Calculate The Covariance Between The Variables In Excel

Use this interactive calculator to compute sample or population covariance from two data series, preview the exact Excel formula you need, and visualize how the variables move together with a responsive chart.

How to calculate the covariance between the variables in Excel

Covariance is one of the most useful measures for understanding whether two variables tend to move together. If one variable increases while the other usually increases too, the covariance is positive. If one tends to rise while the other falls, the covariance is negative. If there is little consistent joint movement, the covariance is near zero. When analysts ask how to calculate the covariance between the variables in Excel, they usually want a practical answer: which function to use, how to set up the worksheet, and how to interpret the result without confusion. This guide covers all of that in a direct, business-ready format.

Excel makes covariance calculations easy because modern versions include two dedicated functions: COVARIANCE.S and COVARIANCE.P. The first is for a sample, and the second is for a population. The distinction matters because the denominator changes. Sample covariance divides by n – 1, while population covariance divides by n. That small difference changes the final value, especially in small datasets.

Quick rule: Use COVARIANCE.S(range1, range2) when your rows are a sample from a larger universe. Use COVARIANCE.P(range1, range2) when your rows represent the entire population you want to describe.

What covariance tells you

Covariance measures the direction of the linear relationship between two variables. Unlike correlation, covariance is not standardized, so its numeric size depends on the units of the original data. If you measure one variable in dollars and another in units sold, the covariance will be in combined units that are harder to compare across datasets. That does not reduce its value. Covariance is still essential in finance, forecasting, operations, econometrics, and experimental analysis because it reveals joint variability.

• Positive covariance: the two variables tend to move in the same direction.
• Negative covariance: the variables tend to move in opposite directions.
• Near-zero covariance: there is little linear co-movement.
• Large magnitude covariance: stronger combined movement, though the scale depends on the units used.

Excel functions you should know

Most users only need two Excel functions:

1. =COVARIANCE.S(A2:A11,B2:B11) for sample covariance
2. =COVARIANCE.P(A2:A11,B2:B11) for population covariance

Older spreadsheets may still reference legacy formulas like COVAR. Although Excel often continues to support older syntax for compatibility, the current best practice is to use the explicit functions above. They are clearer, easier to audit, and less likely to be misunderstood by colleagues reviewing your workbook.

Step by step example in Excel

Suppose column A contains monthly advertising spend and column B contains monthly sales revenue. You want to know whether higher ad spend tends to coincide with higher sales. Here is a clean workflow:

1. Place the first variable in one column, such as cells A2:A7.
2. Place the second variable in the matching rows of another column, such as B2:B7.
3. Make sure both columns contain the same number of numeric observations.
4. Click into an empty result cell.
5. Enter either =COVARIANCE.S(A2:A7,B2:B7) or =COVARIANCE.P(A2:A7,B2:B7).
6. Press Enter.

That is the basic calculation. If the resulting covariance is positive, sales tend to rise when ad spend rises. If negative, sales tend to fall when ad spend rises, though that pattern is less common in that example. If the value is close to zero, there may be little linear relationship, or the data may be noisy.

Worked data example

Consider this small dataset of paired monthly values. Here, ad spend is measured in thousands of dollars and sales are measured in thousands of dollars as well.

Month	Ad Spend X	Sales Y	X minus mean(X)	Y minus mean(Y)	Product of deviations
1	10	55	-5	-8.33	41.65
2	12	58	-3	-5.33	15.99
3	15	64	0	0.67	0.00
4	18	68	3	4.67	14.01
5	20	72	5	8.67	43.35

The sum of the products of deviations is about 115. If this is a sample of 5 observations, sample covariance is 115 / 4 = 28.75. If it is the full population, population covariance is 115 / 5 = 23.00. In Excel, those values would be returned by COVARIANCE.S and COVARIANCE.P respectively.

Sample covariance versus population covariance

This distinction causes many errors in financial models and operational dashboards. If you are analyzing survey responses, monthly performance samples, or any subset of a larger process, the sample version is usually appropriate. If your worksheet contains every observation in the population you care about, then the population function is more accurate.

