Open main menu

A bivariate Gaussian probability density function centered at (0, 0), with covariance matrix given by
Sample points from a bivariate Gaussian distribution with a standard deviation of 3 in roughly the lower left-upper right direction and of 1 in the orthogonal direction. Because the x and y components co-vary, the variances of and do not fully describe the distribution. A covariance matrix is needed; the directions of the arrows correspond to the eigenvectors of this covariance matrix and their lengths to the square roots of the eigenvalues.

In probability theory and statistics, a covariance matrix, also known as auto-covariance matrix, dispersion matrix, variance matrix, or variance–covariance matrix, is a matrix whose element in the i, j position is the covariance between the i-th and j-th elements of a random vector. A random vector is a random variable with multiple dimensions. Each element of the vector is a scalar random variable. Each element has either a finite number of observed empirical values or a finite or infinite number of potential values. The potential values are specified by a theoretical joint probability distribution.

Intuitively, the covariance matrix generalizes the notion of variance to multiple dimensions. As an example, the variation in a collection of random points in two-dimensional space cannot be characterized fully by a single number, nor would the variances in the and directions contain all of the necessary information; a matrix would be necessary to fully characterize the two-dimensional variation.

Because the covariance of the i-th random variable with itself is simply that random variable's variance, each element on the principal diagonal of the covariance matrix is the variance of one of the random variables. Because the covariance of the i-th random variable with the j-th one is the same thing as the covariance of the j-th random variable with the i-th random variable, every covariance matrix is symmetric. Also, every covariance matrix is positive semi-definite.

The covariance matrix of a random vector is typically denoted by or .



Throughout this article, boldfaced unsubscripted   and   are used to refer to random vectors, and unboldfaced subscripted   and   are used to refer to scalar random variables.

If the entries in the column vector


are random variables, each with finite variance and expected value, then the covariance matrix   is the matrix whose   entry is the covariance[1]:p. 177


where the operator   denotes the expected value (mean) of its argument.

In other words,


The definition above is equivalent to the matrix equality







where  .

Generalization of the varianceEdit

This form (Eq.1) can be seen as a generalization of the scalar-valued variance to higher dimensions. Recall that for a scalar-valued random variable  


Indeed, the entries on the diagonal of the auto-covariance matrix   are the variances of each element of the vector  .

Conflicting nomenclatures and notationsEdit

Nomenclatures differ. Some statisticians, following the probabilist William Feller in his two-volume book An Introduction to Probability Theory and Its Applications,[2] call the matrix   the variance of the random vector  , because it is the natural generalization to higher dimensions of the 1-dimensional variance. Others call it the covariance matrix, because it is the matrix of covariances between the scalar components of the vector  .


Both forms are quite standard, and there is no ambiguity between them. The matrix   is also often called the variance-covariance matrix, since the diagonal terms are in fact variances.

By comparison, the notation for the cross-covariance matrix between two vectors is



Relation to the correlation matrixEdit

The auto-covariance matrix   is related to the autocorrelation matrix   by


where the autocorrelation matrix is defined as  .

Relation to the matrix of correlation coefficientsEdit

An entity closely related to the covariance matrix is the matrix of Pearson product-moment correlation coefficients between each of the random variables in the random vector  , which can be written as


where   is the matrix of the diagonal elements of   (i.e., a diagonal matrix of the variances of   for  ).

Equivalently, the correlation matrix can be seen as the covariance matrix of the standardized random variables   for  .


Each element on the principal diagonal of a correlation matrix is the correlation of a random variable with itself, which always equals 1. Each off-diagonal element is between −1 and +1 inclusive.

Inverse of the covariance matrixEdit

The inverse of this matrix,  , if it exists, is the inverse covariance matrix, also known as the concentration matrix or precision matrix.[3]

Basic propertiesEdit

For   and  , where   is a  -dimensional random variable, the following basic properties apply:[4]

  2.   is positive-semidefinite, i.e.  
  3.   is symmetric, i.e.  
  4. For any constant (i.e. non-random)   matrix   and constant   vector  , one has  
  5. If   is another random vector with the same dimension as  , then   where   is the cross-covariance matrix of   and  .

Block matricesEdit

The joint mean   and joint covariance matrix   of   and   can be written in block form


where   and  .

  and   can be identified as the variance matrices of the marginal distributions for   and   respectively.

If   and   are jointly normally distributed,


then the conditional distribution for   given   is given by


defined by conditional mean


and conditional variance


The matrix   is known as the matrix of regression coefficients, while in linear algebra   is the Schur complement of   in  .

The matrix of regression coefficients may often be given in transpose form,  , suitable for post-multiplying a row vector of explanatory variables xT rather than pre-multiplying a column vector x. In this form they correspond to the coefficients obtained by inverting the matrix of the normal equations of ordinary least squares (OLS).

