How to Know When Your Trading Strategy Is Dying (Before It Kills Your Account)

Every trader has been here. Six weeks ago the system was humming — entries clean, losses cutting fast, winners running. Now it's four consecutive red weeks and the setup that used to be obvious looks wrong every time you take it. The journal says the P&L is down 18R from the peak. And yet you keep taking the trades, because it worked before, and because you spent nine months building it, and because you're sure the market is just being difficult.

That last sentence is where accounts go to die. The question of whether your strategy is in a normal drawdown or actively decaying is one of the most important questions in systematic trading — and almost nobody answers it with data. They answer it with hope.

This article gives you the four signals that separate temporary variance from structural decay, a framework for measuring them, and the decision protocol for what to do when the signals start flashing.

Normal Drawdown vs. Real Decay

The fundamental distinction is whether the underlying edge still exists. A drawdown is a period of losses inside a system that is still structurally sound. Every profitable strategy ever built has drawdowns. A trend-following system with 40% win rate and 0.5R expectancy will produce 8-trade losing streaks with some regularity just from normal variance. That's not decay. That's the cost of admission.

Decay is different. It means the statistical edge that produced your positive expectancy historically has diminished or disappeared. Markets repriced the setup. Volatility regimes changed and your stop distances no longer work. The correlation you were exploiting was arbitraged away. Whatever the cause, the mean R-multiple and expectancy of the system are now lower than when you built confidence in it — and the loss might be permanent, not cyclical.

The table below captures the practical difference between the two:

The hard part is that both scenarios feel identical from the inside. Six consecutive losers in a perfectly healthy system and six consecutive losers in a dying one produce the same emotional state. The difference is entirely in the data — which is why most traders who quit during a normal drawdown or keep trading through decay make the same mistake: they're making the call on feeling, not measurement.

The Four Signals of Strategy Decay

1. Rolling Win Rate Trending Down

Not one bad month — a sustained trend. Calculate your win rate over your most recent 30 closed trades and compare it to your full historical average. Now do the same for the 30 trades before that. If the trend line is pointing down across two or more consecutive windows, and the gap from your historical baseline is greater than 10–15 percentage points, that's signal, not noise. One bad window is variance. A directional trend over 60+ trades is structural.

2. SQN Falling Below Its Floor

Every healthy system has a normal SQN range — the band it operates in during normal market conditions. If your system historically ran between 2.5 and 3.2 and your rolling 50-trade SQN is now 1.4 and falling, something has broken. The SQN is the single most diagnostic number here because it captures both the magnitude of your edge (mean R) and its consistency (standard deviation of R) simultaneously. A win rate decline could be explained by a bad setup cluster. An SQN decline that deep means both average performance and consistency have degraded. That's the signature of decay, not variance. Read our full breakdown of SQN benchmarks if you're not yet calculating it from your journal.

3. Losses Running Larger Than Designed

This one is underappreciated. If you designed your system around -1R average losses and your recent 30-trade average loss is -1.5R, that's a 50% bleed on every losing trade above what the system modeled. This isn't a market problem — it's a signal that either volatility has increased enough to push price through your stop before you can exit, or you're giving trades "a little more room" under pressure. Either way, your risk model is now wrong. A system with -1.5R average losses instead of -1.0R needs a higher win rate and larger winners just to break even, let alone remain profitable. If the winners haven't grown to compensate, the expectancy math has flipped.

4. Winners Shrinking in R-Multiple Terms

Your system used to produce average winners of +2.2R. Now they're averaging +1.4R. The losers are the same but the winners are smaller — meaning market conditions have changed such that your targets are no longer being reached before reversal. This is often the earliest sign of regime change. When the market that your strategy was calibrated for (trending, volatile, range-bound) shifts, the winners compress first. Losses stay at -1R because your stop discipline holds. But the move from entry to target that used to reliably complete now reverses at 60% of the way there. Watch the average winner more closely than the average loser.

The Two Mistakes That Destroy Accounts

There are two ways to get this wrong, and they're mirror images of each other.

Mistake 1: Quitting during a normal drawdown. Five consecutive losers inside a system with SQN 2.8, positive expectancy, and clean -1R losses. Every metric is green. The system is working as designed — you're just experiencing a losing streak that the math predicts will happen multiple times per year. The trader who abandons here locks in the loss and then watches the strategy recover without them. This is the single most common way retail traders fail to compound: they keep switching systems at the worst possible moment, right at the bottom of a drawdown.

Mistake 2: Staying in a dying system. Six months of declining SQN, rolling win rate 20 percentage points below baseline, average losses growing, winners shrinking. The account is down 40R from peak. The trader keeps trading at full size because "it'll come back." It doesn't. The edge is gone. Every additional trade at full size is lighting money on fire. The inability to distinguish this from Mistake 1 is what turns a 20R drawdown into an account that needs to be rebuilt from scratch.

"The most expensive thing in trading isn't a bad trade. It's a dead strategy you kept feeding capital to because it used to work."

The Decision Protocol: What to Do When the Signals Appear

When you start seeing decay signals, there's a five-step process that gives you the data to make a real decision instead of an emotional one:

Why You Need 30+ Trades Before You Decide Anything

This is the most underappreciated constraint in all of retail trading. With 10 trades, you cannot statistically distinguish a dying strategy from a healthy one in a bad week. The noise swamps the signal. A system with a true win rate of 55% will produce a 10-trade sample with a win rate below 30% roughly 5% of the time — just from variance. That's one in twenty 10-trade windows, and it would look identical to a system that has genuinely decayed to 30% win rate.

The minimum thresholds for each signal:

The practical implication: if you trade 3 times a week and your rolling window is 30 trades, that's 10 weeks of data per measurement. You can't make a confident decay call in 2 weeks. This is uncomfortable — 10 weeks of pain before a definitive answer — which is why the position size cut in Step 3 matters so much. You're buying time to collect the data without destroying the account in the meantime.

How SignalDeck Surfaces Decay Automatically

The metrics above — rolling win rate, SQN trajectory, average R by outcome — are exactly what SignalDeck was built to compute automatically from your trade journal. The Strategy Decay Chart shows a rolling 20-trade win rate line against a 50% reference, making the trend direction immediately visible. The Realized R-Multiple Over Time chart shows each closed trade's R in sequence with a rolling 5-trade average line, so you can see at a glance whether winners are compressing. The SQN widget updates with every new closed trade and can be filtered to a single strategy.

The goal isn't to automate the decision — it's to make sure you're confronting the data instead of avoiding it. Most traders in a drawdown stop looking at their journal. The charts feel accusatory. But the data is the only thing standing between you and making the same mistake in both directions. Track your edge in real time — free during beta.