Skip to content

Adding weighted standard deviations as portfolio risk

A weighted sum of standalone standard deviations generally differs from portfolio standard deviation.

On this page

Inspect the tempting shortcut

For the 40/60 model with standard deviations 20%/10%, adding 0.4×20%+0.6×10% produces 14%. Calling this portfolio volatility silently drops the entered correlation 0.3.

Use the variance identity

Square the weighted own dispersions and include 2wAwBρσAσB. The resulting variance is 0.01288 and its square root is approximately 11.349009%. Keep the weighted 14% figure as a named comparator rather than deleting it.

Identify when equality is justified

Original pair
CorrelationPortfolio volatilityWeighted standalone
+114%14%
0.311.349009%14%
−12%14%

Review the output label

A correct weighted sum can still answer the wrong question. Before interpreting a result, identify whether the task asks for a mean return, a weighted standalone dispersion, variance or portfolio standard deviation. Each uses a different operation.

Further references