When to Retire a Trading Strategy That Stopped Working

In my second year running a shop, I killed a product that had paid the rent for eighteen months. Sales had slid for a quarter, then another, and I had a folder of reasons why the next quarter would be different. The weather. A road closure. A competitor's promotion. Every reason was true and none of them were the point.

What finally made the decision was not a feeling. It was a number I had written down a year earlier, in a quiet week, when I decided in advance what "this line is finished" would look like. When the number arrived, the decision had already been made by a calmer version of me.

A trading strategy is a product line, and every product line eventually comes up for review. Traders are terrible at this in both directions. Some fire a perfectly good strategy after three losing trades. Others defend one for a year past its expiry date, paying for the privilege every month. Both mistakes come from the same missing document: a written rule for when a strategy retires.

Everything below is how I run that review as an operator. It is education about running your own books, not instructions, and there are no prices, levels or signals here on purpose.

A Strategy Is a Product Line, Not an Identity

The first problem is not analytical, it is emotional, and it needs saying before any of the mechanics matter.

Most traders do not own a strategy, they are their strategy. It is the thing they explain to people, the thing they defend in group chats, the thing they spent months learning. Retiring it feels like admitting that the months were wasted, so the decision stops being a business decision and becomes a question about their competence.

A shop owner does not feel that way about a product. Discontinuing a line that stopped selling is Tuesday's admin, not a personal failure. The owner keeps the shop; the shelf gets restocked with something else. That distance is not coldness. It is what lets the decision get made at the right time instead of two quarters late.

And the distance is what keeps the business alive long enough to matter. US Bureau of Labor Statistics data on new private-sector establishments shows that of those that opened in the year to March 1994, 49.6 per cent were still trading five years later, and 33.6 per cent after ten years (BLS Business Employment Dynamics, Table 7). Roughly half survive five years, and the ones that do are not the ones that never had a bad product. They are the ones that stopped paying for it in time.

So the first move is to demote the strategy in your own head. It is a line item. It earns its shelf space or it loses it, and neither outcome says anything about you.

You Cannot Judge a Strategy on How the Last Few Trades Felt

Here is where most retirements are wrong: they happen after a run of losses that was always going to happen.

Any approach with a genuine edge still loses regularly. Losing runs are not evidence of failure, they are a feature of the arithmetic. And you can put a number on it.

Take a strategy that wins 55 per cent of the time, which would be a good, genuinely profitable business. Run one hundred trades through it. The chance that somewhere in those hundred trades you hit a run of five losses in a row is about 65 per cent. A run of four in a row is about 92 per cent likely. Even six straight losses shows up roughly one time in three. Drop the win rate to a coin flip and five in a row becomes an 81 per cent likelihood.

Those figures are not a survey or an opinion. I calculated them directly from the probabilities, assuming each trade is independent and the win rate is constant, which is the simplest honest model of a run of trades. Real trading is messier, and the messiness generally makes streaks worse rather than better, because losses cluster in the conditions that caused them.

Read that again, because it is the whole reason people fire good strategies. A losing streak that feels like proof your method is broken is, in a profitable business, closer to a near certainty. It will feel exactly like the strategy breaking. It feels that way to everyone.

Which means a run of losses on its own tells you almost nothing. What tells you something is a run of losses set against what your strategy's normal bad patch looks like. And you cannot know that unless you have recorded enough of them, which is why this whole subject collapses without a journal. If you do not have one, the trading journal template is the foundation everything here rests on.

The uncomfortable implication: you are not allowed to retire a strategy you have not measured. Not because the rule is unfair, but because without a record you are not making a decision, you are reacting to the last thing that happened to you.

Four Different Things That Look Identical

When results deteriorate, exactly one of four things is happening, and they demand completely different responses. Most traders never separate them, which is why they fix the wrong thing.

1. Variance

Nothing changed. You are inside a normal losing patch, and it looks like collapse because you are living through it one trade at a time. Correct response: nothing. Keep size where the plan says and let the sample finish.

