Is The Efficient Portfolio Achievable? Investment and communications solutions, Richard Newell

Richard Newell
Tel and Fax
+64 (0)9 529 1611
Mobile
+64 (0)21 534 456

E-Mail

By Richard Newell

A lot of truly terrible houses have been built, apparently inspired by architectural greats like Frank Lloyd Wright or Mees Van der Rohe. But somehow the purity of the original thought has been corrupted and bastardised by people without the wit to appreciate it.

That lack of appreciation is a charge also levelled at financial advisers, by two of New Zealand’s top finance academics, Ed Vos, associate professor of finance at Waikato University and Francis Milner, a corporate finance analyst with the BNZ. Their original 2002 report, ‘The Theory and Practice of Investment Advice in Financial Planning’ concluded that advisers are not putting forward theoretically good advice because they lack the ability to appreciate how to adjust portfolios according to changes in the clients' risk tolerance. That is, they do not think in terms of investing in the optimal 'market portfolio' and matching this with 'risk free' holdings.

From the academic view, it is symptomatic of a lack of understanding of the concept. Advisers though, don’t see it that way. They believe the theory is just that, a theory, and that in the real world, there are other ways of formulating asset allocation.

Generally, strategic asset allocation considers long term return expectations for different asset classes, their risk characteristics and correlation. The aspirations, well-being and risk tolerance of the investor also play a role. The combination of the risk asset and risk-free assets is the basis for the theory of efficient portfolios. The father of modern portfolio theory, Harry Markowitz, established the concept of the efficient portfolio; one that maximises expected return while minimising risk. Taking into account all the potential combinations of assets, it is possible to determine which portfolios provide the greatest expected return for a given level of risk. The efficient frontier is then created by plotting all efficient portfolios on a single graph.

James Tobin extended Markowitz’s thinking by introducing the concept of risk-free assets. Tobin found that by introducing a risk-free asset into the model, a single portfolio along the efficient frontier could be established as the optimal portfolio. This so-called ‘market portfolio’ was embraced as the most appropriate basket of risk assets for all investors, no matter what their level of risk tolerance. The relative proportions of assets within the basket would not change. What investors needed to determine was how much they would invest at the risk-free rate and how much in the market portfolio.

Vos and Milner explain: “Therefore, customising a portfolio of assets to an individual client’s risk/return preference is not a matter of changing assets or weightings within the risky basket. It is a matter of choosing between the risk free rate and market portfolio.” This equally applies to an aggressive investor who wants to move further up the capital market line from the market portfolio (higher expected return) by borrowing to invest. “Any other choice does not optimise the parameters of risk and return,” say Vos and Milner.

The process of formulating a client’s long term strategic asset allocation is one of the most important, if not the most important component of the whole advice process. If, as has been suggested, asset allocation is responsible for 90% of investment performance, why is the efficient portfolio theory not taken more seriously?

Unfortunately, the theory has come under severe criticism over the course of the recent bear market. Norman Stacey of Diversified Investment Strategies, is just one of many advisers who do not have much time for the academic viewpoint: “The efficient portfolio theory has failed investors over the last three years. It has not provided them with a way through the bear market we have just been through. That is perhaps the issue for the academics to address.

“The bear market has shown that it is market risk that many clients and even advisers cannot handle. The volatile conditions have scared off a lot of investors.”

Economist Donal Curtin also has issues with the efficient portfolio theory. In particular, he questions the validity of adherence to efficient portfolios based on false assumptions: “It’s quite clear that the idea of a stable set of correlation coefficients is simply wrong.”

“The pronounced drift between the asset classes is shown by a study done in the United States, which illustrates that between 1802 and 1870, you were paid 1.7% on average to hold equities over bonds. In the period 1871 to 1925, the equity premium was 3.9% and in the period 1926 to 1990, the differential was 6.8%. As Curtin comments: “Over whatever period you take it, you certainly don’t get numbers that would make the efficient portfolio model reliable.”

Curtin believes that relying heavily on traditional strategic asset allocation methods may not be appropriate for many clients. “This is true over a variety of different time periods. If we look at asset class performance over the last 11 years, property is at the top and equities are at the bottom.” There is still scope for using the principles of modern portfolio theory, but Curtin’s view is that “it is possible to structure an inefficient portfolio based on false assumptions.”

Vos has produced other research* that supports his view that: “low and stable correlations are very possible. But care must be taken as to the proper specific investment within a class.” He defends his position by reiterating that his paper was attempting to ascertain where advisers are failing to firstly understand theory and secondly implement it: “The theory does not say that they 'must find stable low correlations' or that they cannot, from time to time, based on their expert analysis of expected returns, co-variances, and standard deviations, change their minds about what they see as 'their best shot'. Everyone receiving their advice should get the same opinion at the same time.

