Futures trading strategies

Without a prop firm there is no deadline, no trailing drawdown and no profit split — what you need instead is capital for the margin and the falls. Overnight margin on one micro Nasdaq contract is about $3,500 (day-trading margin about $100), so a swing rule needs a few thousand dollars per contract just to hold through a night, and a fall of 20% of the contract’s value ($10,000 on one MNQ) is an ordinary year. Costs are low (about $0.80 a side all-in on a micro at a discount broker, $2–3 on a mini) and futures gains are taxed 60/40 under US rules whatever the holding period. The retail record is poor: of Brazilian day traders who kept at index futures for 300 days, 97% lost money (Chague, De-Losso and Giovannetti 2020); most of the edges below are small, so size and costs decide them.

What this tool found: On one micro contract with these costs, the rules that made money over 25 years earned a few hundred to about two thousand dollars a year each — the same order as the margin they tie up (one MES held overnight earned about $900 a year after carry). Nothing here turns a small account into a living. What the better rules do is earn a share of the index’s return with a far smaller worst fall; the rest of the markets (oil, gas, currencies, the note) gave no rule that held up in both halves of the history on reliable prices.

Sources: Chague, De-Losso and Giovannetti (2020), “Day trading for a living?” · CME micro contract specifications; broker margin tables (autumn 2026) · IRS section 1256 (60/40 treatment of regulated futures)

Time-series momentum — buy what went up over the last year, sell what went down — earned a Sharpe ratio near 1 across 58 futures markets from 1985 to 2009 (Moskowitz, Ooi and Pedersen 2012), and the industry built on it (managed futures) made 27% in 2022. The catch for one person on one market: the per-market Sharpe is about 0.15 and most of the return is the market’s own drift (Nilsson); the diversified result needs 40–60 markets (Graham Capital 2017). Since 2010 the edge has been thinner and the falls long — the SG Trend index lost 20% from May 2024 to May 2025. The 200-day rule (Faber 2007) on stock indices earned about what holding did with half the worst fall.

What this tool found: Long only above the 200-day average kept most of holding’s return on the stock indices and gold with a worst fall of 8–16% of capital against holding’s 14–26%, and made money in both halves of 26 years on NQ, ES, YM and GC (front-month prices; on the ETF prices ES, RTY and GC). Twelve-month momentum, long only, did the same with more trades; the moving-average crossover and the Donchian breakout traded too rarely on one market to prove anything (a dozen trades in 25 years) and the breakout lost after costs on most. Shorting with trend rules lost on every stock index. Nothing beat holding the Nasdaq on return — the gain is the smaller fall.

Sources: Moskowitz, Ooi and Pedersen (2012), “Time series momentum”, Journal of Financial Economics · Faber (2007), “A quantitative approach to tactical asset allocation” · Nilsson, “An analysis of expected returns of trend-following strategies” (Quantpedia summary) · Graham Capital (2017), “Market diversification”; Cambridge Associates (2025) on the SG Trend index drawdown

Stock indices mean-revert over days: a sharp two-day dip in an uptrend has been bought back more often than not (Connors’ RSI(2), 2008), a close near the day’s low tends to be followed by an up day (Pagonidis 2013, the IBS effect), and three down closes in a row are more often than not followed by an up close. These rules win 65–80% of their trades and are in the market a tenth to a fifth of the time, so their worst falls are small; their total return is well under holding’s, and in 2008 and 2022 the dips kept dipping. Commodities and currencies trend day to day instead: the same rules lose there.

What this tool found: The steadiest results in the whole sweep. RSI(2) long on the S&P: 80% of 256 trades won over 34 years, both halves positive, worst fall 5% of capital, in the market 13% of the time, for about half of holding’s return. Three lower closes in a row (ES, NQ, RTY, ZN) and the close near the low (IBS: NQ, ES, YM) held up in both halves on the ETF prices and again on the real futures of the last two years (IBS on NQ, ES, YM, RTY: 66–75% won). The Bollinger and big-day fades were weaker; on the commodity and currency front-month histories these rules look spectacular only where the old quotes are unreliable (copper, gas, the euro) — see the warning on those prices.

Sources: Connors and Alvarez (2008), “Short term trading strategies that work” · Pagonidis (2013), “The IBS effect: mean reversion in equity ETFs” · Backtests of RSI(2) on SPY 2000–2024 (several authors): 70–80% wins, far smaller falls than holding, lower total return

Most of the stock market’s rise since the 1990s came between the close and the next open; the day session on its own earned little (Cooper, Cliff and Gulen 2008; Lou, Polk and Skouras 2019; Liu and Tse 2017 on index futures). Futures hold the night directly. The strongest overnight stretch — 2 to 3 am New York, as Europe opens — has faded since 2021 (Boyarchenko, Larsen and Whelan, New York Fed 2026). With a round trip every day, costs decide it: on one micro a few basis points of the contract’s value each night.

