Volatility Is Not Risk
Most people think volatility is danger.
It isn't.
Volatility is how much prices have been moving, not how much they can move. That difference sounds semantic until the day your stops don't work.
Formally, volatility is usually defined as the standard deviation of returns:
Already baked into that equation are two assumptions:
- Returns are reasonably well-behaved
- The distribution is stable enough to estimate
Markets violate both, regularly.
Volatility spikes are not "high volatility." They are distribution changes.
When that happens, every indicator that relies on the past - ATR included - becomes a historian, not a risk manager.
ATR measures realized range. Markets trade expected variance.
The gap between those two is where surprises live.
And surprises are not random. They are simply underestimated.
Practical Implications
If volatility is not risk, then what should you actually do?
Stop using historical volatility as a risk measure.
ATR, Bollinger Bands, and realized volatility indicators tell you what happened, not what can happen. When distributions shift, these become useless - or worse, misleading.
Use wider stops or dynamic stops.
If you're sizing positions based on ATR, you're assuming the future will look like the past. During regime shifts, your stops will get hit repeatedly. Either widen them significantly, or make them volatility-adaptive.
Monitor for distribution changes, not volatility levels.
A sudden spike in volatility isn't "high volatility" - it's a signal that the underlying distribution has changed. When this happens, your entire risk framework needs recalibration. Don't just tighten stops; reconsider your assumptions.
Don't trust backtests that assume stable distributions.
Most backtests implicitly assume the future will resemble the past. They'll show you beautiful equity curves right up until the day they don't. Look for strategies that survive regime shifts, not ones that optimize for the last regime.
The practical takeaway: Risk management based on historical volatility is risk management based on yesterday's news. When the distribution changes, you need a framework that adapts, not one that assumes stability.
Part I of the series: Trading What Breaks Linear Thinking.