This content is from: Portfolio

Strategically Rethinking Your Asset Allocation

It’s time for investors to change the way they look at asset allocation.

Campbell HarveyPierre LaGrange

The art of asset allocation hasn’t changed that much since Harry Markowitz laid the groundwork for Modern Portfolio Theory more than 50 years ago. In almost zombielike fashion, investors typically recalibrate their investment return projections every three to five years. But taking the usual approach to strategic asset allocation will result in missed opportunities, because it ignores the huge impact that economic changes can have on medium-term returns.

The fundamentals of the economy have changed dramatically during the past tumultuous year and a half. In December 2007 many people did not realize we were in a recession. Today we all know we are in a recession and that it is likely nearing its end. As a result, the target returns used in strategic asset allocation need to be adjusted. History shows that prospective returns are much higher for a portfolio formed in the middle of a recession than for one formed midway through a recovery. Furthermore, equity volatility and correlations between asset classes depend strongly on the business cycle: Markets are more likely to move together in a recession.

There has also been unprecedented global money creation during the past six months. There is considerable uncertainty that central banks have the will to tighten monetary policy in the face of very slow employment growth; such uncertainty produces inflation risk and changes expected returns.

Portfolio managers must take all this information into account. We are not just talking about reweighting a portfolio back to its long-term policy objective. A change in expected returns implies a deviation from the policy objective.

Most allocations allow for a target-tracking error. In this framework the risk is essentially the volatility of the portfolio. Using this approach the investment manager is assumed to know exactly the prospective returns, volatilities and correlations. But this cannot be the case. To make things worse, small changes in the “exactly” known inputs can lead to dramatic (and impractical) changes in portfolio weights.

Bayesian methods have been developed that enable the investment manager to specify a distribution of returns, rather than a single number, incorporating uncertainty directly into asset allocation. Suppose an asset class has an expected return of 8 percent and you believe there is an 80 percent chance that the realized return will be 2 percent to 14 percent. You need to use that information. There is very little chance that the realized return will be exactly 8 percent. Incorporating the range of possibilities will lead to a more stable allocation.

Although uncertainty is important, a comprehensive approach to risk goes even further. Suppose a stock in your portfolio has a beta of 1.2. We have just introduced uncertainty, which enables us to use the distribution of the exposure — that is, that the beta falls in the 0.8-to-1.5 range. But what if the stock’s expected return is driven by more than one beta? That is called ambiguity. The ambiguity here is in the choice of model.

One painful lesson that many investors learned during this economic crisis is that reports of alpha were greatly exaggerated. Some managers were claiming alpha that was, in fact, nothing more than simple risk-taking. A comprehensive analysis of risk reduces the chances that investors are purchasing market risk — or beta — when they think they are getting alpha.

One of the great challenges for portfolio managers is integrating asset allocation into stock selection. Often top-down asset allocation is done using benchmark indexes for the asset class. But why use the FTSE all shares index, for instance, as the benchmark for a U.K. portfolio if the investment proposed by our stock selection team has a lower projected volatility or a different industry mix than that of the index? Successful asset allocation strategies integrate the bottom-up with the top-down, using target portfolios for each asset class rather than common benchmarks.

During the recent turbulence many managers were stung because they failed to reduce their strategic equity weights even when there were clear signs that a recession was approaching. One such sign was the inverted U.S. yield curve in 2006; this configuration has presaged a recession in every business cycle since the 1960s, without any false signals. The yield curve and other indicators, such as valuation ratios, were screaming for a reduced strategic equity weight in 2006.

Well, the yield curve isn’t inverted anymore. Valuation ratios are much lower. This is exactly the time to increase equity weights. Of course, you can choose to do nothing. You might even argue that you have missed the upswing in equities. Both would be mistakes. Strategic asset allocation enables you to break from the pack. At a minimum initiate a review — and do it soon.

Pierre Lagrange is co-founder and senior managing director of GLG Partners with overall responsibility for the London-based firm’s global equity products. Campbell Harvey is a finance professor at Duke University and senior adviser to GLG’s investment management team. For their full analysis, visit

Related Content