What this tool found: On the real futures of the last two years the night, above the 200-day average, made money in both halves on NQ, ES, YM, RTY, GC, SI and HG (54–60% of nights won; NQ $3,800 a year on one micro) while the day session earned nothing — the strongest short-history result here. Over 27 years of ETF prices the Nasdaq night still earned 60% of holding’s return with a third of its worst fall, but lost in the first half (1999–2012: the dot-com bear and 3–5% financing) and paid in the second; the S&P night was flat over 34 years after carry. A round trip every night costs about $1.60 on a micro — a tenth of what the night earned on NQ.

Sources: Cooper, Cliff and Gulen (2008), “Return differences between trading and non-trading hours” · Lou, Polk and Skouras (2019), “A tug of war: overnight versus intraday expected returns” · Boyarchenko, Larsen and Whelan (2026), “The overnight drift”, Federal Reserve Bank of New York

The turn of the month (the last session and the first three) has carried most of the S&P’s return since 1980 (Ariel 1987; Lakonishok and Smidt 1988; McConnell and Xu 2008); a Fed working paper found it gone in S&P futures after 1990, others still find it through 2024. November to April has beaten May to October in most markets for a century (Bouman and Jacobsen 2002) — the summer is flat on average, not negative. The day before a holiday paid many times an ordinary day until the 1980s (Ariel 1990) and has faded for large US stocks since. Day-of-the-week effects were large in the 1970s and are mostly gone (Grebe and Schiereck 2025, a meta-analysis).

What this tool found: The turn of the month made money in both halves on NQ, ES, YM and GC (front-month and ETF prices alike) with a fifth of the sessions in the market — $300–600 a year on a micro, a small, steady piece. The session before a holiday paid on most markets but only $100–300 a year. November–April kept holding’s return on the S&P with its 2008 and 2022 summers skipped, and nothing on the others. Picking weekdays found nothing that held in both halves on the stock indices.

Sources: McConnell and Xu (2008), “Equity returns at the turn of the month”, Financial Analysts Journal · Bouman and Jacobsen (2002), “The Halloween indicator”, American Economic Review · Ariel (1990), “High stock returns before holidays” · Grebe and Schiereck (2025), day-of-the-week effects, Eurasian Economic Review

Larry Williams’ volatility breakout (the open plus a share of yesterday’s range) made his 1987 trading-contest record; he writes it has worked less well in recent decades. The opening-range breakout on the Nasdaq ETF (Zarattini and Aziz 2023) reported an annualized alpha of 33% from 2016 to 2023; an independent replication found the edge survives only at very low costs, mostly from 2022, and a test on index futures found the 5-minute range the worst window after realistic costs.

What this tool found: The daily-bar volatility breakout lost on every market and every kind of prices after costs — the stop a share of the range beyond the open fills on the quiet days and gives it back by the close. The minute-by-minute versions in the Challenges Backtest (opening range, momentum breakout) tell the same story on real futures: close to a coin flip after costs, with the Nasdaq the best of a weak field. The one exception is the opening range scalp: stop orders at the 9:30–10:00 high and low that take only a few points (about 1.5% of the day’s range). On real Nasdaq futures it made money in both years and beat the same orders on other 30-minute ranges; on the S&P, Dow, Russell, gold and oil it lost. On the S&P a slower version held instead: a buy stop above the first hour’s high, a target and a stop about 30 points away, out after two hours — a profit factor of 1.4 in both years, but picked from thousands of settings, so treat it as weaker evidence.

Sources: Williams (1999), “Long-term secrets to short-term trading” · Zarattini and Aziz (2023), “Can day trading really be profitable?” · Zarattini, Barbon and Aziz (2024), “Beat the market: an effective intraday momentum strategy for the S&P 500 ETF” · Independent replications of the opening-range breakout with execution costs (2024–2025)

The academic intraday effects are small: the first half-hour’s direction predicts the last half-hour’s (Gao, Han, Li and Zhou 2018; an R² of about 2%), gap fills happen on most small gaps and fewer than half of large ones, and the pre-FOMC drift (Lucca and Moench 2015) is one or two days a month. The retail record on intraday futures is the worst of any style; what the minute strategies here have shown on two years of real futures is a profit factor near 1.0–1.3 before any prop-firm rule, with the Nasdaq’s overnight drift the steadiest part.

What this tool found: The nine minute strategies (momentum breakout, opening range, VWAP pullback and fade, gap fill, candle range, 10 am, opening range scalp, overnight drift + day trades) are run here on their own-account terms: a fixed number of micros, your broker’s costs, no challenge. Over the two years of real futures only the Nasdaq’s overnight drift + day trades, gap fill and opening range scalp made money steadily; the others were near zero or lost after costs. The scalp wins about 96% of its trades by a few points and loses 60–200 points on a handful a year, so size it by that loss — and 1-minute bars cannot show how its stop orders fill.

Sources: Gao, Han, Li and Zhou (2018), “Market intraday momentum”, Journal of Financial Economics · Lucca and Moench (2015), “The pre-FOMC announcement drift”, Journal of Finance · Chague, De-Losso and Giovannetti (2020) on retail day traders