Asset Allocation

Asset allocation is the most important part of the investment process. Manage this process diligently and the likelihood of realizing investment objectives is greatly improved. Get it wrong, and the decisions that follow (sector, style, individual manager and security selection) will have little effect.

In a 1986 study* researchers estimated 93% of the variability in portfolio returns to be attributable to the asset allocation decision. Other studies have generally estimated the contribution at 60-95%, with the variation largely due to differing data sets, time periods, methodology, and underlying assumptions. While the results of these studies have often been misinterpreted, the importance of an effective asset allocation policy cannot be denied.

The commodity and real estate markets in 2007 provide a recent example of the relative importance of asset allocation on portfolio returns. Commodities were on fire that year. Real estate was not. The Goldman Sachs Commodity Index returned 32.7% in 2007, while the Wilshire Real Estate Investment Trust index lost 17.6%. Active managers consider long-term annual performance of 1-2% over their respective benchmark, after fees, to be very good, and it is. However, any value that may have been added (or subtracted) by active management in either of these sectors in 2007 clearly would have been overwhelmed by the big picture – the relative amounts allocated to each sector as a whole.

This example also highlights the importance of valuation and diversification in the asset allocation process. We use measures of value and estimated long-term economic growth, combined with historical and fundamental relationships, to develop a set of capital market assumptions. We apply these assumptions to a broad range of asset classes, using them to form estimates of long-term expected returns, volatilities and correlations for each. These estimates provide the raw material for the quantitative side of the asset allocation study.

Modern Portfolio Theory prescribes the use of mathematical techniques (mean-variance optimization) to derive a set of asset allocations that optimize the tradeoff between risk (variance, or volatility) and return. The resulting set of allocations form what is known as the Efficient Frontier. While the efficient frontier defines the optimal tradeoff of risk and return within the confines of the model, it does not tell us where along the frontier we should be. The specific asset allocation is selected based on specific client objectives, constraints, and risk tolerance.

It is important to note that the optimization routine used is a precision instrument, but the data we feed it (return, variance, and correlation estimates) are very imprecise. As a result, these techniques can produce extreme and often ridiculous results if taken literally. It is for this reason that optimization routines are used only as a tool to guide us in the asset allocation study. Professional judgement, with a keen eye toward the prudent management of risk, are the driving forces in determing an appropriate asset allocation.

*Determinants of Portfolio Performance”; Brinson, Hood, Beebower; Financial Analyst’s Journal, July/August, 1986