Strategy Review
Two failures cost traders roughly the same amount of money: keeping a dead strategy alive out of loyalty, and killing a healthy one during a normal losing run. Both come from the same gap — no pre-agreed test for the difference.
First, is the drawdown even unusual?
Before interpreting anything, check whether the losing run exceeds what your win rate makes likely. A system winning 40 per cent of the time will produce a run of seven losses with uncomfortable regularity. If your current drawdown sits inside the expected range, there is nothing to diagnose.
| Win rate | Longest losing run in 200 trades (typical) | Extreme but plausible |
|---|---|---|
| 60% | 5 | 8 |
| 50% | 7 | 10 |
| 40% | 9 | 13 |
| 33% | 12 | 17 |
Second, the four-question test
- Q1Is my adherence rate over the drawdown above 85 per cent? If not, the trader is the variable.
- Q2Are the losses concentrated in one setup or spread across all of them?
- Q3Has the market condition my edge depends on measurably changed — volatility regime, session ranges, correlation structure?
- Q4Is the average loss bigger than it used to be, or just more frequent?
Q4 is the most diagnostic and the least used. More frequent losses of normal size usually means variance. Losses that are individually larger means your stop placement no longer matches the market volatility — a fixable calibration problem, not a dead edge.
Third, decide before you are in the hole
Write the retirement criteria for each strategy while it is working. Something concrete: “I stop trading this setup if it produces a negative expectancy over any rolling 60 trades with adherence above 85 per cent.” Then the decision during the drawdown is a lookup, not a debate with yourself at 2 a.m.
What a genuine regime break looks like
- The setup still triggers at the same frequency but the follow-through has gone.
- Losses come from the same failure mode repeatedly, not from varied causes.
- A peer trading the same idea reports the same experience.
- A structural explanation exists — a central bank regime change, a liquidity shift, a new participant in the flow.
Three of those four is enough to stop. One of them is not.
Retire a strategy on evidence, not on discomfort. Discomfort arrives every quarter; evidence does not.
Written by Adrian Vestberg — former interbank FX dealer, now running a two-person discretionary desk.