【The More You Trade, the More You Lose】7 strategies tested over 22,786 candlesticks: 0 trades earned 25.71%, while 1,714 trades lost 11%
My AI ran 7 popular strategies on BTC 1h over 22,786 candlesticks (2024-01-16 to 2026-08-22, starting with $10k, 10% position size, and 0.1% fee per side). The results revealed a counterintuitive pattern: the number of trades and annualized returns were almost perfectly inversely correlated—the strategy that traded the most lost the most.
Here are the 7 strategies ranked by number of trades. The numbers speak for themselves:
- 0 trades (buy and hold): +25.71% annualized
- 70 trades (Turtle): +0.89%
- 118 trades (grid): -1.44%
- 188 trades (ICT/SMC): -1.95%
- 294 trades (AI trend): -2.37%
- 1,267 trades (RSI): -8.95%
- 1,714 trades (MACD): -11.27%
And this isn’t a matter of “low win rates.” MACD had a 30.51% win rate, and RSI had a 37.96% win rate—both far higher than the Turtle strategy’s 12.86%—yet every strategy with more trades lost money. A high win rate doesn’t necessarily mean you’re making money, and these numbers make that crystal clear.
What’s behind this? A 0.1% fee per side adds up to more than 3,000 instances of friction across 1,714 trades, wiping out the 30% win-rate advantage and then some. The “win rate” of a high-frequency strategy is an illusion: it wins often, but makes little per trade and gets stopped out quickly. Once costs are deducted, returns turn negative. That’s also why the two strategies with the fewest trades (0 and 70) were the only ones with positive returns.
These results don’t prove that “doing nothing always wins,” but they do show one thing: trading more doesn’t mean earning more. More often than not, you’re just paying more fees to the exchange and exposing yourself to more volatility. How many times do you trade on an average day? Have you actually factored in fees and slippage to see whether you’ve really outperformed simply holding?
$BTC
#Crypto
My AI ran 7 popular strategies on BTC 1h over 22,786 candlesticks (2024-01-16 to 2026-08-22, starting with $10k, 10% position size, and 0.1% fee per side). The results revealed a counterintuitive pattern: the number of trades and annualized returns were almost perfectly inversely correlated—the strategy that traded the most lost the most.
Here are the 7 strategies ranked by number of trades. The numbers speak for themselves:
- 0 trades (buy and hold): +25.71% annualized
- 70 trades (Turtle): +0.89%
- 118 trades (grid): -1.44%
- 188 trades (ICT/SMC): -1.95%
- 294 trades (AI trend): -2.37%
- 1,267 trades (RSI): -8.95%
- 1,714 trades (MACD): -11.27%
And this isn’t a matter of “low win rates.” MACD had a 30.51% win rate, and RSI had a 37.96% win rate—both far higher than the Turtle strategy’s 12.86%—yet every strategy with more trades lost money. A high win rate doesn’t necessarily mean you’re making money, and these numbers make that crystal clear.
What’s behind this? A 0.1% fee per side adds up to more than 3,000 instances of friction across 1,714 trades, wiping out the 30% win-rate advantage and then some. The “win rate” of a high-frequency strategy is an illusion: it wins often, but makes little per trade and gets stopped out quickly. Once costs are deducted, returns turn negative. That’s also why the two strategies with the fewest trades (0 and 70) were the only ones with positive returns.
These results don’t prove that “doing nothing always wins,” but they do show one thing: trading more doesn’t mean earning more. More often than not, you’re just paying more fees to the exchange and exposing yourself to more volatility. How many times do you trade on an average day? Have you actually factored in fees and slippage to see whether you’ve really outperformed simply holding?
$BTC
#Crypto