused it. A gap built on a scheduled release, like CPI at 8:30 ET, has a clear catalyst. One built slowly overnight on thin volume can get reversed more easily once cash traders show up.

Our pre-market checklist puts these three checks in its second step.

Common misreadings

• "Stocks only go up at night." The research describes averages over long periods, and plenty of nights are negative.
• "So I should only hold overnight." Trading twice a day costs commissions and spreads, and in a lot of accounts taxes too. A gross pattern isn't a net strategy.
• "Futures and ETFs are the same." They track the same index, but futures trade overnight, so the split looks different on an NQ chart.

Limits

• The Nasdaq-100 result comes from 1999–2006, a period that includes the dot-com collapse.
• The tug-of-war study covers individual U.S. stocks and strategies, not the index directly.
• Neither study tests the result after a retail trader's costs.

Sources

• Kelly, M. A., & Clark, S. P. (2011). Returns in trading versus non-trading hours: The difference is day and night. Journal of Asset Management, 12(2), 132–145. IDEAS/RePEc
• Lou, D., Polk, C., & Skouras, S. (2019). A tug of war: Overnight versus intraday expected returns. Journal of Financial Economics, 134(1), 192–213. LSE Research Online

FAQ

What is the overnight return? The return from one day's closing price to the next day's opening price.

Do stocks make most of their gains overnight? For the Nasdaq-100 ETF from 1999 to 2006, the overnight half was positive and the intraday half negative, according to Kelly and Clark (2011). Other periods may be different.

Why are overnight returns higher? Researchers have suggested payment for risk, order flow at the open, different groups of investors and global news. None of it is proven.

Does this apply to NQ futures? NQ trades through the night, so its "overnight" is a live session. The cash index's opening gap reflects that session.

Should I only hold positions overnight? Not based on this alone. Costs, taxes and the period studied all matter, so test it on recent data first.

---

Research and education only. Not financial advice. Past results do not predict future returns.

How this article was made: drafted with AI assistance from the research papers listed above; facts and figures checked by Gabriel Décarie and Karel Panneton.