Covariance Matrices

The covariance matrix packages every pairwise variance and covariance among assets into a single PSD matrix that drives portfolio risk.

1. The idea

For a random vector $R = (R_{1}, \dots, R_{n})$ of asset returns, the covariance matrix $Σ$ has $Σ_{ij} = Cov (R_{i}, R_{j})$ , with variances on the diagonal. It is symmetric and PSD by construction.

Portfolio variance for weights $w$ is the quadratic form $w^{T} Σ w$ , which is why $Σ$ being PSD guarantees no portfolio can have negative variance — a basic sanity check on any estimated covariance matrix.

Diversification reduces risk because off-diagonal covariances typically contribute less than the diagonal terms scale; when correlations are less than 1, $w^{T} Σ w$ is strictly smaller than a weighted average of individual variances.

2. Why it matters

Every step of Markowitz optimization, factor risk models, and PCA on returns starts from correctly forming and manipulating $Σ$ .

3. Core mathematics

Covariance matrix

Σ_{ij} = Cov (R_{i}, R_{j}), Σ_{ii} = Var (R_{i})

Portfolio variance

Var (w^{T} R) = w^{T} Σ w

Decomposition via correlation

Σ = D ρ D, D = diag (σ_{1}, \dots, σ_{n})

4. How to recognize it

Think of this when…

Any two-or-more-asset variance/risk computation
Questions giving a correlation matrix and asking for portfolio risk
Setup for Markowitz mean-variance optimization

5. Worked example

Worked example

Two assets have $σ_{1} = 0.2$ , $σ_{2} = 0.3$ , correlation $ρ = 0.5$ . Find the variance of an equal-weighted portfolio.

1
Build the covariance.
$Cov (R_{1}, R_{2}) = ρ σ_{1} σ_{2} = 0.5 (0.2) (0.3) = 0.03$
2
Apply the two-asset variance formula with $w_{1} = w_{2} = 0.5$ .
$Var = w_{1}^{2} σ_{1}^{2} + w_{2}^{2} σ_{2}^{2} + 2 w_{1} w_{2} Cov$
3
Substitute numbers.
$= 0.25 (0.04) + 0.25 (0.09) + 2 (0.25) (0.03) = 0.01 + 0.0225 + 0.015 = 0.0475$
Answer: Portfolio variance ≈ 0.0475, so volatility ≈ 21.8%

6. Common mistakes

Forgetting the factor of 2 on the cross term in the two-asset variance formula
Using correlation where covariance was needed, or vice versa
Assuming diversification always reduces volatility even when correlation is 1

7. Quick check

Off-diagonal entries of a covariance matrix represent:

Portfolio variance is given by:

8. Practice

1 QuantSolve questions use this concept.

Practice this concept

9. Related methods

This lesson is conceptual; see the topic hub for methods.

Create a free account to save lesson progress.