Skip to content

FRM Part I · FRM Exam Part I · Regression Diagnostics

A risk analyst's regression of bond spread changes on two highly correlated regressors, equity index return and equity volatility change, shows large standard errors. She considers remedies. Which action is most appropriate to address the multicollinearity while preserving the model's purpose of predicting spreads?

The appropriate action is to drop one of the correlated variables or merge them into one composite variable, accepting some omitted-variable risk. Robust standard errors, lagged residuals and rescaling do not change the correlation between regressors, so they do not cure multicollinearity.

  1. ADrop one of the two correlated variables or combine them into a single composite variable, accepting some omitted-variable riskCorrect
  2. BApply White's robust standard errors, which remove the correlation between regressors
  3. CAdd lagged residuals as a regressor to eliminate the correlation between the regressors
  4. DMultiply both variables by a constant to reduce their correlation

Explanation

Typical remedies are dropping one of the collinear variables, combining them, or obtaining more data. Robust standard errors address heteroskedasticity, not correlation among regressors. Adding lagged residuals targets serial correlation. Rescaling by a constant does not change correlation, which is scale invariant.

Did you get it right without looking?

One question tells you little. A timed set on Regression Diagnostics shows your real accuracy, how long you take and where you lose marks.

Practise Regression DiagnosticsFree diagnostic mock

More Regression Diagnostics questions