Equity Curves That Mislead

An equity curve — cumulative P&L plotted over time — is the single chart everyone looks at first, and it's also the easiest chart to accidentally (or deliberately) make look better than the underlying strategy actually is. A smooth upward line invites trust, but the smoothness itself can come from things that have nothing to do with genuine, repeatable edge.

A few common ways a curve misleads: a single outsized winning trade can carry an otherwise mediocre or losing strategy, producing a curve that looks steadily profitable when in fact almost all the gains sit in one line of the trade log. Plotting on a linear rather than logarithmic scale makes early, small-dollar gains look flat and later, larger-dollar gains look explosive, even if the percentage returns were identical throughout — misleading anyone comparing performance across different account sizes or time periods. And a curve built on a strategy that was tuned against the very data it's plotted on will look better in-sample than any live version ever will, since the parameters were chosen specifically to make that historical curve rise.

Before trusting an equity curve, check the underlying trade-level P&L distribution for concentration in a handful of trades, confirm the scale (log for anything spanning more than roughly a 2x range), and separate in-sample from out-of-sample periods on the same chart rather than showing only the combined line.