Perspectives from ISB

A diversified asset allocation methodology should be the  hallmark for any investment portfolio which has the objective of providing returns which meet desired objectives, with a relatively low level of risk.

Traditional methods of  asset allocation, based on either  the simple 60(equities)/40 (bonds) approach or more complicated variations based on modern portfolio theory, can suffer from large swings in returns, and permanent loss of capital, as they do not usually take asset  valuations into account and incorporate flawed measures of risk. James Montier, from the well known value fund manager GMO and the author of several popular books on investing, provides an insightful look at the deficiencies of the traditional approaches and offers some helpful suggestions on how to improve the asset allocation process. To summarise:

-The traditional determination of asset allocation is usually based on historical asset returns.  An important academic study in 1986  pointed out that the asset mix between bonds, equities and cash accounted for 93.6% of  the variation in returns, and suggested that the first step in any asset allocation process should be to determine the long-term weights to different assets in the portfolio.

-In addition, academic studies  suggested that portfolios should be constructed along an “efficient frontier” (maximising a return for a certain level of risk) derived from forecasts of returns, standard deviations and correlations of various assets. The standard 60/40 portfolio lay along such an efficient frontier.

-The problems with this traditional approach are manifold:

1. Risk is not volatility: The  definition of risk is the volatility of  returns, a deeply flawed  approach. Risk is not an elegantly derived  number, but in the words of the pioneer of value investing Benjamin Graham,  it represents  the danger of  a permanent loss of capital. In the words of John Maynard Keynes “ It is largely fluctuations which throw up the bargains and the uncertainty due to fluctuations which prevent other people from taking advantage of them” –  i.e.   risk was higher in 2007 rather than 2009, contrary to what modern portfolio theory would say.

2. Valuation indifference: Modern portfolio theory ignores a fundamental principle of investing – increasing allocations to assets which offer cheap valuations and decreasing allocations when assets are expensive.  Asset class returns are higher when starting valuations (based on  tested valuation  measures like the Graham and Dodd p/e or the initial bond yield) are low and vice-versa lower when valuations are high.

3. Benchmarking alters behaviour: Benchmarking  portfolios against a specific index adversely alters investment behaviour  in three ways: a) forces managers  to compete against an index rather than focus on  the value of an investment, b) measurement of risk changes to  measuring deviations from the benchmark-“tracking-error” and  enhances the issue of “career-risk” which forces managers to stay fully invested, and, c) makes managers to think of return in a relative sense rather than focussing on absolute return.

4. Not enough return: A low return environment forces managers to seek new approaches and invest in unfamiliar and riskier  asset classes rather than changing their return assumptions to reflect the new environment.

-A first-generation new approach to enhance portfolio returns  has often involved mimicking the Yale endowment strategy – the “let’s all look like Yale” phenomenon which involved investing in new (and less liquid) asset classes like private equity, commodities and hedge funds. This approach has several problems:

1.Diversification was often performance chasing: This involved managers chasing the latest and hottest fad in the market, rather than diversifying  to seek cheap  valuations  – i.e. the flows into private equity in recent years when valuations were high.

2.Diversification in name only: Often the diversification was in strategies which were highly correlated – for example, a variety of hedge fund strategies have extraordinary high levels of correlation (nearly 90%) indicating that they were all doing the same thing –momentum chasing and selling volatility.

3. It’s poker not roulette: Investing is usually similar to playing poker (where the behaviour of other participants matters) rather than roulette (where it doesn’t). Chasing high returns results in overcrowded trades which significantly alters the risk/return trade-off – for  example, as investors rushed into commodities, the “roll return” turned from positive to negative and reduced the annual return from about 10%since 1970 to about 5% from 2000.

-A second-generation new approach to enhance portfolio returns has been the “let’s all look like Bridgewater approach” which entails investing on a “risk-parity” approach – the current favourite bad idea.

-This approach involves initially  weighting assets according their volatility – so a traditional 60/40 portfolio would be altered to 13/87 by reducing the weight of the riskier stocks. Given that this would dramatically reduce the returns, risk-parity would involve increasing leverage of the portfolio until the risk equals the original 60/40 portfolio, thereby increasing returns.

-However, the risk-parity approach has  two problems: 1) the definition of risk as volatility is flawed as mentioned previously,  implying altering  allocations at the wrong time without paying heed to valuations, and, 2) relying on leverage to boost meagre returns, ever a good idea,  as it can change a temporary loss into a permanent loss.

-A simpler, and more holistic approach to investing is required. This involves  setting a realistic  real return target and incorporates a more sensible measure of risk and a concern for valuations.

Risk should include:

1) valuation risk –  paying too much for an asset,

2) business risk – i.e. there are fundamental problems with the asset, and,

3) financing risk – leverage.

-This puts valuation at its core , making value investing a truly risk-free approach by taking a value-oriented  view across asset classes and making alterations to allocations based on changes in value.

-A value-driven approach requires patience and a willingness to be a contrarian. Patience is needed because valuations are only mean-reverting over relatively long periods of time(for example, it takes about 7 years for stocks  to mean-revert after  they have moved one standard-deviation from their  valuation trend).

-A willingness to be a contrarian is  critical– i.e. doing the opposite of what everyone else thinks as being sensible. This requires the courage to take a stand against the dominant view, being an independent thinker and the firmness of character to stick to your views.  Unfortunately, these are all traits which are unnatural to human beings.

A fascinating perspective and a  classic “out-of-the-box” approach to investing. four key time-honoured investing principles stand out: value, avoid leverage, patience and contrarianism. Warren Buffet was interviewed recently and asked to give his words of wisdom on investing, and he boiled it down to only two principles: avoid leverage and patience! In a way, patience incorporates both value and contrarianism – and to paraphrase Jeremy Grantham, leverage can remove the only advantage an individual investor has over the institutions – patience, which if only utilised would allow them to beat 99% of institutions over time!.