Skip to content

FRM Part I · FRM Exam Part I · Regression Diagnostics

A regression of loan loss rates on GDP growth shows residuals that are systematically positive when unemployment is high and negative when unemployment is low. Unemployment is not in the model and is correlated with GDP growth. Which conclusion is most appropriate?

The model likely suffers from omitted variable bias, and adding unemployment is a sensible fix. Residuals that vary systematically with an excluded variable correlated with GDP growth indicate misspecification, which biases the GDP coefficient. Heteroskedasticity alone would not bias the slope estimates.

  1. AThe model likely suffers from omitted variable bias, and adding unemployment is a sensible remedyCorrect
  2. BThe model suffers only from heteroskedasticity, which biases the slope
  3. CThe problem is perfect multicollinearity and a regressor must be dropped
  4. DThe residual pattern indicates the model is correctly specified

Explanation

Residuals related systematically to an omitted variable that is also correlated with an included regressor signal omitted variable bias, which makes the GDP coefficient biased and inconsistent. The remedy is adding the relevant variable. Heteroskedasticity alone does not bias slopes.

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