Skip to content
Technical Analysis

Risk of ruin: the arithmetic that decides if you survive

A positive-expectancy system can still destroy an account. The variable that decides it is size — and the relationship is far less forgiving than it looks.

Technical AnalysisAdvanced12 min read
Browse Technical Analysis(172)

Everything in this track has been about finding an edge. This lesson is about the thing that decides whether the edge ever gets to pay you: whether you are still trading when it does.

Think of it like this
The favourable game you can still lose

A game pays ₹110 for every ₹100 wagered, on average. Excellent odds. Bet your entire savings on each round and one bad run ends you permanently — before the average ever arrives.

In the market

That is risk of ruin. The edge is real, the expectancy is positive, and none of it matters if a normal losing streak takes the capital away first. Position size is the only variable connecting the two.

Losing streaks are longer than intuition suggests

A system that wins 45% of the time — a perfectly good trend-following hit rate — will produce long strings of losses purely by chance. Not because anything is wrong. Because that is what randomness looks like.

Win rateChance of 6 losses in a rowChance of 10 in a rowOver 200 trades, expect a streak of about
60%0.4%0.01%5–6
50%1.6%0.1%7–8
45%2.8%0.25%8–9
35%7.5%1.3%11–12
A 35% win-rate system with large winners is entirely viable — but it will hand you a run of eleven losses, and you must have planned for that in advance.
Loading interactive demo…

The asymmetry that makes size the deciding variable: losses and the gains needed to undo them are not symmetric, and the gap widens sharply beyond about 30%.

Loading interactive demo…

Now watch it happen: two thousand simulated traders run the same system, and only the risk per trade decides how many reach the drawdown they would call ruin.

Why bigger is not better

The intuition that a larger bet on a favourable edge produces proportionally more money is wrong, and it is wrong in a specific way: returns compound multiplicatively, so a large loss removes the base that future gains would have grown from.

Worked example
The same system at three position sizes
45% win rate, 2R winners, 100 trades
Risking 1% per tradeWorst drawdown roughly 12%. Comfortable throughoutEnds around +35%
Risking 3% per tradeWorst drawdown roughly 33%. Genuinely difficult to sit throughEnds around +70%
Risking 6% per tradeWorst drawdown exceeds 60%. Several sequences never recoverOften ends negative
Risking 10% per tradeA single ordinary streak removes most of the capitalRuin is the likely outcome
Return does not rise with size indefinitely — it peaks and then collapses. Past a certain point, additional size makes the system worse in expectation, not merely more volatile. This is the most counter-intuitive result in trading and the most expensive one to learn from experience.

Kelly, and why nobody trades it

The Kelly criterion computes the size that maximises long-run growth. It is mathematically correct and almost universally reduced in practice.

f = W − (1 − W) ÷ R
f
fraction of capital to risk
W
win rate as a decimal
R
average win divided by average loss

Example: W = 0.45, R = 2 → f = 0.45 − 0.55 ÷ 2 = 0.175, or 17.5% of capital per trade. Almost nobody should trade this.

Practical sizing rules
  1. 1
    Cap risk per trade at 1–2%

    Not because it is optimal, but because it makes every plausible losing streak survivable while you find out whether your edge is real.

  2. 2
    Cap total open risk

    Six positions at 1% each is 6% at risk if correlated — and in a market fall they will be correlated. Limit combined open risk to roughly 5–6%.

  3. 3
    Reduce size in drawdown

    Trading smaller after a 10% drawdown slows recovery slightly and makes ruin far less likely. Increasing size to recover does the opposite of both.

  4. 4
    Size on measured performance, not conviction

    Felt confidence has no relationship to outcome. Size from your recorded win rate and payoff ratio.

Check yourself

A system wins 40% with 2.5R winners — a solidly positive expectancy. The trader risks 8% per trade. What is the likely outcome?

Simple bhasha mein
Jeetne wala khel bhi hara sakta hai

Ek khel mein har ₹100 pe ₹110 milte hain — faayde ka sauda hai. Par agar aap har baar poori jama-poonji lagate ho, toh ek bura din aur khel khatam — average aane se pehle hi aap bahar. Isiliye system achha hona kaafi nahi; size chhota rakhna hi asli bachaav hai.

What to remember
  • A positive-expectancy system can still ruin an account; size decides which.
  • Losing streaks are longer than intuition suggests — plan for the streak your win rate implies.
  • Returns do not rise indefinitely with size: past a point, more size lowers expected outcomes.
  • Full Kelly assumes perfect knowledge of your edge and is far too aggressive in practice.
  • Cap risk per trade, cap total open risk, and reduce size in drawdowns rather than increasing it.
You reached the endMark it done and keep your streak going.
Up nextWhen six positions are really one betPrevious: Scaling in, pyramiding and partial exits
Finished this lesson?

Mark it done to track your progress through the curriculum.

Common questions

Short, direct answers to what people ask about this topic.

risk of ruin meaning in trading
Risk of ruin is the probability that an ordinary run of losses takes your capital below the level at which you can keep trading, even though the system itself has a positive edge. Edge and survival are separate questions: expectancy describes what happens on average over many trades, while position size decides whether you are still there when that average arrives. A game paying Rs 110 for every Rs 100 wagered still finishes you if you stake everything on each round.
how much capital should be risked per trade to avoid ruin
The convention taught in risk management is a cap of 1–2% of capital per trade, with combined open risk across all positions held to roughly 5–6%, because positions correlate in a falling market. The reasoning is survivability rather than optimality: eleven consecutive losses at 1% each is about a 10% drawdown, while the same streak at 5% each is roughly a 43% drawdown needing a 75% gain to undo. That is the arithmetic of the convention, not a recommendation for any particular account.
the formula that calculates the bet size maximising long-run growth is called
The Kelly criterion. In trading form it is written f = W − (1 − W) ÷ R, where W is the win rate as a decimal and R is the average win divided by the average loss. It is mathematically correct and almost universally scaled down in practice, because it assumes you know W and R exactly — a system with W = 0.45 and R = 2 returns f = 17.5% of capital per trade, a size whose drawdowns routinely exceed 50%. Many professionals work at a quarter to a half of Kelly, and plenty use less.
what gain is needed to recover from a 50% drawdown
A 100% gain — you have to double what is left just to get back to where you started. The relationship is not symmetric and it worsens sharply as losses deepen: 20% down needs 25% back, 33% down needs about 50%, 50% down needs 100%, and 60% down needs 150%. That asymmetry is why position size, rather than win rate, is the variable that decides whether an account survives.
how many losses in a row are normal for a 45% win rate system
Across roughly 200 trades, a system winning 45% of the time will typically produce a run of eight or nine consecutive losses purely by chance, with nothing wrong at all. Streaks lengthen as the win rate falls — a 35% win-rate system, which stays perfectly viable if its winners are large, should expect a run of eleven or twelve. The planning question is never whether the streak arrives but what your position size turns it into when it does.