Skip to content

IAI Actuarial Core Principles · Economic Modelling · Single and multifactor models for investment returns

Which of the following is a recognised limitation of using a multifactor model for forecasting investment returns?

A key limitation is that sensitivities to factors are estimated from historical data and may change over time, so the model can give unreliable forecasts. Factor choice also needs judgement, and the model does not guarantee lower variance.

  1. AFactor sensitivities estimated from past data may be unstable over time, so forecasts can be misleadingCorrect
  2. BThe model cannot be used to attribute past performance to different risk sources
  3. CThe model always produces a lower variance than a single-index model for any portfolio
  4. DThe model requires returns to be perfectly normally distributed to estimate any parameter
  5. The model eliminates the need to choose factors, since the data selects them without judgement

Explanation

Factor loadings are estimated from historical data and can change with market regimes, which weakens forecasts. Factor models do support performance attribution, do not guarantee lower variance, and require judgement in choosing factors. Regression estimation does not need perfect normality.

Did you get it right without looking?

One question tells you little. A timed set on Single and multifactor models for investment returns shows your real accuracy, how long you take and where you lose marks.

Practise Single and multifactor models for investment returnsFree diagnostic mock

More Single and multifactor models for investment returns questions