2. Execution drift

The strategy is fine and you stopped following it. You widened a stop once and it worked. You started taking setups that were nearly right. You skipped the boring part of the process on a busy morning. The strategy did not stop working, you stopped running it. Correct response: fix the operator, not the product. This is the gap I wrote about in the gap between knowing and doing, and it is by far the most common answer of the four.

3. Changed conditions

The strategy needed something the market is no longer offering. A range approach in a market that now trends, a breakout approach in a market that keeps failing at the edges, an approach built for wide daily movement in a quiet month. Nothing is broken, the environment simply moved out from under it. Correct response: reduce or pause, note the conditions it needs, and revisit when they return. This is where genuine retirement or hibernation belongs.

4. It never worked

There was never evidence, only a good explanation and a promising first week. Correct response: retire it without ceremony. There is nothing to mourn, because there was never anything there.

Notice that only two of these four call for retirement, and neither of them is the one that hurts most in the moment.

Write the Retirement Rule Before You Need It

Every business I respect makes discontinuation decisions against criteria set in advance, when nobody is panicking and nobody is defending anything. You can do exactly the same on one line of your plan.

Three components, all of them decided in a calm week:

• A minimum sample. A number of trades below which you will not judge anything at all. This single line stops most premature retirements, because it makes "three bad days" an inadmissible argument.
• A maximum drawdown. A loss level for this strategy at which it stops trading regardless of how convinced you are. Not a level at which you think about it, a level at which it stops. This is the line that stops the second failure mode, the year of one-more-month.
• A review date. A fixed point in the calendar when the line comes up for assessment whether or not it is in trouble. Reviews that only happen during pain get made during pain.

Write those three next to each strategy on your one-page trading business plan. The value is not in the numbers being clever. It is that the decision gets made by the version of you that is not currently down money.

Demote Before You Retire

This is the step almost nobody takes, and it is the most useful thing in this article.

A shop does not usually go from full shelf to gone. It cuts the order, gives the line less space, and watches for a period before clearing it out entirely. You can do the same, and it turns a binary decision into a reversible one.

Demotion looks like this: cut the strategy to your smallest workable size, keep following it exactly as written, and keep recording. You are no longer risking meaningful money on something you have doubts about, but you are still collecting evidence rather than throwing the sample away. If it recovers, you have proof rather than a hunch, and you scale it back up deliberately, the way I described in when to scale up your trading business. If it keeps deteriorating at minimum size, retirement becomes an easy call made on a record instead of an argument.

Demotion also solves the emotional problem. Nobody has to be wrong. You are not admitting failure, you are reducing exposure to uncertainty, which is the most ordinary business decision there is.

Firing a strategy after a bad week and defending one for a year are the same mistake. Both replace evidence with feeling. The only difference is which feeling won.

The Hidden Cost of Strategy Hopping

There is a reason I push demotion rather than replacement, and it is a cost that never appears in anyone's accounts.

Every time you swap strategies, you reset the clock. Your sample starts from zero, so you cannot tell for months whether the new approach is any good. The learning curve restarts too, which means your execution gets worse for a while regardless of how sound the method is. And crucially, the switch usually happens at the worst possible moment: right after a losing run, when your confidence is lowest and your judgement is at its least reliable.

The pattern that follows is familiar to anyone who has run a chaotic business. New approach, hopeful start, first losing patch arrives, and because there is no sample and no rule, it gets fired too. Eighteen months of that and a trader has three years of experience in nothing.

An operator running two or three lines properly beats one running twelve badly, in trading exactly as in retail. The metrics that let you tell them apart are in the key metrics every trading business should track, and they only mean something if you stay in one place long enough to gather them.

What to Do With the Capital You Free Up

When a line is retired, its shelf space does not have to be filled the same week.

