When you build a strategy, the temptation is to optimize for the equity curve going up and to the right. Bigger return, higher Sharpe, prettier chart. I did this for months. Then I blew an eval not because my edge was bad, but because I’d never seriously looked at my worst stretch.
Here’s the reframe that fixed it: your strategy isn’t defined by its average. It’s defined by its drawdown. The average tells you what you make in a good year. The drawdown tells you whether you survive long enough to see one.
Two numbers matter more than your return:
- Max drawdown — the deepest peak-to-valley drop in the backtest. If it’s bigger than your prop firm’s limit, your strategy is mathematically incompatible with that account. The expectancy doesn’t matter; you’ll be halted before it plays out.
- Daily loss — the worst single day. Funded accounts die on daily loss limits more than anything else, usually right after a loss, when the bot (or the trader) presses.
So size for the drawdown. If your worst historical day is −$1,200 and your daily limit is −$1,000, you don’t have a strategy problem — you have a sizing problem. Cut the contracts until your worst day fits inside the rule with room to spare.
It feels backwards. You’re capping your upside on purpose. But getting funded isn’t about the best week — it’s about never having the week that ends you. Build for the valley, and the peaks take care of themselves.