“So, the recommended mix of what is in the risky basket of investments can, and will, change from time to time. But the proportion of what is in that mix should not change as the client's risk profile changes.”

One criticism levelled at the theory is that traditional beliefs about diversification have not held up, that in the bear market situation, all major markets are highly correlated. But Vos says it is not correct to suggest that asset correlations are all high in bear markets: “This is just another example of people not being willing to undertake the necessary research and analysis to create optimal portfolios.

“Planners seem to feel that the mathematics of diversification are no longer valid. I can prove to you that diversification reduces variability. If planners are not up to using the maths, this is one thing. But if they reject the math as 'not valid' this is quite wrong. My paper only suggests that they should use the math and then come up with their 'best shot'.”
Clearly there is a credibility gap between the theory and the practice. Strictly adhered to, the efficient portfolio model should work and, as such, there is risk for the client in not applying it. Do advisers have the tools at their disposal to bridge this gap? In terms of asset allocation, they rely heavily on their product suppliers. And in terms of risk assessment, no they don’t have the tools. For many clients, the concept of investment risk is something they have learned about the hard way.

The problem of defining a client’s tolerance was illustrated by a controlled study of risk assessment carried out by advisers in Australia. It was shown that advisers’ estimates would have been more accurate if they had made no attempt to understand their clients’ risk tolerance, but had simply assumed all were average. This is not a criticism of advisers. Other studies involving managers and subordinates, doctors and patients, teachers and students, etc. have shown similar inaccuracies in assessing personal attributes. Where the industry has failed is in not using the available knowledge to devise more effective methods of investment risk assessment.

Donal Curtin encourages investment advisers to think outside the box when formulating client asset allocation. “The industry trend is to follow your favourite investment managers of the moment, for example Wellington, or Alliance Capital and just vary the weighting according to the client. A better practice is to use different funds in different proportions across different risk profiles, rather than the same fund manager in different proportions for all levels of client.”

Not unnaturally, Curtin does not follow the advice of the academics: “No, I am guilty on all counts there,” he says. “I am prepared to make strategic shifts in asset allocation. And as an economist, I like to think I have information that will be some guide to the way things are going, so I’m prepared to make the tactical calls. I am also happy to make changes in style and to replace fund managers.”

Diversified Investment Strategies are also firmly in the active asset allocation camp: “It adds value and reduces risk,” says Stacey. He recognises the complexity of the issue and that misinterpretation, or ‘shibboleths’ are apt to confuse the issue. For example, the oft-quoted theory, originally put forward by Brinson, Singer and Beebower, that over 90% of the return of a portfolio could be explained by its asset allocation. Brinson’s research indicated that 90% of the variability of fund performance was due to asset allocation. More recent research by Ibbotson Associates suggests the return attributable to asset allocation is more like 40% for mutual funds and 35% for pension funds. The remaining 60% of fund performance variation results from such other factors as the timing of moves between asset classes, security style (e.g. value or growth stock), security selection and expenses.

The view put forward by Ed Vos is an idealized one, where advisers are constantly monitoring the efficient frontier and the capital asset pricing model in order the structure the most efficient portfolio at that time. The advisers’ view is that they don’t trust the theory and besides, they don’t have time to seek out the optimal solutions.

There would appear to be some consensus here, that the complexity of trying to aggregate various assets to come up with an optimal risk/return profile is beyond the scope of most advisers. Vos and Milner’s research confirms this and industry evidence suggests that advice is proffered on the basis of a set menu of investment options that can only approximate an optimal portfolio.

Vos maintains that “the goal of an 'optimal' portfolio using mathematics is still valid, and all clients should get the same advice. This does not mean that the opinion of optimal cannot change. Indeed, it should as more products become available to enhance the 'optimal' performance.”
Published in ASSET Magazine, October 2003

Further reading:
Theory and Practice of Investment Advice in Financial Planning, Ed Vos and Francis Milner, University of Auckland Business Review, Vol 4, No.2 2002.

* Macroeconomic Drivers of the Non-correlation Between Equity and Commodity Indices. Co Author: Frank Aarts. The Global Business and Finance Research Conference. London, Vol 1 No 1, pp 380-393. 14-17 July.

Private Equity: A Portfolio Approach, Francis Milner and Ed Vos. Journal of Alternative Investments. Vol 5 No 4, Spring 2003. pp 51-65.

Does Asset Allocation Policy Explain 40,90 or 100 per cent of Performance? Roger Ibbotson and Paul Kaplan. www.ibbotson.com

Website by Webtrix.