TradingLab’s “86% win rate” MACD strategy, tested on real data

TradingLab’s “BEST MACD Trading Strategy [86% Win Rate]” has been watched about 4.5 million times. It pairs the MACD with a 200-period moving average and promises one of the highest win rates in trading. We wrote its rules down exactly as the video describes them and backtested them on three and a half years of real crypto futures data, with trading fees, slippage and funding included.

The rules, as the video gives them

• Long: the MACD line crosses above its signal line while both are below zero, and the price is above the 200 moving average.
• Short: the mirror image. The MACD crosses below its signal line above zero, with the price below the 200 moving average.
• Stop loss: just beyond the 200 moving average, which “acts as a wall”.
• Take profit: 1.5 times the risk.

How we tested it

Where the video leaves something open, here is exactly what we chose, so you can disagree with it:

• MACD 12, 26, 9, the standard settings that TradingView plots. (The video describes the two lines as a 12- and a 26-period moving average; the standard MACD is the difference between those two, against a 9-period signal line.)
• 200 EMA for the moving average. The video doesn’t say which kind.
• Stop 0.1% beyond the 200 EMA, as it stood when the trade opened, and a target at 1.5× that distance.
• 1% of the account risked per trade. The video doesn’t give a size.
• 15-minute, 1-hour and 4-hour charts on BTC, ETH and SOL perpetual futures, January 2023 to September 2026. The video doesn’t give a timeframe.
• Trades fill at the next candle’s open, with Binance’s fees, slippage and real funding payments.

Two things we did not test. The video’s “improved” version only takes trades at a support or resistance level you draw yourself, which is a judgement call no backtest can reproduce fairly. And the video was made with forex and stocks in mind; this is a test on crypto, where our tool works.

The results

The best result, Ethereum on 4-hour candles, finished +0.7% over three and a half years, which is break-even. The 15-minute versions were the worst by far: over a thousand trades each, and on Bitcoin the fees alone came to $5,030 on a $10,000 account.

Is it just the fees?

Partly. So we ran it again the flattering way many tools do by default: no fees, no funding, and every trade filled at the price that gave the signal. Bitcoin on 1-hour candles went from −29.0% to −3.9%, and on 15-minute candles from −81.3% to −19.2%. The win rate barely moved: 38–40%. Even with every cost removed, it still didn’t make money.

Why 40% is the number that matters

With a target of 1.5 times the risk, each win makes 1.5 units and each loss costs 1. To break even you need to win 1 trade in 2.5, which is 40%, before any fees. This strategy’s win rate sat right around that line in every test. In other words, the MACD signal added almost no edge over chance, and trading costs pushed it below zero.

The video’s title says 86%, but the video itself doesn’t show a test of that number, only one example trade. That’s the gap a backtest fills. For more on why a high win rate on its own tells you nothing, see our test of the 90% win rate strategy.

See every trade

All nine backtests are public. Open one for the full report, every trade on the chart, and month-by-month results.

Try your own version

Think a different timeframe, a wider target or another coin would change the verdict? Describe it and see: