Telling a broken edge from a bad month
Every edge eventually stops working, and almost every edge occasionally looks like it has when it has not. Distinguishing the two is genuinely hard, and there is no test that settles it — but there is a way to be wrong less often, and it starts by refusing to answer the question from the equity curve alone.
The equity curve is the last place a real change shows up and the first place noise shows up. A more useful set of questions sits underneath it. Has the win rate moved beyond its historical range, or only within it? Has the average win shrunk, the average loss grown, or both? Is the number of signals per month still normal? Did your own behaviour change — larger size, more trades, later entries — before the results did?
That last question catches the most common case by far. Far more edges are broken by the trader than by the market: the setup was tightened after a losing run, or loosened after a winning one, or the sizing crept up, or the trades were taken twenty minutes later than the rule specified. None of that appears as a strategy failure, and all of it is visible in a journal that recorded the plan alongside the trade.
When the market really has changed, it usually shows up in the mechanics rather than the money: the signal fires far less often, or fires as often but the distance it used to travel has collapsed. Volatility regimes do exactly this — a mean-reversion method that needs a certain daily range simply stops having anything to work with, and the honest description is "the conditions are gone", not "the edge is gone".
The practical procedure is unexciting and works. Before concluding anything, check that you followed the rules, check the sample is large enough to say anything at all, check whether the signal count and the average move changed, and only then look at the money. If the first three are unchanged and the money is not, you have a genuine result. If they changed, you have your explanation and it is not the market.
What to take away
- The equity curve is the last place a real change appears and the first place noise does.
- Most broken edges were broken by the trader, and the journal shows it.
- A real regime change usually shows in signal count or average move, not first in P&L.
Where it goes wrong
- Judging an edge from the curve without checking rule adherence first.
- Concluding "the edge is gone" when the conditions it needs are simply absent.
- Changing the method and the size at the same time, so neither can be evaluated.
This chapter, measured against your own trades
In the app the same chapter ends in your figures rather than an example: how often you did the thing it describes, over your last ninety days. You pick one change to make, and Evidence checks afterwards whether it actually changed — from your journal, arithmetic, no opinion involved. Questions you get wrong come back a week later and again a month after that.
Open the free plan →Check that it stuck
Answers shown — in the app these are asked before you see them, and the ones you get wrong come back after a week.