One man’s meat is another man’s poison is an old adage that can well describe the world of investment. What are today’s traditional asset classes can look alternative to an Oxbridge college with anywhere up to 700 years history of ownership of agricultural land, forestry and silver plate, with requirements for liquid assets catered for through extensive cellars of fine wines! However, the last 50 years of institutional investment for pension funds and other investors have been dominated by investment in what are now seen as the core mainstream liquid assets of developed market equities, bonds and cash.
The development of Modern Portfolio Theory (MPT) by the Nobel laureates Harry Markowitz and William Sharpe et al. has been spectacularly successful in setting a disciplined framework for the process of asset allocation between these classes, while the subsequent development of efficient market theory has clearly led to a transformation of the way that equities in particular, are managed. But behind this story, there is a darker side to the success. The beauty and simplicity of MPT can beguile the unwary and nowhere is this more dangerous than in the assessment and inclusion of alternative assets.
Private equity, hedge funds, property and commodities are increasingly being seen as a useful and in many cases, essential component of a balanced investment portfolio primarily due to expectations of higher returns, but also because of a perceived low correlation with existing mainstream asset classes.
While liquidity among many alternative assets can be poor, they may offer high yields or capital gains over long time periods that make them very attractive for long-term investors. Adding lowly correlated alternative assets, which may offer higher returns or comparable returns to traditional assets but with less downside risk, should conceptually improve both returns and the risks of a portfolio. Despite the difficulties of incorporating that view into a more rigorous intellectual framework, there is little doubt that there is much to be gained by the inclusion of a wider range of assets including alternatives within long-term investment portfolios.
When investors seek to incorporate alternatives into an asset allocation framework they face serious problems that cannot be readily solved using a traditional MPT type mean-variance optimisation. MPT is based around the premises that high returns are directly correlated with higher risks, and that the risk can be described mathematically by the standard deviations of returns symmetrically around the mean.
Diversifying across asset classes that are not highly correlated with each other then reduces the risk of the portfolio as a whole with the objective of portfolio optimisation being to maximise the return for a given level of risk producing an “efficient frontier” of possible portfolios to choose from. There are however, a number of implicit assumptions behind this analysis, which need to be understood when considering the addition of alternative assets.
In particular, MPT assumes that returns and risks can be measured accurately; that asset class returns are symmetrically distributed in mathematically what is described as a ‘normal’ distribution; that liquidity is irrelevant; that markets as a whole are reasonably efficiently priced so that an index return represents what an investor can be expected to receive; and that volatility is the best measure of risk for investors. Incorporating alternative assets into the context of a traditional MPT asset allocation framework based on a mean-variance type optimisation process is not only misleading but also dangerous for a number of reasons.
Assets that have a high ‘Sharpe ratio’, which can be described as the extra return above risk free bonds divided by the standard deviation of the returns, will have high allocations in any asset allocation process using an MPT mean-variance optimisation. Successful hedge fund returns often have such high Sharpe ratios on a historical analysis that they can completely crowd out other asset classes.
However, such analysis completely ignores the fact hedge fund returns are not normally distributed, and the downside risk may be far higher than implied by a standard deviation figure.
Indeed, returns for many alternative assets are not symmetrically distributed around a mean and investors are more concerned about the downside risk than a misleading figure for a standard deviation of returns.
Dispersions of returns can be high
The dispersal of returns among different fund managers are generally much higher in alternatives than in the management of traditional asset classes. As a result, in private equity and hedge funds, the choice of manager is far more important than the choice of strategy, and index results do not generally give any realistic representation of what an investor will actually see in their own portfolio. Hedge funds are arguably too heterogeneous to be regarded as a separate asset class at all that can be analysed using market indices.
Alternative assets apart from hedge funds are generally illiquid. This can also create significant distortions in a conventional asset allocation framework. Private equity, for example, is often and erroneously seen as an asset class that is not highly correlated with listed equity markets purely because the lag and inefficiencies of pricing means that any prices used in analysis will inevitably be out of synch with those seen in the listed markets. Private equity is better regarded as “super-charged” equity, offering potentially higher returns in exchange for lower liquidity and greater dispersion among managers.
Because there is no readily accepted method for including alternatives within a portfolio, there is no real consensus on the optimum allocation of assets given an investor’s liabilities and risk tolerances. What is perhaps the most useful approach is to undertake scenario analysis that looks at the effects on a portfolio of a wide variety of both acceptable and disaster scenarios.
The tolerance to a disaster type scenario will vary tremendously between institutions. An insurance company operating in a highly competitive environment may feel happy to take an aggressive stance if, in a disaster scenario where everyone falls off a cliff, it is the last to do so, while a pension fund may have a view that it is not operating in a competitive environment at all and seeks solely to minimise the cost for meeting liabilities with an acceptable level of risk.
In such an analysis, the characteristics of alternative assets need to be described in more detail than can be done using returns, standard deviations and correlations alone. One approach is that of Citigroup’s Vineet Budhraja and others. who suggest in a recent article1 that all asset class returns can be described by just four “drivers of return”.
These are firstly the fundamental drivers of the economy, often described as the “beta” and reflecting economic growth, credit cycles and public debt that drive other metrics such as corporate earnings and interest rates.
Four paths to return on investment
The easiest way for investors to access this source of returns is through index funds; the skill of individual managers, often referred to as alpha, which could be through buying and selling over and undervalued securities, or through taking control of a firm and adding value; liquidity - investors who purchase illiquid assets give up an option to trade out of the assets and the authors argue, should be compensated for “selling” this option and so obtain a “tradability” premium; finally, investors should be compensated for taking on the risk of a large loss. The low probability high impact and potentially catastrophic risks that reside in the downside tail of a return distribution for an alternative asset are more akin to insurance type risks, which insurance companies are willing to underwrite in exchange for a premium. Investors in alternatives should be compensated in a similar way.
When looking at incorporating alternatives, such a classification of risk and return drivers can provide a better conceptual handle on the choices that need to be made. Budhraja and his colleagues argue that individual investment opportunities must be broken down to determine what proportion of return is attributable to each return class. This needs to be combined with a methodology that can assign costs to each return class on a comparable basis. While fundamental risk is measured using standard deviation, similar metrics would need to be utilised for the other risks. Investors then need to understand their tolerance for not only fundamental or beta risk, but also illiquidity, active or alpha risk and downside risk.
Conceptual approaches such as that described by Citigroup can add value to the process of incorporating alternative assets within a portfolio. The issue for institutional investors is that it is not clear who has the skill set and the objectivity to be able to undertake the analysis and provide impartial advice.
1Citigroup Alternative Investments Journal: Summer 2006, Asset allocation: A New Paradigm, Vineet Budhraja, Rui de Figueiredo, Janghoon Kim and Ryan Meredith
Citigroup Alternative Investments Journal: Summer 2006, Asset allocation: A New Paradigm, Vineet Budhraja, Rui de Figueiredo, Janghoon Kim and Ryan Meredith