The fundamentals of asset allocation and modern portfolio theory date back to 1952, when Harry Markowitz showed that the risk of an asset was related not only to its volatility, but also to its correlation with the other assets in a portfolio.
This finding quantified the benefits of portfolio diversification. It also led to the construction of an efficient frontier of investment portfolios, with the characteristic that each portfolio on the frontier had the greatest possible return for a particular level of risk.
At the time, the techniques for identifying these optimal portfolios were computationally difficult, but with the development of modern computers, the practical application of modern portfolio theory became possible. Today, the Markowitz model is widely used by investment professionals to allocate assets among various asset classes.
The importance of a consistent and effective asset allocation policy has been highlighted in a series of US-based studies that consider the relationship between long-term, or strategic, asset allocation policy and other elements of the overall investment strategy.
Among the most notable is a study that was published by Brinson, Hood and Beebower in 1986. This analysed the returns of 91 large US pension plan portfolios in the SEI Large Plan Universe from 1974 to 1983. The results showed that the strategic allocation to stocks, bonds and cash explained, on average, 94% of the volatility of total portfolio return.
Further, the contribution of strategic asset allocation to the average portfolio return was found to be approximately equal to the total portfolio return. In other words, the average portfolio return was nearly explained by asset allocation alone. These results demonstrated empirically that asset allocation is the primary determinant of total portfolio return. The analysis has been updated several times since the original study and in each instance reached similar conclusions.
Other techniques for quantifying investment risks and returns have become more popular. Among the most commonly used approaches are Monte Carlo simulations and asset/liability modelling. Through Monte Carlo simulations, investment strategies can be tested under a broader range of statistical assumptions than traditional analysis allows. Asset/liability models help investors understand the impact of alternative investment approaches on pension variables such as the rate of contributions and pension surplus.
In a particular situation, the choice of tools to assist in the asset allocation decision will depend largely on the investors’ objectives, liability profile and investment horizon. Once these are determined, analyses can be structured to target the results the investor is most concerned about.
There are often key trade-offs that summarise an investor’s objectives and drive the asset allocation decision. Modern portfolio theory is based on such a trade-off, namely, portfolio standard deviation versus portfolio return. Most institutional investors are more concerned with satisfying their liabilities or minimising costs than with targeting a particular standard deviation or return. For example, a pension scheme may wish to minimise pension contributions while meeting the minimum funding requirement.
To the extent that the final investment decision relies on the results of a model, it is important to test the chosen strategy under a range of different assumptions. Performance in adverse and volatile markets should be evaluated, particularly as it is during periods of high volatility that the usual relationships among asset classes and economic variables tend to break down.
The investor’s comfort level with the full range of potential outcomes should be questioned, to ensure that the strategy will be maintained in adverse markets. Otherwise, it is probably not the right strategy.
Using similar techniques to those used in the empirical analysis of the asset allocation decision, newer studies have focused on the importance of investment style as a determinant of individual manager and individual fund return. The difference in returns to investment styles such as value, growth and capitalisation are comparable to the differences in returns between stocks and bonds.
Because of the volatility with which investment styles move in and out of favour, a style-neutral portfolio will be more consistent with the overall asset allocation strategy than a portfolio which is tilted towards one style or another.
The tendency of managers and funds to rotate among investment styles, without evidence that this can be done effectively, also suggests that style-specific managers are preferred for implementing asset allocation strategies. The benefits of a manager of managers are seen when implementing a chosen asset allocation strategy.
Asset allocation is important, as is understanding the role it plays in the overall investment process of a manager of managers. Certainly, it is not the only step. Once an investment strategy is set it must be implemented accurately, maintained, and nurtured. To preserve integrity and increase the predictability of results, the asset allocation strategy must be implemented with money managers who can deliver on the characteristics of the asset classes that they are hired to represent.
Our asset allocation process has two levels. The first is a strategic determination based upon asset/liability modelling. This focuses on the optimal combination of equities, bonds and cash considering the pension’s liabilities, risk appetite and other such factors. This level also considers many challenges in the management of investment assets. Foremost are the evolving capital markets and the mechanisms to access those markets. There is an increasing array of complex instruments, which often have attractive features but can be loaded with hidden risk.
Investment opportunities have also become global, presenting a dilemma for pension trustees, who have neither the time nor the resources to identify opportunities nor the ability to execute the individual portfolio diversification. So we take it a step beyond recommending a 60/35/5 (equity/fixed income/cash) portfolio and consider the pension fund’s exposure to emerging market and developed countries.
The second level concentrates on the actual country weightings and regional exposures. As a ‘manager of managers’, we determine the optimal country breakdown while being cognisant of a selected benchmark.
Once managers are selected for a specific mandate, they are given strict country/regional bands to adhere to. This process relies on specific attention to the level of risk in investment portfolios and keeping risk in line with specific benchmarks at the strategy, portfolio and manager level. Portfolio management is a continuous process of controlling risk and maximising performance, and should be researched and monitored on an ongoing basis, once an asset management strategy has been defined.
Joseph Ujobai is managing director of SEI Investments, based in London

Topics