FRM Part II · FRM Exam Part II · Regression Hedging and Principal Component Analysis
A desk wants to hedge a bond portfolio against level, slope and curvature moves using principal component analysis rather than a single DV01 number. Which description of the approach is correct?
The desk should measure the portfolio's exposure to each main principal component (level, slope, curvature) and size hedge positions so net exposure to each is about zero. This goes beyond DV01, which only covers a parallel shift.
- AHedge a single parallel shift because the first component always explains all variation
- BCompute exposure to each of the main principal components and choose hedge instruments so the net exposure to each is approximately zeroCorrect
- CUse only the third component, since it captures the largest variance
- DChoose hedges so that the components are made correlated with each other
Explanation
PCA hedging measures the portfolio's sensitivity to each major factor (level, slope, curvature) and sizes hedges to neutralize each. The first component does not explain all variance, the largest-variance component is the first (not the third), and components are uncorrelated by construction.
Did you get it right without looking?
One question tells you little. A timed set on Regression Hedging and Principal Component Analysis shows your real accuracy, how long you take and where you lose marks.
More Regression Hedging and Principal Component Analysis questions
- A bank holds USD 50 million face of a 10-year bond priced at par with a DV01 of 0.0800 per 100 of face value. It hedges with a 5-year bond p…
- A risk analyst applies principal component analysis (PCA) to the covariance matrix of daily changes in six par swap rates. Which statement a…
- Which is a recognized limitation of using PCA factors from historical data to hedge a fixed income portfolio?
- A risk manager is hedging a 10-year bond position using a PCA-based approach with level and slope factors, and compares it with a single-var…
- A risk analyst applies principal component analysis (PCA) to the covariance matrix of daily changes in eight key-rate points on a government…
- PCA on a yield curve gives eigenvalues of the covariance matrix of 8.0, 1.5, 0.3 and 0.2 (units of bp squared) for the four principal compon…