Which Trading Strategies Actually Work? 15 Backtested on Real Data
Every trading strategy looks good in a hand-picked chart. The real question is what happens when you take every signal, for years, and pay the spread on each one. That is what the WickCheck strategy library does: popular strategies, tested exactly as written on real candles, with spread, commission and 1% risk per trade.
Here is the scoreboard for 15 of the most popular setups, what held up, what did not, and the patterns behind the results.
- 6 of 15 strategies made money on their main test after costs; the other 9 lost or roughly broke even.
- The ones that held up were simple, slower rules on daily charts: RSI(2) pullbacks, the 20-day breakout, the Bollinger bounce and the golden cross.
- Most popular intraday ICT and smart money setups landed below a profit factor of 1 once every signal and costs were counted.
- The market mattered as much as the rule: the same strategy often won on indices and lost on forex, or the reverse.
In this guide: How the strategies were tested · The scoreboard · Pattern 1: slow, simple rules held up best · Pattern 2: many intraday setups lose to costs · Pattern 3: the market matters as much as the rule · How to use these results
How the strategies were tested
Each strategy is written in plain English, exactly as it is usually taught, and run on real historical candles. Every trade risks 1% of the account, spread and commission are included, signals are only taken on closed candles, and when a stop and a target fall inside the same candle, 1-minute data decides which came first where it is available.
The main market and timeframe for each test are listed below, together with the number of trades. A profit factor above 1 means the rules made money after costs. Numbers are from the WickCheck strategy library as of 18 November 2026. The library re-runs every test as new candles arrive, so the live pages may differ slightly.
The scoreboard
- Golden cross (SPY, daily, 33.8 years): 15 trades, profit factor 3.53.
- RSI(2) pullback (S&P 500 futures, daily, 10 years): 87 trades, profit factor 1.88.
- 20-day breakout (gold, daily, 10 years): 82 trades, profit factor 1.68.
- Bollinger band bounce (S&P 500 futures, daily, 10 years): 66 trades, profit factor 1.42.
- NY session liquidity sweep + FVG (gold, 15m, 3 years): 67 trades, profit factor 1.16.
- MACD crossover (EUR/USD, 4h, 10 years): 498 trades, profit factor 1.09.
- Break of structure (EUR/USD, 1h, 10 years): 2,005 trades, profit factor 0.95.
- EMA crossover 9/21 (EUR/USD, 1h, 10 years): 1,257 trades, profit factor 0.94.
- Equal highs and lows sweep (gold, 15m, 3 years): 608 trades, profit factor 0.92.
- Fair value gap (EUR/USD, 1h, 10 years): 2,753 trades, profit factor 0.92.
- Liquidity sweep reversal (EUR/USD, 1h, 10 years): 4,872 trades, profit factor 0.91.
- RSI oversold bounce (EUR/USD, 1h, 10 years): 154 trades, profit factor 0.91.
- ICT order block (EUR/USD, 15m, 3 years): 1,859 trades, profit factor 0.81.
- London session sweep + FVG (GBP/USD, 15m, 3 years): 42 trades, profit factor 0.75.
- ICT Silver Bullet (Nasdaq futures, 5m, 1 year): 27 trades, profit factor 0.46.
Pattern 1: slow, simple rules held up best
The four strongest results all come from daily charts and rules you can write in one sentence. They trade rarely (from about 9 trades a year for RSI(2) to a handful per decade for the golden cross), so costs barely touch them, and each trade has room to move.
That does not make them easy to trade. Our losing streaks guide shows why slow strategies still test your patience: the 20-day breakout had two losing years on gold, and the golden cross can go years between signals.
Pattern 2: many intraday setups lose to costs
The setups with the most trades, fair value gaps, liquidity sweeps, order blocks, EMA crossovers, all sit below a profit factor of 1 (0.81 to 0.95). That is roughly what you would expect from rules with no real edge: close to break-even before costs, slightly negative after them.
It does not mean these concepts are useless. It means the plain, textbook version taken on every signal is not enough. The edge, if there is one, comes from what you add: which market, which session, which filter. See overtrading for why extra trades rarely help.
Pattern 3: the market matters as much as the rule
The same rule can win on one market and lose on another. RSI(2) was profitable on 9 of 10 markets in its test, including every index, but lost on Apple. The New York sweep + FVG model made money on gold and GBP/USD and lost on EUR/USD, Nasdaq and S&P 500 futures. Our NY session strategy write-up covers that one in detail.
So the useful question is never just 'does this strategy work?' but 'does it work on the market and timeframe I trade?'
How to use these results
- Start from something that held up, then test it on your own market and timeframe.
- Check the year-by-year table on each library page. An edge that came from one great year is fragile.
- Look at the trade count. Under about 30 trades, as with the golden cross and Silver Bullet, the result can be luck either way. See sample size.
- Test your own version. Every library page has a button that opens the exact test in WickCheck, where you can change any word and run it again. Our guide on how to backtest a trading strategy covers the process.
Bottom line. Most popular strategies did not survive real data and real costs. The ones that did were simple and slow, and even those depended on the market. Test before you trust.
Open a strategy from the library, change a word, pick your market and see the result in seconds.
Questions
- Which trading strategy is the most profitable?
- In WickCheck's library the highest profit factors came from slow daily strategies: the golden cross (3.53 on SPY, but only 15 trades), RSI(2) pullbacks (1.88 on S&P 500 futures) and the 20-day breakout (1.68 on gold). Results depend heavily on the market and period.
- Do ICT strategies work?
- The textbook versions tested in WickCheck's library mostly lost money after costs. The New York sweep + fair value gap model was an exception on gold (profit factor 1.16) and GBP/USD (1.48), but it lost on EUR/USD and US index futures.
- How were these strategies tested?
- On real historical candles with spread and commission included, 1% risk per trade, signals only on closed candles, and 1-minute data deciding which came first when stop and target fall in the same candle (where available).
Keep reading
Backtest a trading strategy properly: precise rules, real data, costs, enough trades, out-of-sample checks and the mistakes that create fake results.
20-Day Breakout (Donchian) Strategy: 9 Markets, 10 Years TestedThe Turtle-style 20-day breakout, backtested on gold, indices, oil, EUR/USD and stocks over 10 years with costs. Profitable on all 9 markets tested.
ICT New York Session Strategy: Liquidity Sweep + FVG TestedAn ICT-style New York session model, liquidity sweep then 15-minute fair value gap, backtested for 3 years on gold, GBP/USD, EUR/USD, Nasdaq and S&P 500.
EMA Crossover Strategy on Gold: What the Backtest ShowsWe tested the 9/21 EMA crossover with a 200 EMA filter on gold, EUR/USD and the Nasdaq, with costs included. Gold lost money after costs over 10 years.
Educational content, not investment advice. Trading involves risk of loss.