Covariance matrix as a parameter of a distributionEdit

If a vector of n possibly correlated random variables is jointly normally distributed, or more generally elliptically distributed, then its probability density function can be expressed in terms of the covariance matrix.[6]

Covariance matrix as a linear operatorEdit

Applied to one vector, the covariance matrix maps a linear combination c of the random variables X onto a vector of covariances with those variables:  . Treated as a bilinear form, it yields the covariance between the two linear combinations:  . The variance of a linear combination is then  , its covariance with itself.

Similarly, the (pseudo-)inverse covariance matrix provides an inner product  , which induces the Mahalanobis distance, a measure of the "unlikelihood" of c.[citation needed]

Which matrices are covariance matrices?Edit

From the identity just above, let   be a   real-valued vector, then


which must always be nonnegative, since it is the variance of a real-valued random variable. A covariance matrix is always a positive-semidefinite matrix, since


Conversely, every symmetric positive semi-definite matrix is a covariance matrix. To see this, suppose   is a   positive-semidefinite matrix. From the finite-dimensional case of the spectral theorem, it follows that   has a nonnegative symmetric square root, which can be denoted by M1/2. Let   be any   column vector-valued random variable whose covariance matrix is the   identity matrix. Then


Complex random vectorsEdit

Covariance matrixEdit

The variance of a complex scalar-valued random variable with expected value   is conventionally defined using complex conjugation:


where the complex conjugate of a complex number   is denoted  ; thus the variance of a complex random variable is a real number.

If   is a column vector of complex-valued random variables, then the conjugate transpose is formed by both transposing and conjugating. In the following expression, the product of a vector with its conjugate transpose results in a square matrix called the covariance matrix, as its expectation:[7]:p. 293


where   denotes the conjugate transpose, which is applicable to the scalar case, since the transpose of a scalar is still a scalar. The matrix so obtained will be Hermitian positive-semidefinite,[8] with real numbers in the main diagonal and complex numbers off-diagonal.

Pseudo-covariance matrixEdit

For complex random vectors, another kind of second central moment, the pseudo-covariance matrix (also called relation matrix) is defined as follows. In contrast to the covariance matrix defined above Hermitian transposition gets replaced by transposition in the definition.



  • The covariance matrix is a Hermitian matrix, i.e.  .[1]:p. 179
  • The diagonal elements of the covariance matrix are real.[1]:p. 179


If   and   are centred data matrices of dimension   and   respectively, i.e. with n rows of observations of p and q columns of variables, from which the column means have been subtracted, then, if the column means were estimated from the data, sample correlation matrices   and   can be defined to be


or, if the column means were known a priori,


These empirical sample correlation matrices are the most straightforward and most often used estimators for the correlation matrices, but other estimators also exist, including regularised or shrinkage estimators, which may have better properties.


The covariance matrix is a useful tool in many different areas. From it a transformation matrix can be derived, called a whitening transformation, that allows one to completely decorrelate the data[citation needed] or, from a different point of view, to find an optimal basis for representing the data in a compact way[citation needed] (see Rayleigh quotient for a formal proof and additional properties of covariance matrices). This is called principal component analysis (PCA) and the Karhunen–Loève transform (KL-transform).

The covariance matrix plays a key role in financial economics, especially in portfolio theory and its mutual fund separation theorem and in the capital asset pricing model. The matrix of covariances among various assets' returns is used to determine, under certain assumptions, the relative amounts of different assets that investors should (in a normative analysis) or are predicted to (in a positive analysis) choose to hold in a context of diversification.

See alsoEdit


  1. ^ a b c Park,Kun Il (2018). Fundamentals of Probability and Stochastic Processes with Applications to Communications. Springer. ISBN 978-3-319-68074-3.
  2. ^ William Feller (1971). An introduction to probability theory and its applications. Wiley. ISBN 978-0-471-25709-7. Retrieved 10 August 2012.
  3. ^ Wasserman, Larry (2004). All of Statistics: A Concise Course in Statistical Inference. ISBN 0-387-40272-1.
  4. ^ Taboga, Marco (2010). "Lectures on probability theory and mathematical statistics".
  5. ^ Eaton, Morris L. (1983). Multivariate Statistics: a Vector Space Approach. John Wiley and Sons. pp. 116–117. ISBN 0-471-02776-6.
  6. ^ Frahm, G.; Junker, M.; Szimayer, A. (2003). "Elliptical copulas: Applicability and limitations". Statistics & Probability Letters. 63 (3): 275–286. doi:10.1016/S0167-7152(03)00092-0.
  7. ^ Lapidoth, Amos (2009). A Foundation in Digital Communication. Cambridge University Press. ISBN 978-0-521-19395-5.
  8. ^ Brookes, Mike. "The Matrix Reference Manual".

Further readingEdit