Print this article

Analytics: Getting to the core of model portfolios

Ronald Surz

6 April 2007

A look at an important and elegant enhancement to model portfolio programs. Ronald Surz is president of PPCA, a San Clemente, Calif.-based software firm that provides advanced performance-evaluation and attribution analytics, and a principal of RCG Capital Partners, a Denver, Colo.-based fund-of-hedge-funds manager.

Model portfolios are being touted as the investment industry's newest and best solution. Investment managers submit their models to broker sponsors who then package these style and asset classes into various diversified programs, tailored to investors with varying risk appetites.

Despite all the hoopla, these programs could stand improvement, and certainly will evolve. For example "core" -- the stuff in the middle between "value" and "growth" -- has gone largely unnoticed.

Strategic bets

Most model portfolio programs acknowledge a middle range in company size by employing large-, middle- and small-capitalization managers. But none realize that there is a similar middle in styles: a core in between value and growth. Ignoring this core leads to poor diversification and the consequent exposure to poor performance.

Mid-cap companies do not perform like large and small companies, so there are good reasons to treat them as a separate category. Similarly, "homogenized core" stocks -- the stuff in the middle as opposed to a blend of value and growth -- perform differently than value and growth.

Using the Surz style index definitions, core surprises about a third of the time on a quarterly basis by either outperforming or underperforming both value and growth. In other words, core doesn't fall in between these two styles about a third of the time. Accordingly, ignoring homogenized core is like throwing away the filling in an Oreo cookie - it's the sweet spot of diversification, around which value and growth revolve.

Here's what's happening in model portfolio programs. Value and growth models receive allocations that are roughly the market weights in these styles, plus the sponsor usually applies strategic bets above or below market weights to enhance performance.

But these bets unknowingly underweight homogenized core all the time. Research conducted by Frank Sortino of the Pension Research Institute indicates that allocations to skillful value and growth managers systematically underweight the middle of the market.

This is understandable in light of the scrutiny that most managers are under to maintain style consistency. Managers have incentive to sell companies that drift toward the middle, away from their declared style. The result is an unintended bet in model portfolios away from homogenized core -- and of course that's at odds with effective diversification.

This is an easy problem to fix. Broker sponsors can refer to the composition of homogenized core as another, albeit passive, model. The constituents of the current large cap core are provided in the following table.

Homogenized large-cap core model portfolio: Q1 '07