A smooth equity curve in a backtest is a warning, not a prize
Posted: Sat Oct 31, 2026 6:57 am
Equity curves that rise in a neat line with tiny drawdowns look like a dream, and they usually signal a problem.
Common causes include overfitting, ignored costs, unrealistic fills, or hidden risk such as martingale or grid logic. A system that adds to losers can produce a smooth curve for a long time, then wipe out everything in one event.
Checking the distribution of trade results helps. A strategy with a high win rate and a small average win next to rare, enormous losses is concealing tail risk.
The longest losing streak, the largest single loss, and the ratio of worst loss to average win are all useful numbers. If one loss equals fifty wins, the curve is a time bomb.
Real edges tend to have visible bumps. Accepting that drawdowns exist, and sizing for them, is healthier than chasing a curve that looks too perfect.
Have suspiciously smooth results ever turned out to hide a structural risk?
Common causes include overfitting, ignored costs, unrealistic fills, or hidden risk such as martingale or grid logic. A system that adds to losers can produce a smooth curve for a long time, then wipe out everything in one event.
Checking the distribution of trade results helps. A strategy with a high win rate and a small average win next to rare, enormous losses is concealing tail risk.
The longest losing streak, the largest single loss, and the ratio of worst loss to average win are all useful numbers. If one loss equals fifty wins, the curve is a time bomb.
Real edges tend to have visible bumps. Accepting that drawdowns exist, and sizing for them, is healthier than chasing a curve that looks too perfect.
Have suspiciously smooth results ever turned out to hide a structural risk?