February 2010 Issue: Key Points

On the economic front, we believe we have recently passed a number of potentially important tipping points, involving recognition of the true nature of the predicament we face, the ability of Western societies and governments to meet the challenges that lie ahead, and the likely future behavior of China and Iran. Unfortunately, none of this bodes well for the future. We continue to believe that most investors still underestimate the probability that we will return to the high uncertainty regime. We also believe that this is generating significant asset class over and undervaluations.

This month's feature article reviews the debate underway at Norway's sovereign wealth fund over active versus passive management, and the use of a risk factor based approach to asset allocation. We use this as an opportunity to review the confusing terminology that is used in this debate, including the meaning of terms such as active return, alpha, beta, exotic beta, alternative beta and the like. Our approach starts with the observation that the only portfolio that all investors can, if they choose, passively and simultaneously hold is the market cap weighted portfolio. Any deviation from this therefore represents some type of active management, whether it is based on holding preferences that differ from those of the average investor, or believing that one has a superior ability to forecast fundamental asset values and/or future investor behavior. As we have repeatedly noted over the years, we also believe that financial markets are a complex adaptive system that usually operate out of, but are drawn to, equilibrium and efficient asset pricing. This creates the opportunity for successful active management. However, more often than not, such efforts fail to produce anticipated results because of the constantly evolving nature of economic and financial market relationships and/or because of the incremental costs associated with active management, including trading commissions, bid/ask spreads, and market impact, as well as research and taxes. Given the challenges involved, we continue to believe that only certain active strategies make sense (those focused on uncorrelated alpha) and only for some investors (those who must achieve relatively high real returns to achieve their long term goals).

The issue of risk factor based asset allocation is also extremely interesting. However, we are a long way from endorsing it as an alternative to the traditional asset-class based approach. Why? First because given the continued debate over the definition and existence of different risk factors, and the underlying sources of the observed (and time varying) risk premiums for bearing them, we do not see a solid theoretical basis for estimating the equilibrium return an investor should expect to earn for exposure to different risk factors, or how these returns should vary under disequilibrium conditions (e.g., our high uncertainty and high inflation regimes). Second, as we have demonstrated in our Principal Components Analysis of asset class returns between 1990 and 2006 and then in 2007-2008, the loading of different asset classes on different statistical risk factors varies over time. This should also be the case for risk factors based on different methodologies, whether it be macroeconomic variables, psycho-social variables (e.g., investor sentiment), or financial market variables (e.g., the value, small cap, credit, and momentum risk factors). As a practical matter, this seems to make the implementation of a risk factor based approach to asset allocation very difficult in an environment where both risk factor premiums and risk factor loadings on different asset classes are constantly changing over time. So while we continue to closely follow this extremely interesting debate, we won't be changing our approach to asset allocation any time soon.