All calculators

Correlation Matrix and Portfolio Volatility

Your return series

Separate values with commas, semicolons, tabs or spaces; you can paste columns from a spreadsheet. Every series needs the same number of periods. The defaults are made-up figures; this tool has no market data. Write 2 for +2% and −1 for a 1% loss.

Portfolio weights (optional)

Each weight is from 0% to 100%. The portfolio figures assume these weights hold every period.

Portfolio standard deviation per month
1.0109%
Per month, from 12 periods. About 3.5017% a year (approximate). Diversification ratio 1.4255.

Line 1 read as series names: Series A, Series B, Series C.

These figures describe one past sample of the returns you entered. They are not a verdict on any mix and say nothing certain about what comes next.

Shading repeats the number in each cell: 0.70 to 1 0.30 to 0.69 −0.29 to 0.29 −0.69 to −0.30 −1 to −0.70. Blue shades are positive, orange negative.

Notes on these inputs

    How it is worked out

    • Covariance = Σ (x − mean x)(y − mean y) ÷ (n − 1). Each sample standard deviation (sd) is the square root of a series' own covariance.
    • Correlation = covariance ÷ (sd of x × sd of y), always from −1 to 1. It is n/a when a series does not vary.
    • Portfolio variance = w′Σw = Σi Σj wi wj covij, with weights as fractions of 1. It is per period, like the returns; the portfolio sd is its square root.
    • Annualised sd ≈ per-period sd × √(periods a year), an approximation that assumes independent returns and ignores compounding.
    • Diversification ratio = weighted average of the sds ÷ portfolio sd. A ratio of 1 means combining the series removed none of the ups and downs in this sample; above 1 means the mix moved less than the weighted average of its parts.

    Limits

    • Correlation only measures straight-line co-movement. Two series can be closely linked in a curved or one-sided way and still show a correlation near 0.
    • Correlations change over time. In past sharp market declines, correlations between many risky assets have often risen, so a mix that looked spread out in calm periods may move more together when it matters.
    • Short samples are noisy, and one extreme period (an outlier) can move a correlation a lot. Check the sample size in the notes above.

    Diversification is covered in risk and diversification and asset allocation basics; concentration in drawdown and concentration. For return per unit of volatility, try the Sharpe and Sortino calculator. This is education, not a recommendation about any security.

    What this calculator shows

    Paste two to six return series in percent, one column per series and one row per period, with an optional header row of names. Every series needs the same number of periods, from 3 to 600. The calculator returns the Pearson correlation for every pair, the covariance matrix and each series' sample standard deviation. If you enter weights that add up to 100%, it also returns the portfolio's standard deviation and its diversification ratio. It has no market data: every number is one you type in.

    How the numbers are worked out

    Worked example (hypothetical)

    The defaults are three made-up series of 12 monthly returns, each with a mean of 1%. Series A and B have a correlation of about 0.82, A and C about −0.69, and B and C about −0.93. With weights of 50%, 30% and 20%, the portfolio variance is 11.24 ÷ 11, about 1.0218, so the standard deviation is about 1.0109% a month. The weighted average standard deviation is about 1.4410%, which gives a diversification ratio of about 1.4255.

    Limits of this model

    Learn investing basics

    For educational purposes only. This material does not provide investment advice.