Correlation Risk
Learn why correlations rise during crypto stress, how to distinguish historical correlation from dependency, and how correlated positions can multiply portfolio loss.
Reading progress — saved on this device
Correlation risk is the danger that positions expected to diversify one another move together—especially during stress—causing aggregate losses far larger than standalone risk estimates imply.
Learning objectives
- Interpret correlation as a relationship rather than a guarantee.
- Estimate the effect of correlation on combined risk.
- Recognise stress correlation and common-factor exposure in crypto portfolios.
What it is
Correlation ranges from −1 to +1 and describes the linear co-movement of returns over a chosen sample. A value near +1 means the assets tended to move in the same direction; near −1 means opposite direction; near zero means little linear relationship in that sample.
Correlation depends on timeframe, lookback and market regime. BTC and altcoins can show moderate daily correlation in calm markets but become highly correlated during broad deleveraging.
Correlation is not causation and it is not the same as economic dependency. Two positions can be structurally linked even before historical data reveals a strong coefficient.
How it works
For two positions with equal volatility, portfolio variance rises as correlation rises. At +1 correlation there is no diversification benefit between equal-direction exposures; at lower correlation some variability can offset.
Stress correlation often matters more than average correlation. During exchange failures, stablecoin depegs or macro risk-off events, common liquidity and collateral channels can synchronize previously separate assets.
Rolling correlation can reveal changes but is backward-looking. Scenario analysis should therefore supplement statistics: ask what happens if every risky crypto asset falls 25% together.
Long/short strategies face correlation-break risk. A hedge can fail when the long asset falls more or the short asset rallies more, even if historical relationship looked stable.
How to analyse and apply it
| Check | Why it matters | What to verify |
|---|---|---|
| Lookback/timeframe | Changes measured correlation. | Compare several windows and frequencies. |
| Stress regime | Shows diversification under pressure. | Measure or scenario-test crisis periods. |
| Common factor | Explains co-movement. | Identify BTC beta, liquidity, sector, collateral or venue links. |
| Hedge stability | Determines spread risk. | Stress correlation breakdown, not just average fit. |
Risk rules should be written before a live position is opened and evaluated across many trades or scenarios. A control that is changed only after losses appear is discretionary damage control, not a repeatable risk system.
Worked example and thought exercise
Two equal-weight assets each have 40% volatility. If correlation is zero, the simplified portfolio volatility is about 28.3%. If correlation rises to 0.8, it increases to about 38%. Diversification that looked meaningful nearly disappears.
A portfolio with five altcoins each risking 1% to individual stops may not have 1% total risk. If a BTC liquidation cascade drives all five through their stops with slippage, aggregate loss can approach or exceed 5%.
Thought exercise: why is a correlation matrix from a quiet three-month period a weak stress test?
Common mistakes and practical workflow
- Treating low historical correlation as permanent diversification.
- Using ticker count as a substitute for factor diversification.
- Ignoring correlation between strategy P&Ls, venues or collateral sources.
- Assuming a hedge remains effective when market structure changes.
Practical workflow
- Measure correlations over multiple horizons.
- Group positions by common economic factor.
- Run stress scenarios with correlations approaching one for risky assets.
- Cap aggregate risk for correlated clusters.
- Review hedge effectiveness during the worst historical and hypothetical regimes.
✅ Knowledge checkpoint
- What does a correlation of +1 imply for same-direction positions?
- Why can stress correlation exceed normal correlation?
- How does correlation affect two-asset portfolio variance?
- Why is owning many altcoins not automatically diversified?
FAQs
❓ Does zero correlation mean independence?
No. It only indicates little linear co-movement in the measured sample; nonlinear or regime-dependent dependence can remain.
❓ Should I use Pearson correlation only?
It is a common starting point, but rank correlations, tail dependence and scenario analysis can reveal other relationships.
❓ Can stablecoins be correlated with risk assets?
Their price target may differ, but portfolio losses can still become linked through collateral, liquidity or issuer events.
❓ How often should correlations be updated?
Regularly enough for the strategy horizon, while remembering that no rolling estimate replaces stress scenarios.
📋 Summary
Correlation risk is fundamentally about failed diversification. Statistical estimates help, but crypto portfolios need common-factor mapping and stress scenarios because relationships can strengthen sharply during exactly the periods when diversification is most needed.
Want this in a personalised order?
Take the crypto assessment and get a custom path of 10 modules matched to what you already know. Free, no card required.
Build my path →