Covariance
Overview
Covariance measures how two assets move together over time. It is the mathematical engine behind diversification, volatility, portfolio variance, correlation matrices, and the entire risk structure of a portfolio.
Most advisors talk about “risk” as if it’s a property of individual assets — high‑risk stocks, low‑risk bonds, volatile sectors, stable ETFs. But portfolio risk is not the sum of individual risks. Portfolio risk is a relationship problem, not a component problem.
ClearLedger Analytics computes covariance deterministically using adjusted‑close return series, centered around each asset’s mean return, and annualized to institutional standards.
1. What Covariance Represents
Covariance answers one question:
“Do these two assets move together, or do they move differently?”
Interpretation:
Positive covariance → assets rise and fall together
Negative covariance → assets move in opposite directions
Zero covariance → movements are unrelated
Covariance is the foundation of diversification.
2. Conceptual Definition & Covariance of Two Assets Formula
Covariance represents the average product of each asset’s deviation from its mean return.
If two assets deviate from their means in the same direction at the same time, covariance is positive. If they deviate in opposite directions, covariance is negative.
Cov(X, Y) = Σ [ (Xᵢ − μX) × (Yᵢ − μY) ] / (n − 1)
3. How ClearLedger Analytics Computes Covariance
Step‑by‑Step Calculation:
Build a stable date spine to align trading days.
Retrieve daily adjusted returns for both assets.
Compute mean return for each asset.
Center returns around mean using (X − μX) and (Y − μY).
Compute covariance as the average of (X − μX)(Y − μY).
Annualize covariance by multiplying by 252 trading days.
This produces a clean, annualized covariance value used directly in portfolio variance, correlation matrices, diversification scoring, and efficient frontier optimization.
ClearLedger Analytics also computes covariance using Excel’s MMULT engine inside VertiPaq for deterministic, high‑speed columnar performance.
4. Covariance Builds the Portfolio’s Risk Matrix
How to Compute the Covariance Matrix of a Portfolio
ClearLedger Analytics constructs a full covariance matrix (Σ):
Σ = covariance matrix
Each cell represents the covariance between two assets. The matrix is stored in VertiPaq for:
high‑speed columnar access
deterministic recalculation
stable risk geometry
The covariance matrix is the risk map of the portfolio.
5. Portfolio Risk Comes From Covariance, Not Volatility
Individual volatility matters — but only inside the covariance structure.
Portfolio risk is computed using the quadratic form:
Risk = wᵀ Σ w
This shows:
risk depends on weights
risk depends on covariance
risk depends on interactions, not just components
Two assets with identical volatility can produce radically different portfolio risk depending on their covariance.
6. Why Covariance Is the True Engine of Diversification
Diversification is not owning many stocks, sectors, ETFs, or bonds.
Diversification is owning assets with low or negative covariance.
If assets move differently, risk spreads out. If assets move together, risk concentrates.
ClearLedger Analytics quantifies this precisely.
7. Covariance Determines Risk Contribution
Risk contribution tells you which assets are actually driving portfolio risk.
An asset with high volatility but low covariance may contribute less risk than expected.
An asset with moderate volatility but high covariance may contribute more risk than expected.
ClearLedger Analytics computes risk contribution using:
RCᵢ = wᵢ × (Σw)ᵢ
This reveals the true risk drivers.
8. Covariance Determines How Weights Should Change
When ClearLedger Analytics optimizes weights, covariance determines:
which assets reduce risk when increased
which assets increase risk when increased
which assets amplify each other
which assets offset each other
how diversification behaves under constraints
Covariance is the geometry of the optimization problem.
Without covariance, optimization is blind.
9. Covariance Makes Risk Explainable to Clients
Clients understand:
“These two stocks move together.”
“This ETF offsets volatility in your tech sleeve.”
“This position reduces risk because it behaves differently.”
ClearLedger Analytics makes this story visible through:
risk tables
correlation matrices
risk contribution charts
diagnostics comparing current vs optimized risk
This is how advisors defend their recommendations.
Conclusion
Covariance is the mathematical engine that determines portfolio risk.
It explains how assets interact, how diversification works, how risk spreads or concentrates, how weights should change, and how the portfolio behaves under stress.
ClearLedger Analytics computes covariance deterministically, transparently, and at institutional quality — giving advisors a risk model they can trust and explain.