Beta Exposure
For a high-level definition of beta and beta exposure see the main glossary entry here.
The default beta exposure in the application is calculated using a weighted least squares (WLS) regression, that includes a lag of the factor. The regression equation can be represented by
The lagged variable is necessary to account for asynchronous data (see our white paper, Asynchronous Data and Serial Correlation in Financial Time Series for more).
The value of the beta exposure in the application is the sum of the two betas in the equation. If you would like to see the two beta values separately, you can use the Beta(t) and Beta(t−1) fields.
Our default beta calculation uses a window length of 252 business days, with a decay factor of 0.99. This combination has a half-life of 61 days, and captures 92% of the weight of the infinite series.