The instinct is to deploy that capital immediately, usually into whatever looked good while you were losing. Resist it. There is nothing wrong with a business holding cash while it decides, and a strategy retired properly is often followed by a quieter month by design. That is not lost time. It is the period where you are reading conditions rather than paying for opinions, which is the whole argument in what a business owner does when the market goes quiet.

If the freed capital goes anywhere first, the reserve is a better home than a new experiment. A business with a buffer makes better decisions than one that needs the next month to work, and that logic is laid out in how to build a cash reserve for your trading business.

The Review That Makes All of This Routine

None of this needs to be a dramatic event. It fits inside the reviews you should already be running.

Your weekly review is where execution drift gets caught, which handles the most common of the four causes before it ever looks like a strategy problem. Your monthly profit and loss review is where the line gets held against its retirement criteria, and where you write one sentence on whether the conditions it needs are still present.

Do those two consistently and the retirement decision stops being a crisis. It becomes an entry in a meeting you were having anyway, which is exactly what it should be.

What I Would Want You to Take Away

Write the retirement rule while things are calm. Minimum sample, maximum drawdown, review date. One line per strategy. Ten minutes of work that will one day save you from an expensive decision made at the worst moment.

Diagnose before you decide. Variance, your own execution, changed conditions, or never worked. Four different answers, and only two of them mean retirement.

Demote instead of firing. Minimum size, same rules, keep recording. Reversible beats final.

Do not swap in a replacement out of frustration. A retired line and an empty shelf is a fine position for a business to be in for a few weeks.

And the one that took me longest to accept: the strategy is not you. The business is you. Products come and go, and the operator who can retire one without flinching is the one still trading in five years.

Frequently Asked Questions

How do I know if my trading strategy has stopped working? You cannot know from results alone, and certainly not from a handful of trades. You need enough recorded trades to know what a normal losing patch looks like for that strategy, then you check whether current results fall outside it. Before concluding anything, rule out the two more likely explanations: ordinary variance, and your own execution drifting away from the written rules.

How many losing trades in a row mean a strategy is broken? There is no universal number, and any specific figure someone quotes you is invented. Losing runs of several trades occur naturally in approaches with a real edge. This is exactly why the minimum sample and maximum drawdown belong in writing beforehand, so the answer is your own recorded number rather than a feeling on a bad afternoon.

Should I change my strategy after a losing month? Usually not, and a losing month is the worst possible moment to make the decision. Diagnose first. If the month broke your written retirement criteria, act on the criteria. If it did not, the month was data and not a verdict.

What is the difference between a strategy that stopped working and one that never worked? Evidence. A strategy that stopped working has a recorded period where it performed as intended under conditions you can describe. One that never worked has an explanation, a promising first week and no record. The second needs no ceremony to retire.

Can a retired strategy be brought back? Yes, and this is why I prefer the word retired to dead. If a strategy needed conditions the market stopped offering, it may be worth running again at small size when those conditions return. Write down what it needed at the moment you shelve it, because in six months you will not remember accurately.

How many strategies should a trading business run? Fewer than most people think. Each one needs its own sample, its own metrics and its own review to be judged at all, and attention divides faster than results improve. Two or three run properly is a business. A dozen run loosely is a hobby with a spreadsheet.

About Rex

I'm Rex. Before I ever placed a trade I spent five years running a real business, and the habit that transferred best was unglamorous: decide what "finished" looks like before you are emotionally involved. I killed a product line once on a number I had written down twelve months earlier, and it was the easiest hard decision I ever made, precisely because the deciding was already done. Traders rarely give themselves that gift, which is why so many of them fire the good approaches and marry the bad ones.

Today I run the REX Trading Signal channel, around 11,900 people, on three rules I don't break: every setup carries a stop loss; I post the losing trades, not only the winners; and I never promise profit, no "guaranteed," no "fixed," no "risk-free." Nothing here is financial advice, and no entry, stop or target discussed should be treated as a signal. An operator who can retire a strategy calmly is running a business. Everyone else is defending a purchase.