On this page
Read the complete default record
Asset A has 40% capital weight, B 60%, dispersions 20%/10%, and assumed correlation 0.3. All describe one hypothetical return period. Covariance is 60 percentage points squared; variance is 128.8 percentage points squared.
Interpret the different scales
| Output | Value | Interpretation |
|---|---|---|
| Portfolio standard deviation | 11.349009% | Square root of total variance |
| Weighted standalone dispersion | 14% | Comparator before covariance reduction |
| Difference | 2.650991 percentage points | Comparator minus portfolio dispersion |
| A weighted contribution | 6.908092 percentage points | Signed Euler component |
| B weighted contribution | 4.440916 percentage points | Signed Euler component |
Change one assumption deliberately
Holding weights and dispersions fixed, increasing correlation increases variance when both weighted exposures are positive. Changing a weight also changes the own variance terms, so a weight comparison is not simply a correlation comparison. Use the curve to inspect the entered model, not to infer future performance.
Inspect endpoint and zero cases
At weight A=100%, the result is A’s dispersion. At equal positive dispersions, equal weights and correlation −1, model volatility is zero and contribution ratios are omitted. If a dispersion is zero, its product with correlation is zero and that correlation is only a stipulated unused parameter.