There is a pattern that shows up on every trading forum, every week. A trader posts a screenshot of their last fifteen or twenty trades, the curve is sloping down, and the question is always some version of the same thing: is this strategy still working, or is it time to switch? A handful of replies say "trust the process," a handful say "obviously broken," and the trader walks away no closer to an answer than they were before. So they switch. Three weeks later, they post the next chart with the next strategy, and the next question, and the loop begins again.
That loop is the single most expensive habit in retail trading. Not bad setups, not poor risk, not even tilt - it is the constant rotation away from anything before it has had a fair chance to prove itself, and toward whatever felt good in the last week. The reason it happens is simple: most traders have no objective way to tell a real failure from ordinary variance. So they default to feel, and feel is wrong almost every time. This article is about how to actually answer the question - how much data you need, what to look at, what numbers separate "in drawdown" from "broken," and the kill-switch you write down before any of it matters.
The one-line version: "is my strategy working" is a math question, not a feeling. You need a written expectancy, a written maximum acceptable drawdown, and a sample size large enough to mean anything. Without those three, every answer is just your last few trades talking - and your last few trades are noise.
The brain that tells you the strategy is broken after five losers is the same brain that told you it was a goldmine after five winners three weeks ago. Recency bias makes the most recent trades feel like the most representative, when statistically they are nothing of the sort. A profitable system with a 50 percent win rate produces strings of four or five consecutive losers as a matter of routine - the math demands it. To the trader living through one, it does not feel routine. It feels like the moment the edge died.
This is why retail traders chronically quit strategies that work and persist with strategies that do not. A losing streak inside the system's normal range gets read as a fatal break, so they bail. A winning streak inside the system's normal range gets read as confirmation, so they size up and stay - right until variance corrects on the upside. Both reactions come from the same broken instrument, which is the human eye scanning ten data points and trying to extract a verdict.
The single most important number in this whole conversation is sample size. Until you understand how few trades is "not enough," you will keep killing strategies prematurely. The rough guide is uncomfortable for most traders to hear.
If your read on the strategy is built on fewer than a hundred trades, you are not measuring the strategy. You are measuring a sample that is too small to separate skill from coincidence, and any conclusion you draw is essentially a guess wearing a chart. Worse, traders who switch strategies every twenty or thirty trades guarantee they will never accumulate a sample large enough to judge anything - they are trapped in noise forever, blaming the setups for what is really a sample-size problem.
The fix is not glamorous. It is to commit to a single, written strategy for long enough to accumulate a real sample before you allow yourself to render any verdict. That usually means months, not weeks. The discipline to wait is what separates the trader who eventually finds an edge from the trader who spends a decade rotating through strategies that probably had one.
Sample size buys you the right to ask the question. These three metrics answer it. Forget the cumulative P&L curve - it bounces around with luck. The real verdict lives in the underlying numbers, and any modern trading journal calculates them for you the moment trades sync.
Expectancy is the single most important number to know about a strategy. It tells you how much, on average, the strategy makes or loses per trade once win rate and risk-to-reward are blended together. Positive expectancy means that if you keep trading the same way, the math is on your side and your equity will grind up over a large enough sample. Negative expectancy means it does not matter how patient you are - you are losing money slowly on purpose. Our explainer on trading expectancy walks through the formula in full, but the core idea is that one positive-expectancy strategy traded consistently beats five neutral strategies traded enthusiastically every time.
Profit factor is gross profit divided by gross loss. A profit factor below 1.0 means the strategy loses money no matter how you frame it. Above 1.0 means it is profitable. Above roughly 1.3 means it is solid. Above 1.5 it is genuinely good. The reason it is more honest than P&L is that it is scale-free and position-size-free - you cannot flatter it by adding leverage or hide it by trading small. It is the strategy's underlying signal, stripped of the noise of how much you happened to risk on each trade. See our deeper write-up on profit factor for the ranges and what they actually mean.
This is where most traders find out their "broken" strategy is actually two strategies stitched together - one that works and one that does not. If you trade three setups, look at the expectancy of each one separately. It is extremely common for a perfectly good A-setup to be dragged into the red by a B-setup the trader keeps taking out of boredom or impatience. The aggregate looks broken; the breakdown shows it never was. This is the same diagnostic logic covered in our guide on the core trading performance metrics, applied specifically to the "should I quit" question.
Our free Excel journal auto-calculates win rate, profit factor, expectancy and equity curve. One row per trade and the three numbers above stop being a guess.
Get the free templateThis is the question every trader is really asking when they reach for a new strategy. The good news is there is an objective answer, and it depends on one comparison: how does this drawdown stack up against the drawdowns the strategy has historically produced?
