190 strategy tests on 10 years of data: what really works on Gold, Bitcoin and Indices
Every week a new "secret strategy" goes viral: Asia range breakouts, liquidity sweeps, New York open scalps. We wanted facts, not opinions, so we tested them on real broker history from 2016 to 2026 on 10 markets (Gold, Silver, Bitcoin, US500, US30, Nasdaq, EURUSD, GBPUSD, USDJPY, AUDUSD), always with spread and commission, and always split into two periods (2016–2021 and 2022–2026). A strategy only "passes" if it makes money in both periods.
1. Session strategies (100 tests)
- Liquidity sweep fade (London takes the Asia high/low and comes back): loses on every market. Win rate 25–30%, profit factor about 0.74. The most shared idea on social media was the worst one.
- Asia range breakout at London and New York opening range: lost in 2016–2021, small profits after 2022. Not stable.
- Previous day high/low breakout: close to zero after costs.
- Surprise: buying Gold only during the Asian session was profitable in both periods (PF 1.15 and 1.19), while buying Gold during London or New York lost money. Much of Gold's rise in the last decade happened in Asian hours.
2. Quant-style strategies (90 tests)
- Holding indices overnight: strongly negative after costs.
- Larry Williams volatility breakout and inside-day breakouts: around break-even.
- Buying short pullbacks inside an uptrend on stock indices (RSI-2 style, with a protective stop): the most robust result. US500: 97 trades, 80% winners, profit factor 2.27; US30 PF 1.77; Nasdaq PF 1.74 – profitable in both periods. It only works on the daily chart: on H1/H4 the costs eat the edge.
- Turn of the month (last day + first 3 days): positive on Gold and the three US indices in both periods, but with few trades per year.
3. Trend following on H4 (what still holds after 10 years)
A breakout with the main trend plus a volatility-based trailing stop – the family of rules used by CTAs and the original Turtles – was the most stable approach on Gold and Bitcoin. Tested as an automated strategy in the MT5 Strategy Tester (1% risk per trade, spread, commission and swaps included): Bitcoin H4 2017–2026 turned 10,000 into 34,294 USD with a 8.5% maximum equity drawdown; Gold H4 2016–2026 was also positive (+40.7%). Past results are not a promise, but they show which ideas survive real costs.
Conclusions
- Short intraday "session tricks" rarely survive spread and commission.
- What survives: following trends on H4/D1, and buying pullbacks in uptrends on stock indices.
- Size every trade from your risk, not from your feelings: the stop distance decides the lot.
We built these results into RGT Trend Rider (signals, trailing stop, lot calculator and real statistics on your own chart – free demo in the Strategy Tester) and a free lot calculator. Questions about the tests are welcome in the comments.


