Skip to main content
The Risk-of-Ruin Equation

Why Positive Expectancy Is Not Enough

Pomegra Learn

Why Do Traders with Winning Systems Still Go Broke?

Positive expectancy trading is often cited as the Holy Grail: if your average profit per trade is larger than your average loss, you have an edge, and over hundreds of trades, that edge should compound into wealth. Yet history is littered with traders who had positive expectancy systems and still faced ruin. The answer lies in a brutal fact: expectancy is only one dimension of risk. A strategy with a +$100 average edge per trade can still bankrupt you if that edge comes surrounded by catastrophic drawdowns, correlated losses, or position sizes mismatched to your bankroll.

Understanding why positive expectancy trading is insufficient forces you to confront the full distribution of outcomes—not just the average—and to recognize that ruin depends more on how your wins and losses are ordered and how much capital you risk per trade than on whether your expected value is positive or negative.

Quick definition: Positive expectancy means your average profit per trade, weighted by win probability, exceeds your average loss per trade—yet ruin can still occur if a sequence of losses depletes your account before the edge manifests.

Key takeaways​

  • A winning system (positive expectancy) is necessary but not sufficient for survival; you also need adequate capital and position sizing
  • Variance (the scatter of outcomes around the average) determines how severe drawdowns are and how long it takes expectancy to "kick in"
  • High-variance systems require much larger bankrolls than low-variance systems with similar expected value
  • Sequential risk—the order of wins and losses—matters more than average expectancy for determining ruin probability in the short to medium term
  • Many traders fail with positive expectancy systems because they size positions as though their expected value were guaranteed

The Paradox: Positive Expectancy Yet Ruin​

Imagine two trading systems:

System A: Win rate 90%, average winner $100, average loser $900. Expected value: (0.90 × $100) − (0.10 × $900) = $90 − $90 = $0. This system is break-even.

System B: Win rate 51%, average winner $200, average loser $190. Expected value: (0.51 × $200) − (0.49 × $190) = $102 − $93.10 = +$8.90. This system has positive expectancy.

By pure expected value, System B is superior. But imagine a trader starting with $1,000, risking $500 per trade on System A. They win the first nine trades (+$4,500 gain, new balance $5,500). On the tenth trade, they lose ($500 loss, balance $5,000). Over ten trades, they're exactly break-even in expectation, but they've accumulated significant capital. Ruin was never at risk.

Now the same trader runs System B with $1,000, risking $500 per trade. The win rate is 51%, so roughly half the time they lose $190 and half the time they win $200. After 50 trades, they expect about 25 wins and 25 losses, for a net gain of 25×200 − 25×190 = $5,000 − $4,750 = $250. Positive expectancy manifests as planned.

But here's the problem: the trader might experience a different sequence. Suppose the first 20 trades are all losses: $500 per loss × 20 = $10,000 in losses. Their starting capital was only $1,000. They're bankrupt on trade 3 or 4, long before the law of large numbers can rescue them. The positive expectancy is mathematically certain to win in the long run, but the trader is already ruined.

This is the core paradox: positive expectancy trading promises ultimate victory but offers no guarantee of short-term survival.

The Role of Variance in Ruin Risk​

Variance is the villain in this story. Variance measures how spread out your returns are around the average. Two systems can have identical expected value but vastly different variances.

Compare:

Low-Variance System: Wins are $1,000 and losses are $950. Win rate 52%. Expected value = (0.52 × 1000) − (0.48 × 950) = 520 − 456 = +$64 per trade. Variance is small; outcomes cluster tightly around the average.

High-Variance System: Wins are $10,000 and losses are $9,500. Win rate 52%. Expected value = (0.52 × 10000) − (0.48 × 9500) = 5200 − 4560 = +$640 per trade. Variance is large; outcomes are widely scattered.

The high-variance system has ten times the expectancy, which sounds great. But it also has ten times the volatility. If you risk the same number of dollars per trade on both, the high-variance system will force drawdowns ten times deeper, even though its edge is larger.

Here's the ruin implication: for a given starting capital and position size, high-variance systems face much higher ruin risk than low-variance systems with similar or even lower expected value per trade.

How Sequence Order Overrides Average Expectancy​

Monte carlo simulations reveal an uncomfortable truth: the order of your wins and losses often matters more than the average outcome. Two traders might have identical 51% win rate and identical average profits, yet face vastly different ruin probabilities depending on whether their winning systems cluster all wins together or scatter them evenly.

Consider a simple example. A trader has a system with 51% win rate, +$100 per win, and −$100 per loss. Expected value = 0.51 × 100 − 0.49 × 100 = $1 per trade.