Every profitable strategy goes through losing stretches. Even ones with strong long-term expectancy will give back ten to twenty percent of equity at some point - that is the price of admission for trading anything with variance. The question is not am I in a drawdown, it is am I in a drawdown that is consistent with how this strategy has always behaved, or am I in one that has now exceeded what it ever did before?
The rule of thumb: if your current drawdown is inside the range the strategy has produced before, you are in a normal losing stretch and quitting is almost certainly a mistake. If your current drawdown has clearly broken past the strategy's prior worst stretch - by say 50 percent more, or has lasted dramatically longer - that is the first real signal something underlying may have changed.
The reason this works is that markets do shift. Volatility regimes change, correlations rotate, a setup that thrived in a trending environment can stop working when range conditions arrive. But those shifts produce drawdowns that look unmistakably different from ordinary variance - deeper, longer, and unaccompanied by the normal recovery rallies. If you have never thought hard about the shape of normal drawdowns, our piece on what drawdown actually is covers why the recovery math gets brutally non-linear once the hole goes deep.
The reason traders quit at the wrong moment is that they make the decision in the middle of the pain. Tilt, frustration and recency bias all turn into "this is broken" right at the point where they should turn into "this is normal." The fix is to make the quit decision in advance, in writing, when your head is clear and the equity curve is flat - and then refuse to make a different decision in the heat of a drawdown.
A real kill-switch is concrete enough that you cannot argue with it later. It has numbers. These six points are what a working version of it looks like.
Pick a number - 100, 150, 200 trades - and commit to it. Until you reach it, the strategy is in evaluation, not on trial. No verdict, no switch, no exceptions. This single rule kills most of the strategy-hopping loop on its own.
Before you start, write down what a working version of this strategy should look like in numbers: target expectancy in R, target profit factor, target win rate at your RR. If you do not know what to expect, you cannot tell when reality has deviated from it.
Pick a percentage drawdown - usually 15 to 20 percent of account equity - that is your line. Above it, the strategy comes off the desk for review. Below it, no matter how bad it feels, you stay the course. The number is decided once, in calm, and never moved during a drawdown.
When you do hit a deep stretch, your reference point is not the last twenty trades. It is the strategy's previous worst drawdown over its full history. Worse than the prior worst by a clear margin is a signal. Inside the historical range is noise.
If something does feel off, the right first move is almost never to switch. It is to cut risk in half and keep trading the same setups. This lets you collect more data on whether the edge is genuinely gone without bleeding through the kill-switch level while you figure it out.
The moment you meaningfully change entry rules, stop placement, or filters, you are no longer trading the same strategy - you are trading a new one. The sample size resets to zero, and you go back to step one. A written trading plan makes this honest, because changes have to be logged instead of quietly absorbed.
Tracker Fx syncs your cTrader, Bybit, OANDA or MetaTrader trades and calculates expectancy, profit factor, win rate and drawdown per setup - so "is this strategy working" stops being a feeling and becomes a one-glance answer from your real fills.
Start Free TrialThe framing so far has been about the trader who quits too soon. The opposite mistake is rarer but real - the trader who clings to a strategy long past the point its numbers have stopped supporting it, usually because of identity or sunk cost. "I have been a breakout trader for two years" stops being a description and becomes a self-image worth defending, even when the expectancy chart has been negative for the last six months and the profit factor has slipped under 1.0 on a sample that long.
This is why the kill-switch has to be a real number, not a vibe. Without a written line, the trader who quits too soon abandons working systems on a feeling, and the trader who quits too late protects broken ones on a feeling. The numbers do not care which trader you are. They will tell you the same answer either way - if you let them. The framework for sorting that out is the same one we cover under what a trading edge actually is: an edge is a positive-expectancy statistical pattern that survives across a meaningful sample, not a recent run of winners.
"Is my strategy still working?" is the wrong question to ask after a losing week, because the answer at that resolution is almost always "I cannot tell yet." The right version of the question is harder and slower: over a sample large enough to mean anything, is my expectancy positive, is my profit factor above 1.0, and is my current drawdown inside the range this strategy has always produced? Three numbers, calmly read, beat a thousand gut decisions about whether to chase the next setup.
So fix the sample size in writing before you start. Track expectancy and profit factor per setup with a real trading journal so the verdict comes from your actual fills, not your memory of them. Write your kill-switch drawdown number now, while it costs nothing, and refuse to move it in the middle of a drawdown when moving it feels like the most reasonable thing in the world. Then trade the strategy long enough to mean something - and accept that the silent, repetitive months between switches are what an edge actually looks like in real life. The traders who get there are not the ones who picked the right strategy on the first try. They are the ones who stopped picking long enough to find out.
We use essential cookies to run the site and analytics cookies (Google Analytics) to understand what's working. You can accept or reject. See our Privacy Policy.
Tracker Fx Team