Night and day: overnight against intraday returns in S&P 500 futures

Abstract

We backtested the overnight drift rules on Micro E-mini S&P 500 (MES) 5-minute bars from 2016-01-11 to 2026-10-30: 2,643 trades, with commission and a tick of slippage on every market and stop fill. In sample (to 2023-08-03) it made −0.00R per trade with a profit factor of 0.97; out of sample (from 2023-08-04) +0.02R per trade and 1.22 over 794 trades. 16 of the 16 parameter settings we tried were profitable out of sample. Sized at $250 of risk a trade, simulated 50K evaluations passed 0% to 2% of the time across 36 tracked plans.

1Hypothesis

Split every day into the hours the cash market is closed and the hours it is open. Then hold the future only overnight, from 15:55 to 09:35 ET.

If, as studies of stock indices found, most of the long-run gain came overnight, holding only overnight should keep most of the return with less time in the market.

2Data

Instrument

Micro E-mini S&P 500 (MES), traded on the ES price series (ES.c.0: front month, rolled on the calendar). CME Globex via Databento.

Period

2016-01-11 to 2026-10-30: 2,746 sessions, 3,770,901 one-minute bars, built into 210,816 5-minute bars of the regular session (09:30 to 16:00 ET).

Costs

$0.75 commission a contract each way and 1 tick ($1.25) of slippage on every market and stop fill. Limit orders fill when price touches them; market orders fill at the next bar's open and stops at their price, each with one tick of slippage; when a bar touches both the stop and the target, the stop counts.

Position size

About $250 of risk a trade: micro contracts, rounded down, at least one. R is a trade's net P&L divided by its risk at entry.

Out of sample

The last 30% of weekdays: 2023-08-04 on (846 days), after 1,974 in-sample days. The rules were fixed before either side was run.

3Method

Our reading of the publicly described rules; the creator may trade it differently. The rules, exactly as the code runs them:

1. Entry: buy at entry_min (15:55 ET, as minutes after midnight: 955), every regular session.
2. Exit: sell at exit_min (09:35 ET: 575) the next session, after its first 5-minute bar. Weekends and holidays are held through.
3. Stop: 2 times the 10-day average true range below the entry, for a defined risk.
4. Size: about risk_usd dollars of risk a trade (micro contracts, rounded down, at least 1 and at most max_contracts).

1"""Overnight against intraday: hold Micro E-mini S&P 500 (MES) from the close to the next open, long only.2 3Rules:4  1. Entry: buy at `entry_min` (15:55 ET, as minutes after midnight: 955), every regular session.5  2. Exit: sell at `exit_min` (09:35 ET: 575) the next session, after its first 5-minute bar. Weekends6     and holidays are held through.7  3. Stop: 2 times the 10-day average true range below the entry, for a defined risk.8  4. Size: about `risk_usd` dollars of risk a trade (micro contracts, rounded down, at least 1 and at9     most `max_contracts`).10"""11import qy12 13 14class Bot(qy.Strategy):15    """Own the market overnight only."""16 17    symbol = "MES"18    timeframe = "5m"19    session = "rth"20    params = {"entry_min": 955, "exit_min": 575, "risk_usd": 250, "max_contracts": 50}21    max_position = 5022    overnight = True23 24    def setup(self):25        self.day = None26        self.prev = None          # the close of the session before the last completed one27        self.prev_close = None    # the latest bar's close28        self.trs = []29 30    def _size(self, risk_points):31        per_contract = risk_points * self.point_value32        n = int(self.p.risk_usd // per_contract) if per_contract > 0 else 133        return max(1, min(self.p.max_contracts, n))34 35    def on_bar(self, bar):36        if bar.date != self.day:37            if self.day is not None and self.prev_close is not None:38                if self.prev is not None:39                    self.trs = (self.trs + [max(self.dh, self.prev) - min(self.dl, self.prev)])[-10:]40                self.prev = self.prev_close41            self.day, self.dh, self.dl = bar.date, bar.high, bar.low42        else:43            self.dh, self.dl = max(self.dh, bar.high), min(self.dl, bar.low)44        self.prev_close = bar.close45        t = bar.dt.hour * 60 + bar.dt.minute + 546        if self.position.is_long and t == self.p.exit_min:47            self.close_position("morning exit")48            return49        if t != self.p.entry_min or not self.position.is_flat or len(self.trs) < 10 or self.orders:50            return51        atr = sum(self.trs) / 1052        self.buy(self._size(2 * atr), stop_price=bar.close - 2 * atr, tag="Overnight long")53 

4Results

5Robustness

We reran the strategy on a grid of 16 settings of its two main parameters, each on the full history with the same split. 5 were profitable in sample and 16 out of sample.

6Prop-firm implications

The backtest's sessions run through the rules of 36 tracked 50K evaluations: 2,000 runs of up to 60 sessions each, with the drawdown and daily limits checked against each trade's intraday low. Intervals are Wilson 95% intervals of the simulation. A simulation of the past, not the chance that anyone passes.

7Limitations

• Long only in a rising market: the result mixes the overnight effect with the market's drift.
• Weekends and holidays are held, with their gap risk.
• A backtest on the past, with fills simulated from one-minute bars; live fills, queues and outages differ.

8Conclusion

Depends on the period: it lost in sample (profit factor 0.97) and made money out of sample (1.22). Nothing on the grid helps. Its in-sample best, entry (minutes after midnight ET) 900 and exit (minutes after midnight ET) 630, had profit factor 1.02 in sample but 1.16 out of sample, against 1.22 for the defaults. A rule that loses for years and then wins says more about the market changing than about the rule: the out-of-sample gain needs data after 2026-10-30 before it means anything.

9References

1. Cliff, M., Cooper, M. J., and Gulen, H. (2008). Return Differences between Trading and Non-Trading Hours: Like Night and Day. Working paper, SSRN.