PromptDesk Pro ← Back to the blog

Risk of ruin: the metric that decides whether you're still trading next year

Two traders run the same system: 52% hit rate, 1.5:1 ratio. The first risks 1% per trade and is still trading five years later. The second risks 8% "because the system works" — and goes broke in month four, with the system still working. The difference between the two wasn't the edge: it was the risk of ruin.

What risk of ruin is

It's the probability that your account falls to a point you can't (or aren't allowed to) come back from, before your statistical advantage has time to show up. That point doesn't have to be zero:

• At a prop firm, "ruin" is touching the trailing drawdown: $4,500 on a $150K Topstep account.
• In a personal account, it can be the level where you can no longer trade your minimum size, or the drawdown where you know you'll psychologically quit (for most people, well before −50%).

The central point is uncomfortable but mathematical: you can have a winning system and still have a high probability of going broke, if your position size isn't consistent with your system's variance.

What it depends on (the three levers)

1. Your edge: win rate and win/loss ratio. The thinner the edge, the more brutal the effect of size.
2. Size per trade: the dominant lever. Risk of ruin doesn't grow linearly with size — it explodes. Doubling your risk per trade can multiply your probability of ruin by ten.
3. The distance to your ruin threshold: how many "units of risk" fit between your balance and the point of no return. With 1% per trade and a threshold at −20%, 20 straight losses fit. With 5%, 4 fit.

The Kelly criterion: the ceiling, not the target

The Kelly criterion calculates the fraction of capital that maximizes long-term growth:

f* = W − (1 − W) / R

W = win rate (e.g. 0.55)
R = average win / average loss (e.g. 1.4)

f* = 0.55 − 0.45 / 1.4 = 0.229 → 22.9% of capital

Risk 22.9% per trade? Never. Full Kelly assumes you know your exact win rate (you don't: you estimated it from a sample), and it tolerates 50%+ drawdowns that no human trades through with discipline. That's why the standard practice is half Kelly or quarter Kelly: you capture most of the growth with a fraction of the pain. In the example: between 5.7% and 11.4%... which for a prop firm with a trailing drawdown is still extremely high — there, the firm's rules are in charge, not Kelly.

The most useful reading of Kelly is the inverse one: if Kelly comes out negative, you don't have an edge — no position size fixes a losing system. And if you're risking more than your system's full Kelly, you're mathematically doomed to ruin even if the system wins.

How to estimate it with YOUR data (not textbook formulas)

Closed-form risk-of-ruin formulas assume fixed-size trades and binary outcomes. Your real trading isn't like that: it has $80 wins and $600 wins, mixed losses, concentrated days. For real data the honest method is Monte Carlo:

1. Take your real per-trade P&Ls.
2. Resample them thousands of times (thousands of alternative "runs" of your same trading).
3. Define your ruin threshold (e.g., your prop firm's trailing drawdown, or −$3,000).
4. Risk of ruin = % of simulations that touch that threshold at any point.

We explain the full method in the Monte Carlo simulation guide.

How to lower it without killing profitability

• Lower your risk per trade first. It's the lever that moves the needle most. Going from 2% to 1% reduces growth somewhat; it reduces risk of ruin enormously.
• Fixed risk in % (not in $): risking a percentage of the current balance makes losses shrink during bad streaks — an automatic brake against ruin.
• Stop the day, not the account: a personal daily loss limit (even if your firm doesn't require one) turns spiral days into bounded losses.
• Beware of "making it back": sizing up after losing is the exact recipe for maximizing risk of ruin. It's doing anti-Kelly.

This content is educational and does not constitute financial advice. The calculations described estimate probabilities from past data and do not guarantee future results. Trading involves risk of loss.