Adding liabilities as a reference point to MPT

Frank Sortino and Hal Forsey, along with Robert C. Merton, make some very important points about the importance of asset-liability management considerations not only for defined benefit pension funds, but also for defined contribution funds (Pensions & Investments, Other Views, Nov. 24).

Elsewhere, Mr. Merton has made a similar argument about how endowment funds need to be managed on a liability-aware basis. Interestingly, while we can debate which products can achieve the desired results — whether it is a managed account in a qualified default investment alternative or a desired target return — or which performance metric may be used, the even more critical implication of all of these statements is that we have been taught and are teaching our students in MBA, Ph.D. and CFA programs the wrong model.

The cause of our retirement crisis may have been sown by a simple single assumption early in the academic development of modern portfolio theory and the capital asset pricing model, which are the basis of every pension fund's asset allocation, and asset allocation accounts for 90% of fund risk. Sadly, this oversight has not been corrected and the problems could be perpetuated if not fixed immediately.

For all its wonderful appeal of being embedded in a simple equilibrium setting of investor behavior, CAPM assumes that investors try to maximize wealth — which Messrs. Merton, Sortino and Forsey rightly point out is the wrong assumption. CAPM makes no mention of liabilities, which is quite interesting as one would be believe that the existence of liabilities, however liberally defined, is the reason d'etre for investing.

Instead we should be assuming that all investors maximize funded status (assets divided by liabilities) or surplus (assets minus liabilities). I have recently shown that if you make this adjustment and focus on funded status, or surplus, you can derive a “relative asset pricing model” or RAPM, which is a liability-centric approach and provides a new asset pricing model and liability-centric recommendations for asset allocation. The CAPM is a very, very special case of this more general model.

There is extensive literature on liability-aware asset allocation, including seminal papers by William F. Sharpe, although surprisingly little to none on its impact on asset prices. RAPM attempts to correct this by basing the model on how investors actually behave. I have also shown how making this adjustment for liabilities might provide a path to reconciling MPT with its major detractors: proponents of prospect theory, developed by Daniel Kahneman and Amos Tversky and described as a behavioral finance framework for making decisions under risk. They criticized MPT for lacking a reference point, which I believe could be liabilities.

It is time we corrected this mistake in our academic models and began to show the current and future generation of pension chief investment officers that the correct models to use for their asset allocation studies is RAPM and not CAPM, or else the retirement crisis may be exacerbated.


Chairman and co-founder

Mcube Investment Technologies Inc.

Great Falls, Va.

This article originally appeared in the December 8, 2014 print issue as, "Adding liabilities as a reference point to MPT".