Immunization Using Principal Component Analysis

Abstract

oped a useful method for classifying term structure shifts. A given shift is thought of as a vector whose elements correspond to rate changes at different maturity dates. Viewed in this manner, Fisher-Wed immunization protects a bond portfolio against a shift in the direction of a vector with all components equal. Different duration measures given in the literature are designed to protect portfolios against term structure shifts in a particular direction. Given that we do not know a priori the direction in which the term structure will shift, we are faced with the practical problem of choosing the best single direction in which to anticipate a shift. This is an important problem, because the success of an immunization strategy depends upon how well the assumed term structure model approximates reality. A more general and accurate model allows the term structure to shift in multiple directions. Under a multiple-direction model, an immunized portfolio must be protected against shifts in each fundamental (independent) direction. The problem is determining the set of fundamental directions. The only guide we have is the history of term structure movements. This article shows how principal component I

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 74 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.