Function	Excel syntax	Denominator	Best use case	Example result from sample data above
Sample covariance	COVARIANCE.S(array1,array2)	n – 1	When your rows are a sample from a larger population	28.75
Population covariance	COVARIANCE.P(array1,array2)	n	When your rows represent the full population	23.00

How to interpret covariance correctly

Interpretation should focus first on the sign, then on the business context. A positive sign means the variables generally move together. A negative sign means they move oppositely. The raw magnitude is useful inside the same project, especially when units stay consistent, but less useful for comparing unrelated datasets. For example, a covariance of 250 in one workbook and 12 in another does not automatically mean the first relationship is stronger. The variables may simply have larger measurement scales.

If you need a scale-free measure, calculate correlation too. Correlation standardizes the relationship onto a range from -1 to 1. In many Excel analyses, covariance is used in the background for portfolio variance, regression preparation, or matrix algebra, while correlation is used for quick reporting and interpretation.

Common reasons Excel covariance formulas return unexpected results

• Mismatched ranges: both arrays must have the same number of observations.
• Blank or text cells: non-numeric values can distort your data preparation workflow.
• Wrong function choice: using population instead of sample, or the reverse, changes the denominator.
• Outliers: a few extreme values can strongly influence covariance.
• Units mismatch: mixing dollars, percentages, and transformed values can make the result harder to interpret.

Manual formula behind Excel covariance

It helps to know what Excel is doing internally. The sample covariance formula is:

Cov(X,Y) = Sum[(Xi – mean(X)) * (Yi – mean(Y))] / (n – 1)

The population formula is similar, but divides by n instead of n – 1. This matters because the sample denominator corrects for the fact that the sample mean is itself estimated from the data.

Why analysts use covariance in finance, forecasting, and operations

Covariance is foundational in portfolio analysis because portfolio risk depends not only on the volatility of each asset, but also on how assets move together. In demand planning, covariance can help reveal whether promotions and sales move jointly. In education or healthcare analytics, it can indicate whether two indicators tend to rise or fall together across time or across groups. In each case, Excel provides a convenient entry point before more advanced statistical software is needed.

For example, if a retailer tracks discount percentage and units sold, a positive covariance may indicate that deeper promotions coincide with stronger sales volumes. That alone does not prove causation, but it is often a strong signal that deserves deeper investigation. Similarly, a manufacturer may compare machine temperature and defect counts. If covariance is positive, higher temperatures may be associated with more quality issues.

Practical Excel tips for cleaner covariance analysis

1. Keep paired observations on the same row.
2. Use structured tables so formulas expand automatically when new rows are added.
3. Remove non-numeric formatting issues before analysis.
4. Plot a scatter chart to visually confirm the relationship.
5. Consider adding correlation with =CORREL(range1, range2) for easier communication.
6. Document whether your result is sample or population covariance.

Comparison of covariance outputs from a realistic business example

The table below compares several paired business scenarios using plausible numerical results. It shows why sign and context matter more than raw size alone.

Scenario	Variable X	Variable Y	Sample covariance	Interpretation
Retail marketing	Weekly ad spend	Weekly sales	31.42	Positive joint movement, higher spending usually aligns with higher sales
Manufacturing quality	Machine temperature	Defect rate	4.88	Positive relationship, hotter production periods align with more defects
Energy efficiency	Outside temperature	Heating demand	-52.17	Negative relationship, warmer weather aligns with lower heating demand

Authoritative references for deeper statistics guidance

If you want academically grounded explanations of covariance, variance, and interpretation, these sources are strong references:

• NIST Engineering Statistics Handbook
• Penn State Statistics Online, introductory statistics resources
• National Center for Education Statistics

When covariance is not enough on its own

Covariance should be seen as one part of a broader analysis workflow. A positive covariance does not prove that changes in X cause changes in Y. There may be hidden variables, time effects, seasonal patterns, or random noise. If your goal is prediction, you may need regression. If your goal is communication, correlation is usually easier for non-technical audiences. If your goal is comparing many assets or factors, a covariance matrix may be more informative than a single pairwise value.

Still, if your immediate task is to calculate the covariance between the variables in Excel, the process is simple and reliable: organize paired data carefully, use the correct function, verify your ranges, and pair the result with a chart. That combination gives you both a numerical answer and a visual sense of whether the relationship is positive, negative, or weak.

Final takeaway

To calculate covariance in Excel, place your two variables in matching ranges and use either COVARIANCE.S or COVARIANCE.P. Choose sample covariance when your data is a subset, and population covariance when it is the complete set. Interpret the sign first, use the magnitude carefully, and add a scatter chart if you want a more intuitive view. The calculator above automates the same math, generates the matching Excel formula, and plots the paired values so you can move from raw data to insight faster.