Drawdown

Drawdown measures the drop from an equity peak to the next trough, in percent or currency, and how long the account takes to climb back to a new high. It's the one number that tells you whether you could have actually held the system without closing the platform and walking away. A strategy with an attractive average return and a drawdown nobody could stomach is not a strategy anyone finishes.

There are two ways to measure it, and they tell different stories. Closed-trade drawdown tracks the equity curve after each position is shut, trade by trade. It's honest because every number on it already happened. Open-position drawdown tracks the account's real-time value, including trades still sitting open and losing. For most systems the two roughly agree. For grid and DCA bots they diverge violently, and the gap is where the real risk hides.

The reason is structural. A grid or martingale system refuses to close a loser. It adds to the position instead, waiting for price to come back. So the trade log shows nothing but wins, closed one after another, while the unrealized loss on whatever is still open quietly balloons in the background. Read the trade history alone and the system looks flawless. Look at the account balance during the trend that broke it, and it's a different story entirely.

That's why we measure open drawdown, not the closed curve a vendor screenshots for a sales page. Reconstructing it means walking every bar with every open position marked to the current price, not just the price at the moment a trade finally closes. It's slower to compute. It's also the only version that matches what your own broker statement would show you at three in the morning, mid-trend, ladder still open.

The payoff is that a strategy stops being one number and becomes two. A win rate near 100 percent, paired with a drawdown that can end the account, are not contradictory findings. They're the same strategy, described from two angles, and the second one decides whether you're still trading it next year. Sometimes the damage isn't spread evenly either. A single tail-concentrated event can do more to an account than the whole drawdown history combined, which is why we look at both together instead of trusting either alone.