Scenario 1 (Clustered wins): The sequence is W, W, W, W, W, L, L, L, L, L, W, W, W, W, W, ... After the first five wins, the trader is +$500. After five losses, they're at break-even. But they had a $500 cushion during the loss cluster, so ruin never threatened.

Scenario 2 (Alternating): W, L, W, L, W, L, ... The trader gains and loses $100 each turn, staying near flat. With a $1,000 starting balance, ruin is not at risk.

Scenario 3 (Clustered losses): L, L, L, L, L, W, W, W, W, W, ... With a $500 starting balance and −$100 per loss, the first loss leaves $400. The second leaves $300. By the third loss, the trader has only $200. By the fourth, $100. By the fifth, $0. Ruin on trade 5, even though the system has positive expectancy.

All three sequences are statistically plausible given a 51% win rate. Yet only Scenario 3 leads to ruin, despite the identical expected value.

Why High Win Rate Doesn't Guarantee Safety​

Traders often assume that a 90% win rate system is safer than a 50% win rate system. That's intuitive but dangerously wrong if the trade sizes differ.

Consider:

System 1: Win rate 90%, average winner $50, average loser $900. Expected value: (0.90 × 50) − (0.10 × 900) = 45 − 90 = −$45. Negative expectancy, system is a loser.

System 2: Win rate 55%, average winner $500, average loser $400. Expected value: (0.55 × 500) − (0.45 × 400) = 275 − 180 = +$95. Positive expectancy, system is a winner.

System 1 has a much higher win rate (90% vs. 55%), yet it's a pure money loser. A trader with System 1 will fail gradually over time, despite winning 9 out of every 10 trades. A trader with System 2 will eventually grow wealth despite losing nearly half their trades.

The lesson: win rate in isolation is meaningless. Only the ratio of average winner to average loser—combined with win probability—determines positive expectancy.

The Risk-Reward Ratio Trap​

Even traders aware of expectancy often misunderstand the risk-reward ratio's role in ruin. A "3:1 risk-reward" system (risk $100 to make $300) sounds attractive, but it doesn't tell you the win rate.

A system with 3:1 risk-reward and only a 20% win rate is a catastrophic loser: Expected value = (0.20 × 300) − (0.80 × 100) = 60 − 80 = −$20. You lose money on average.

A system with 1:1 risk-reward (risk $100 to make $100) and a 55% win rate is a strong winner: Expected value = (0.55 × 100) − (0.45 × 100) = 55 − 45 = +$10. You win money on average.

Yet many traders treat "risk-reward ratio" as the primary metric, forgetting that the ratio only matters when multiplied by win probability. An excellent risk-reward ratio combined with a terrible win rate is still a loser.

The Bankroll-to-Position-Size Gap​

Here's where positive expectancy trading meets reality: many traders correctly calculate their edge but size positions as if that edge guarantees short-term survival.

Suppose a trader calculates:

  • Expectancy: +$50 per trade
  • Typical win: +$200
  • Typical loss: −$150
  • Win rate: 55%

Starting capital: $5,000. The trader calculates that with a +$50 edge and 100 trades, they'll make $5,000 in profit, doubling their capital. So they risk $500 per trade to accelerate the process.

But risk-of-ruin math tells a different story. With these parameters and a $500 risk per trade on a $5,000 account, ruin probability is approximately 15–20% (depending on exact variance). The positive expectancy is real, but the position size is reckless given the starting capital. The trader faces a 1-in-5 to 1-in-7 chance of wiping out before the edge manifests.

A safer approach: risk 2% per trade ($100 on the $5,000 account). This dramatically reduces ruin probability to under 1%, and the trader still compounds wealth, just more slowly.

Decision Tree for Positive Expectancy Traders​

Real-World Example: A System with +$50 Edge and 25% Ruin​

A currency trader has a system with these statistics:

  • Expectancy: +$50 per trade
  • Win rate: 54%
  • Average winner: $180
  • Average loser: −$160
  • Historical data: 250 trades
  • Starting capital: $10,000
  • Position size: $500 per trade (5% risk per trade)

Risk-of-ruin calculation (using formula from earlier chapters): Ruin probability ≈ 25%

The trader is shocked. The system has positive expectancy—they've calculated it carefully. Yet there's a 1-in-4 chance they'll be bankrupt before accumulating significant profits.

They run a monte carlo simulation (as described in prior articles) and confirm: in 250 out of 10,000 trials, the account hits zero before 100 trades have been executed.

The decision: reduce position size to $250 per trade (2.5% risk per trade). New ruin probability ≈ 0.8%. This is acceptable. The trader's edge will compound more slowly, but ruin becomes a non-issue across any reasonable time horizon.

