Skip to content

FRM Part II · FRM Exam Part II · Portfolio Construction

A portfolio manager uses mean-variance optimization with expected returns estimated from a short history. The optimizer produces extreme, concentrated long and short positions that change sharply when inputs are slightly altered. Which action most directly addresses this problem in practice?

Adding position-size constraints or shrinking estimated inputs toward a prior reduces the extreme, unstable weights. Mean-variance optimization magnifies estimation error in expected returns, so constraining or regularizing inputs is the practical remedy, whereas more iterations or dropping risk would not fix the sensitivity.

  1. AAdd sensible position-size constraints or shrink the inputs toward a priorCorrect
  2. BIncrease the number of optimization iterations
  3. CReplace the covariance matrix with a diagonal matrix of zeros
  4. DSwitch to maximizing expected return without a risk term

Explanation

Mean-variance optimizers act as error maximizers: noise in expected returns is amplified into extreme weights. Constraints or shrinkage (e.g., Bayesian priors) stabilize weights. More iterations do not change the sensitivity, a zero covariance matrix is invalid, and ignoring risk worsens concentration.

Did you get it right without looking?

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

Practise Portfolio ConstructionFree diagnostic mock

More Portfolio Construction questions