# Martingale and why it fails

> Doubling after losses looks safe on short samples, but the risk of ruin grows with every trade.

Source: https://quantraider.com/learn/risk-management/martingale

Course: Risk management and position sizing (lesson 6 of 10) · Level: Intermediate · 3 min

Martingale sizing multiplies your risk after each loss, hoping a single win will recover everything. It looks safe on short samples, but its risk grows faster than most accounts can absorb.

## The rule

Start with a base amount. After a loss, multiply the last risk by a multiplier (often 2×); after a win, return to the base. With a reward-to-risk of 1 and a multiplier of 2, a win after any number of losses recovers all earlier losses and gains exactly the base amount.

**Worked example: A recovery that works, and the ladder behind it**

1. Base risk €50, multiplier 2, reward-to-risk 1.
2. Three losses of €50, €100 and €200 total €350.
3. The next stake is €400. A win gains €400, so the net result is +€50, the base amount.
4. The stakes keep doubling: €50, €100, €200, €400, €800, €1,600, €3,200, €6,400.
5. Seven straight losses cost 50 × (2^7 − 1) = €6,350, and the eighth stake would be €6,400.

With €10,000, seven losses leave €3,650, which cannot cover the €6,400 stake, so the sequence cannot continue.

## Why it fails in practice

Martingale only guarantees recovery if you have unlimited capital, no position limits and no time limit. None of these exist: capital is finite, brokers cap size and margin, and you cannot trade forever. The method turns many small wins into rare, very large losses.

A run of seven losses at a 50% win rate has probability 0.5^7 = 0.78%, which feels negligible. Over 200 trades, however, a run of that length occurs at some point with a probability of roughly one in two (a rough estimate), and one such run in the example ends the account.

**Strengths**

- Simple rule that is easy to automate.
- Many small wins in a row can make the equity curve look smooth for long periods.

**Weaknesses**

- Risk grows exponentially during a losing streak, exactly when the account is weakest.
- A single long streak can erase all earlier gains and the whole account.
- It changes the shape of the losses, not the underlying expectancy of the strategy.
- Position limits, margin requirements and costs make the theory unreachable.

> **A smooth curve is not low risk**
>
> A backtest of martingale can look excellent until the first long streak. Check the maximum drawdown and the largest stake in the sample, not only the final result.

## Key takeaways

- Martingale multiplies risk after losses and resets to the base after a win.
- It recovers losses only with unlimited capital, no position limits and no time limit, which never apply in practice.
- Seven losses in a row at 2× from €50 already require a stake of €6,400.
- It reshapes returns into frequent small wins and rare huge losses without improving the edge.

> Educational content, not financial advice. Trading involves risk of loss.