Alternatively, they could maintain $500 per trade but increase starting capital to $30,000. Now ruin probability drops to under 3%, and the absolute dollar growth is still strong.

The lesson: positive expectancy is only a starting point. Risk management—driven by bankroll, variance, and position sizing—determines whether you live to see that edge compound.

Why Disciplined Traders Still Fail with Positive Expectancy​

Even traders who follow strict position sizing rules sometimes underestimate the role of correlation and regime shifts. A system might have 52% win rate in the data used to calculate expectancy, but if the next 100 trades occur in a regime with 48% win rate (a plausible drawdown in a bull market turning to bear), the positive expectancy evaporates or reverses.

Similarly, if your trading system wins by "fading reversals" but the market suddenly enters a strong trend, your winning assumption (reversals happen 52% of the time) breaks down. Now you're facing worse odds and larger losses than your historical data suggested.

The remedy is conservative position sizing with an even larger buffer. Many professional traders size positions assuming worst-case regime shifts, not best-case historical averages. This means they risk less per trade than the raw math suggests, but they survive market regime changes that would wipe out competitors.

Common Mistakes​

  1. Confusing mathematical expectancy with short-term guarantee — Positive expectancy promises long-term edge; it says nothing about the next 10, 50, or even 100 trades. Ruin can strike during a statistically unlucky but entirely plausible drawdown.

  2. Using historical win rate without accounting for regime shifts — Your last 100 trades had 55% win rate because of a particular market regime. The next 100 might be different. Position sizing should reflect the possibility that win rate drops to 48% or even 45%.

  3. Ignoring variance in position sizing decisions — A system with 55% win rate and $100 average winner/loser pairs needs much more capital than a 55% win rate system with $1,000 average winner/loser. High variance demands larger bankrolls.

  4. Believing that larger edge means you can afford larger position size — A system with +$100 expectancy per trade does not automatically permit 10x larger positions than a +$10 expectancy system. Variance, not expectancy, determines position sizing limits.

  5. Treating ruin probability as a constant — Your ruin probability changes as your account grows or shrinks. If you start with $10,000 and lose to $8,000, your new ruin probability is (usually) lower. If you grow to $15,000, your ruin probability drops further. Recalculate ruin risk quarterly.

FAQ​

If My Expected Value Is Positive, Why Would I Ever Go Broke?​

Because ruin depends on surviving the journey to long-term profits. A short sequence of losses can deplete your account before the positive expectancy has time to compound. This is especially true if position sizes are large relative to your bankroll.

How Much Capital Do I Need to "Guarantee" That Positive Expectancy Works?​

There's no amount that "guarantees" success, but the more capital you have relative to position size, the lower your ruin risk. A practical rule: if your ruin probability is below 1%, your capital is almost certainly sufficient to survive any reasonable number of trades before positive expectancy manifests.

Should I Always Take the Trade with the Highest Risk-Reward Ratio?​

Not if the win rate is low. A 3:1 risk-reward with a 30% win rate is a loser. A 1:1 risk-reward with a 55% win rate is a winner. Risk-reward ratio must be evaluated alongside win probability to determine expectancy.

Can Positive Expectancy Systems Still Fail After 1,000 Trades?​

Extremely unlikely if your position sizing is conservative. After 1,000 trades, the law of large numbers strongly favors the system with positive expectancy; the chance of a 1,000-trade drought in a 55% win rate system is virtually zero. But with large position sizes and insufficient capital, ruin can strike within the first 100 trades.

Does Increasing Leverage Always Increase Ruin Risk?​

Yes. Doubling your position size while keeping capital constant roughly doubles your ruin probability. This is why many traders use fixed fractional position sizing (e.g., 2% risk per trade) rather than fixed dollar amounts; it automatically scales down position size as your account shrinks and vice versa.

If I Have Positive Expectancy, Can I Just Trade Until I'm Rich?​

You can if you size positions conservatively and monitor your account growth. But if you size positions aggressively (risking 10% or more per trade), ruin can interrupt the journey, especially in the early stages when your account is smallest relative to position size.

Summary​

Positive expectancy trading is necessary but not sufficient for financial survival. A system that makes money on average can still bankrupt you if position sizing is mismatched to your bankroll, or if an unlucky sequence of losses strikes before the edge has time to compound. The path from positive expectancy to wealth creation is lined with ruined traders who understood the math but underestimated the role of variance, sequence risk, and bankroll management.

The traders who survive are those who calculate their edge carefully, then size positions conservatively—often risking far less per trade than the raw math would permit. They recognize that ruin probability, not expected value, is the binding constraint on position size. By respecting that constraint, they give their positive expectancy time to work.

Next​

→ How Drawdown Size Drives Ruin Risk