Here's a question that sounds academic but absolutely isn't: how do you measure the risk of losing everything versus just losing a lot? Most traders encounter Value at Risk early in their education, nod along, and assume the maths has their back. It doesn't — not always, and especially not when you're trading the kind of volatile commodity equities that populate the ASX small-cap end.

VaR gives you a number with satisfying precision. "There's a 5% chance of losing more than $12,400 tomorrow." Sounds rigorous. The problem is that number is built on a bell curve assumption — it treats extreme moves as nearly impossible. Australian commodity equities, particularly junior miners and energy explorers, routinely do things that a normal distribution considers a ten-sigma event. Taleb has been screaming this from rooftops since the 1990s, and markets keep proving him right.

CONCEPTProbability of Ruin asks whether your account survives long-term — VaR only asks how bad tomorrow might look on a normal day.
WARNINGA Gaussian VaR model on ASX iron ore or lithium explorers will systematically underestimate tail risk — sometimes by an order of magnitude.
KEY IDEAFat tails mean the rare catastrophic event is far more likely than your model believes — and it only needs to happen once.

Probability of Ruin (PoR) approaches the problem differently. Rather than asking "what's the worst normal day?", it asks "given my edge, position sizing, and drawdown tolerance, what's the probability my account hits zero before my strategy proves itself?" It's a survival question, not a daily snapshot. Think of VaR as checking tomorrow's weather and PoR as asking whether your house can survive a decade of seasons.

Gaussian VaR vs Fat-Tail RealityLoss →FreqGaussianFat TailMore riskout here

For traders wanting to go deeper, the academic groundwork is well established. The core flaw in RiskMetrics-style VaR — its reliance on normally distributed returns — is extensively documented, and Investopedia's Value at Risk explainer is a clean starting point for understanding what the metric actually measures before questioning it. The concept of fat-tailed distributions, formally called heavy-tailed distributions on Wikipedia, explains mathematically why extreme outcomes cluster more than Gaussian models predict. Practical PoR frameworks, rooted in Kelly Criterion and ruin theory, are covered well in the Investopedia ruin theory article — worth an hour of any quant trader's time.

The practical takeaway: run your position sizing through a PoR lens, not just a VaR headline. If a single position's worst realistic outcome — not its modelled normal worst — threatens more than 2% of capital, the model is flattering you.

VaR tells you the cost of a bad day. Ruin theory tells you whether you'll still be trading in three years. Only one of those questions actually matters.

This content is for educational purposes only and does not constitute financial product advice. Past performance is not indicative of future results. Profit Logic Ltd (ACN 688 669 936) accepts no responsibility for errors or omissions in this content or anywhere on this website. Always seek advice from a licensed financial adviser before making investment decisions.