Syntax
COVARIANCE.P(array1, array2)
| Parameter | Description |
|---|---|
| array1 | Parameter of the COVARIANCE.P function. |
| array2 | Parameter of the COVARIANCE.P function. |
Examples
Portfolio covariance
Formula
=COVARIANCE.P(C2:C253, D2:D253)
Returns the population covariance of two assets' returns. A negative value suggests the assets are good diversification candidates.
Sensor correlation
Formula
=COVARIANCE.P(A2:A1000, B2:B1000)
Measures how two sensor readings vary together across 999 measurements.
Standardized version
Formula
=COVARIANCE.P(A2:A50, B2:B50) / (STDEV.P(A2:A50) * STDEV.P(B2:B50))
Dividing covariance by the product of standard deviations gives the correlation coefficient — same as CORREL.
Common Errors
#N/A
Arrays have different numbers of data points.
#DIV/0!
One or both arrays contain no numeric values.
Tips
Same as COVAR
COVARIANCE.P and COVAR are identical. Use COVARIANCE.P for clearer code that explicitly states population covariance.
For samples, use COVARIANCE.S
COVARIANCE.S divides by n-1 and is appropriate for sample data. The difference is significant for small data sets.
Relationship to CORREL
CORREL = COVARIANCE.P / (STDEV.P of X * STDEV.P of Y). Correlation is just covariance normalized to a -1 to 1 scale.
Try COVARIANCE.P in Viztab
Import your data and use COVARIANCE.P with 370+ other formulas. No signup required.
Open Viztab