Marty Schwartz, the trader profiled in Jack Schwager’s Market Wizards, described a habit of cutting his position size sharply after a loss and trading small until he felt right again. It is one of the most quoted risk maxims in trading, and it is repeated as though it were a law. It is not a law. It is a hypothesis about your own behaviour, and it happens to be one you can test.
Why Reduce Position Size After a Loss?
The case rests on a claim about state, not about markets. It says that after a loss you are, briefly, a worse trader: more likely to force a setup, more likely to size by feeling, more prone to the revenge trade. If that is true, then the trades you take immediately after a loss carry lower expectancy than your average trade, and betting less on lower-expectancy trades is straightforwardly correct.
Notice how much rides on that “if”. The rule is not making a claim about price. Nothing about a loss changes the market’s next distribution; the market has no memory of your trade. The rule is making a claim about the trader. Cutting size after a loss is only rational if the post-loss version of you is measurably worse than the baseline version.
The Case Against
Here is where the maxim earns scrutiny. If your post-loss trades carry the same expectancy as your other trades, then cutting size after a loss systematically reduces your position in a positive-expectancy bet for no reason. Over hundreds of trades, that is a real cost paid for an emotional benefit.
Worse, there is a shape of this rule that damages trend traders specifically. Trend systems earn most of their return from a small number of enormous winners, the fat tails. Losing trades cluster right before big winners as often as anywhere else, because the losses are the cost of staying in the game until a trend arrives. Halving size after every loss means arriving at the year’s best trade with a fraction of your intended position. The rule can quietly amputate the tail it was never designed to touch.
The generic version also collides with something worth respecting: volatility is the fee, not the fine. Ordinary losses inside your system’s normal range are not signals. Reacting to them as if they were is precisely the behaviour that turns a system into a mood.
DO THIS
Do not adopt this rule. Test it. Split your trade log into two sets: trades taken immediately after a loss, and all others. Compare win rate and average R across the two. If your post-loss trades genuinely underperform, the rule is earning its keep in your account and you should implement it mechanically. If they perform identically, cutting size is costing you money to buy comfort.
Testing It Properly
Two refinements make the test honest. First, ensure both sets contain enough trades to mean anything; 30 in the post-loss set is a working minimum, and fewer than that is a coin flip with an opinion. Second, define “after a loss” precisely before you look: the next trade, the same-day trades, the trades within an hour. Different definitions capture different behaviours, and choosing the definition after seeing the results is how a trader proves whatever he already believed.
If the answer is that your post-loss trades are worse, you have learned something more valuable than a sizing rule. You have discovered that your state degrades measurably after a loss, which points toward every intervention in the Mind pillar: tilt profiles, a circuit breaker, a mandatory pause. Cutting size treats the symptom. The pause treats the cause, and the pause is usually cheaper, because it does not shrink your position in the good trades that follow.
The Deeper Lesson
Schwartz’s rule fits Schwartz’s psychology, and it may fit yours. The error is not in the rule. It is in the reflex that adopts a maxim because a great trader said it, which is borrowing a rule you never verified. Every risk maxim you inherit should be treated as a question addressed to your own record, and your record is the only witness qualified to answer.
On the Trader’s Roadmap, this sits at tier eight of the Money pillar, one of the tree’s contrarian nodes, requiring both the weekly loss limit and your win rate statistics. It is placed at the summit deliberately. It is not a beginner’s rule. It is an experiment, and you need a trade record before you are allowed to run it.
Run the test on your own data. The free Edge Companion journal tags trades in R, and the Trader’s Roadmap shows what has to come first.
The Complete Trader's Edge
The full Mind · Method · Money framework. 70 chapters.
View on Amazon →
Market Mayhem
400 years of bubbles, crashes, and the pattern that keeps repeating.
Buy on Amazon →
Greatest Companies
How the world's greatest companies were built — and what traders learn from them.
View on Amazon →
