- Matched periods
- 5
- Slope (benchmark beta)
- 2.1
- Fitted intercept per period
- 1.2 %
- Residual standard error
- 0.6325 percentage points
- Benchmark sample variance
- 2.5 percentage points²
- Sample covariance
- 5.25 percentage points²
- In-sample R²
- 97.351 %
- Sample correlation
- 0.9867
Equal-weight matched period returns; least squares includes an intercept. The fitted intercept is not automatically Jensen alpha or evidence of skill. No significance test, prediction interval, causation or future-return estimate is supplied.
Dots are the supplied paired observations. The solid line is the least-squares fit with an intercept; the dashed line marks zero residual. Points are not joined as a time path. The table retains period order and duplicate observations. This is an in-sample descriptive fit.
View paired observations and residuals
| Period | Benchmark (%) | Asset (%) | Fitted (%) | Residual (percentage points) |
|---|---|---|---|---|
| 1 | -2 | -3 | -3 | 0 |
| 2 | -1 | -0.5 | -0.9 | 0.4 |
| 3 | 0 | 1 | 1.2 | -0.2 |
| 4 | 1 | 2.5 | 3.3 | -0.8 |
| 5 | 2 | 6 | 5.4 | 0.6 |
On this page
Use matched compatible returns
Enter 3–100 comma-separated observations for each series in matching period order. Both use percentages and the same period, currency and income/cost basis. The calculator uses equal observation weights and fits asset return on the chosen benchmark with an intercept.
Read the original default record
Benchmark returns −2/−1/0/1/2 percent and asset returns −3/−0.5/1/2.5/6 percent give slope 2.1, intercept 1.2% per supplied period and residual standard error approximately 0.632456 percentage points. They are hypothetical observations.
Inspect both graphical views
The scatter plots actual observations and the fitted line within the observed benchmark range. The residual view plots signed deviations against benchmark returns. Points are not connected as a time path. The expandable table retains input period order, repeated observations, fitted values and residuals.
Handle non-identifiable or constant records
A constant benchmark cannot identify a slope and is rejected. A constant asset record can give slope zero and a defined intercept, but its correlation and R² ratios are undefined and omitted.
Retain the descriptive model scope
The tool does not choose a market proxy, infer matching dates, subtract a risk-free series or perform significance/causality/prediction analysis. A raw fitted intercept is not automatically Jensen alpha or proof of skill. Use the record and model appropriate to the learning task.