Correlation and Portfolio Risk

The Diversification Illusion

You have positions in BTC, ETH, SOL, and a few altcoins. You're diversified, right?

Then March 2020 happens. Or May 2021. Or November 2022. Every asset in your "diversified" portfolio drops 30-50% together. Your careful position sizing went out the window because you didn't account for the one thing that matters most: correlation.

Understanding Correlation

Correlation measures how two assets move together, ranging from -1 to +1:

• •+1: Perfect positive correlation (always move together)
• •0: No correlation (independent movements)
• •-1: Perfect negative correlation (always move opposite)

In a crypto bull market, BTC and ETH might have 0.6 correlation. Manageable. But in a crash? That correlation spikes to 0.95+. Everything moves together when everyone is panicking.

This phenomenon - correlations increasing during stress - is called correlation breakdown and it destroys portfolios that looked safe on paper.

The Math of Combined Risk

If you have two uncorrelated positions each with 2% risk, your total risk is roughly:

Total Risk = √(2%² + 2%²) = √8 = 2.83%

Not 4%. The lack of correlation actually reduces combined risk.

But if those positions are perfectly correlated:

Total Risk = 2% + 2% = 4%

The correlation directly scales your risk. And in crypto, assume high correlation during the moments that matter most.

Measuring Portfolio Correlation

Here's a simple approach: calculate the correlation matrix of your holdings over different time periods.

These correlations tell you: in a BTC crash, expect ETH and SOL to crash too. Maybe not as much, but definitely in the same direction.

Now do the same calculation for only crash periods (>10% weekly drops). The correlations will be higher. That's your real portfolio risk.

Position Aggregation

Given high correlations, you need to think about your total crypto exposure as essentially one bet:

Effective Exposure = Sum of all position sizes × Average Pairwise Correlation

If you have 20% BTC, 15% ETH, 10% SOL with average correlation of 0.8:

• •Simple sum: 45% exposure
• •Correlation-adjusted: 45% × 0.8 = 36% effective single-asset exposure

And in a crisis when correlations hit 0.95? That 45% becomes effectively 43% bet on one thing: "crypto goes up."

Risk Budgeting Across Correlated Assets

Given correlation reality, how do you allocate?

Method 1: Treat crypto as one allocation. Decide your total crypto risk budget (say, 30% of portfolio), then divide among assets. This prevents concentration while acknowledging correlation.

Method 2: Volatility-weighted allocation. Higher volatility assets get smaller allocations:

Weight_i = (1/Volatility_i) / Sum(1/Volatility_j)

This gives equal risk contribution from each asset despite different volatilities.

Method 3: Maximum Diversification Portfolio. Mathematically optimize for the portfolio with lowest correlation to any single risk factor. Complex to calculate but most robust.

The Regime Factor

Remember regime filtering from our Discovery lessons? It applies here too.

In a risk-on regime, correlation between crypto and traditional risk assets increases. Your crypto portfolio starts moving with stocks. Suddenly your "alternative investment" is just more equity beta.

In a risk-off regime, crypto often correlates with other speculative assets, not safe havens. Don't expect BTC to act like gold when fear spikes.

Track your portfolio's correlation to SPY (S&P 500), QQQ (Nasdaq), and VIX (volatility). These correlations reveal when your "diversified" crypto exposure is actually just leveraged tech speculation.

Hedge Positions

True diversification requires assets with low or negative correlation. In crypto, options are limited:

Cash/Stablecoins: Zero correlation, zero return. But preserves capital for reentry.

Funding Rate Arbitrage: Market-neutral strategies have near-zero correlation to direction.

Cross-Exchange Arbitrage: Similarly direction-agnostic.

Short Positions: Negative correlation by definition, but timing-dependent and expensive in bull markets.

The uncomfortable truth: in crypto, the best hedge is often smaller position sizes rather than complex hedging strategies.

Concentration Risk

The opposite of diversification is concentration, and it has its place.

Warren Buffett famously said diversification is protection against ignorance. If you truly know your edge, concentration maximizes returns. But concentration also maximizes drawdowns.

In crypto edge trading, you might have one or two coins where you've validated strong edges. Concentrating on these makes sense - you have statistical backing. Diversifying into coins without validated edges "for diversification" is actually risk increase disguised as risk management.

Portfolio Risk Monitoring

Track these metrics daily:

Total Exposure: Sum of all position notional values ÷ account equity.

Sector Concentration: Are you overweight Layer 1s? DeFi? Memes?

Rolling Correlation: How correlated have your positions been over the last 30 days?

Beta to BTC: Each position's recent correlation and sensitivity to Bitcoin moves.

Value at Risk (VaR): Statistical estimate of maximum expected loss at given confidence level.

When metrics exceed your risk budget, reduce exposure regardless of individual position attractiveness.

The Correlation Spike Playbook

When correlations spike (usually during market stress), execute this playbook:

1. •

   Reduce gross exposure immediately. Don't try to pick winners. Reduce everything.

2. •

   Move to highest-quality positions. In stress, quality outperforms. BTC over alts, large caps over small.

3. •

   Increase cash allocation. Cash is optionality for reentry at better prices.

4. •

   Avoid adding correlation. Don't buy the dip across multiple correlated assets. You're just increasing the same bet.

5. •

   Wait for correlation normalization. When correlations drop back to normal levels, diversification works again.

Practical Application

Here's how we think about it at TargetHit:

Maximum single-asset exposure: 25% of trading capital Maximum total crypto exposure: 60% of trading capital Minimum cash buffer: 20% (always) Target average pairwise correlation: <0.7

When our portfolio correlation exceeds 0.8, we automatically reduce positions until it drops. This happens mechanically, without discretion.

Takeaway

Diversification in crypto is largely an illusion. Most assets move together, especially when it matters most. Your position sizing needs to account for this correlation, and your risk budgets need to treat correlated positions as single exposures.

True risk management in crypto means accepting that you're often making one bet - will crypto go up or down - and sizing that bet appropriately.

Next, we tackle one of the most controversial topics in trading: stop losses. Are they essential protection or profit-killing overhead?