Sunday, December 09, 2007

CDI vs. LDI?

Wilshire has a fairly interesting white paper out with the title Commitment-Driven Investing (CDI): LDI as We 'C' It. The paper's rather semantic focus on commitment rather than liability is probably a consequence of its US accounting context, so we are not going to hold that against it. The thrust is certainly correct, namely that CDI, i.e. the Wilshire variety of LDI is aimed at managing the true economic funding ratio of pension plans and is therefore inherently a risk management technique. 

Confusing at best however is the frequent reference to two equally important (but partially conflicting) goals of CDI, whereas the only correct objective (from the sponsor's perspective) must be to minimise the cost of running the plan at a given level of benefit security. Also, the authors seem to be struggling - quite fashionably, one might add - with the rationale behind marking-to-market of pension liabilities. The impending revision of IAS 19 will deal with that issue.

Their conceptual conclusion, nevertheless, is an intriguing one: The commitment discount rate is set to be a random variable described by the expected inflation characteristics of the individual plan's benefit stream. This is where the authors can start their optimisation engines, hopefully without being blinded by the light of past correlation stability. 

While we feel more comfortable with the more comprehensive approach recently proposed by Waring & Siegel, the authors deserve a prize for originality as this paper is the first that I know of to include the complete lyrics of a Springsteen song.

No comments: