Backtests of the gold, US30 and NAS100 scalping strategies popular in trading reels, alongside research-based ones, with real spreads, commissions, swaps, minimum sizes and an out-of-sample check. Generated by the open-source Ledger Terminal backtester. Not financial advice.
The short answer
Best strategy found—
Gold, long from the 18:00 New York reopen for about 9 hours. Profitable in both test periods.
Its typical month at 1% risk—
Median over 10 years of hourly data.
Highest win rate seen—
—
30–40% a monthNot found
Not for any strategy, market or risk level. The risk table below shows why.
What higher risk does to the best strategy
Each row replays the strategy's real 10-year trades 2,000 times at that risk per trade. Monthly return grows with risk until losing streaks start to compound, then the chance of halving the account rises steeply.
Risk per trade
Typical month
Bad month (1 in 10)
Account halves within a year
Wiped out
Every strategy, judged on data it was not tuned on
Settings were chosen on the first 70% of each history and then tested on the last 30%. Profit factor above 1 means the strategy made money after costs; win rate alone says nothing about that.
Market · bars · period
Strategy
Win rate in → out
Profit factor in → out
Typical month (1% risk)
Verdict
Month by month: gold Asian-hours drift, 1% risk
How the tests were run
Data: Dukascopy spot XAUUSD, US30 and NAS100 candles. 10 years of hourly bars, plus minute bars for the 5- and 15-minute tests (gold 2 years; indices a shorter, preliminary sample).
Entries at the next bar's open. Stop-loss checked before take-profit when both fall inside one bar. Buys pay the spread; shorts stop on the ask.
Costs: gold $0.30/oz spread plus commission; NAS100 1.0 and US30 2.0 points of spread; estimated overnight swap at the 17:00 New York rollover; 1:100 leverage; minimum size 1 oz (gold) or 0.1 (indices).
Past results do not predict future returns. Live trading adds slippage, requotes and spread spikes that a backtest can only approximate.