I’ve been watching systematic trading strategies for about three years now. The gap between backtest perfection and live performance still surprises me. You read about a strategy returning 47% annually across 2,400 trades, then you put actual money behind it during a crude oil shock and those clean equity curves look different.

Automated trading works brilliantly until the market does something your dataset never saw.

I learned this during the March 2024 swing when Brent spiked from $83 to $97 in 11 days. My momentum system kept buying into strength while smart money was already rotating out. Cost me 8.2% that month.

The First Real Test Nobody Warns You About

Most backtests assume you’ll execute every signal perfectly. You won’t. I’ve missed 14 entry signals over the past year because I was in a meeting or my internet cut out or I second-guessed the setup.

In September I skipped what looked like a false breakout signal on a banking stock. It ran 12% over six days and my system’s annual return dropped from 31% to 26% because of that single decision. Now I track “system return” versus “actual return” in separate columns, and the gap tells me how much my emotions cost.

You can game this stuff for entertainment like people do with superbull gaming setups where there’s no real capital risk. But when your rent money is riding on whether a 20-day EMA crosses a 50-day EMA, your brain does weird things. I’ve watched myself rationalize skipping bearish signals four times in a row because the macro picture looked strong.

Why I Started Caring About Position Sizing More Than Win Rate

Early on I obsessed over win rate. If a strategy won 58% of the time versus 54% I thought that mattered. It doesn’t. What matters is how much you lose when you’re wrong versus how much you make when you’re........

© Blitz