Alpha-Beta Soup

It's all in the mix

Originally published in the February 2008 issue

Two major trends in the asset management industry have been the growth in assets managed within index funds and also assets managed within hedge funds. These apparently polar opposites are in fact part of a common movement that seeks to meet the increased demand from investors for transparency, value-for-money and diversification in their sources of return.

All investment trends carry along with them a corresponding jargon. The central definitions here are of ‘alpha’ and ‘beta’:

  • Beta refers to returns that can be obtained by passive market exposure, typically investing to match the return on an index;
  • Alpha refers to returns that are obtained as a result of investment skill independent of market movements, ie. ‘active management’.

In this article, we will explain the origins of this strange terminology, examine some of the reasons for the trend towards separating alpha and beta management and suggest some ways in which the market may develop further.

Finance in a state of regression (a technical aside)

Regression, as the many readers of Freakonomics will know, is a mathematical technique for teasing out the hidden relationships between two (or more) data sets. It amounts to plotting the two data series against each other in a scattergram – each pair of values make up the x-y co-ordinates of a point on the scattergram.

The regression then amounts to determining whether (in a statistical sense) the resulting points tend to lie near a straight line. The slope of this line is known (almost always!) as the beta co-efficient and the vertical position of the line, as the alpha co-efficient. (It is strangely common for mathematicians always to use the same letters in a particular context – it eases communication and goes back at least to Descartes’ use of the ‘x’ and ‘y’ axis in the 1630s). In a finance context, the returns from a fund can be regressed against the returns from the market over the same periods. If the fund matches the market, the points in the scattergram will lie on a straight line: the two coordinates are identical. Thus, beta is one and alpha is zero. A fund that consistently beats the market by say 0.5% per annum, will have a beta of one and an alpha of 0.5%. A ‘market-neutral’ fund will aim to have a beta of zero.

Distinguishing alpha and beta

Since achieving a market exposure is mechanical in nature, beta returns can be obtained without specific investment skill or knowledge. In contrast, beating the market requires skill and truly skilful managers should be able to command premium fees for their ability to beat the market. This is one reason why it is important for investors to be able to clearlydistinguish alpha and beta returns.

Regression appears to make finding the market component of a fund’s returns easy. In practice though, the program described above is overly simplistic. There is not a single market index – rather, there are a large number of recognised market subsets (eg. small-cap stocks, growth stocks). Regression has to be run against these many possibilities to illuminate the true beta components of a portfolio’s return and thereby identify the residual active-management component.

Separating the alpha from the beta

It is one thing to distinguish the alpha and beta, but fund managers have gone further. A traditional active manager would manage a portfolio against, say, the S&P 500 index and look to generate active returns over and above the index. An investor who wanted to access the active returns had to be willing to take on the underlying S&P 500 market exposure. But if the investor really wanted index-linked-gilt market exposure he faced a quandary: live with the S&P 500 market risk or forego the active returns from equities? Portable alpha and ‘pure’ alpha products are part of the industry’s answer to this problem.

The basic idea of portable alpha is that the original market (beta) exposure embedded within a fund (S&P 500 in this example) is removed and the investor’s desired market exposure is bolted on instead. Beta returns, as we have said, are usually relatively cheap to implement, and indeed changing the beta exposure can often be most efficiently achieved using derivatives. In this way, the investor can capture the market exposures (betas) they want with the active manager returns (alphas) from wherever they believe they are most likely to be found.

Hedge funds often claim to be pure alpha. Alpha returns are the result of the portfolio expressing the manager’s active views. If portfolio constraints are relaxed (chiefly, the long only constraint that limits a manager’s ability to express a 2-way view on many investments), it is usually possible to build a portfolio that picks up all the returns from the manager’s active views without having to take on the exposure to the broad market. The market exposure can be minimised or even eliminated by using long/short pairings or by using derivatives, as in a portable alpha strategy. The industry trend therefore seems clear: traditional funds mixing alpha and beta are being unbundled.

The mixing challenge

Unbundling alpha and beta so far appears to be very positive; investors are better able to assess what they are buying and higher standards are required from managers.

The problem is that the range of possible strategies has suddenly become enormous. Now, the investor needs to determine their desired market exposures and skill exposures – preferably in the context of their liability exposures if they are a pension fund. Determining the appropriate mix often requires advice, which could come from several sources; and there is a large investment consulting community which aims to do just that. Fund managers themselves however, are engineering mixes of the different sources of return. Since it is they who disaggregated them in the first place, they are well placed to quantify the efficiency gains of a ‘better beta’ mix compared to the original indices in each separate market. The resulting funds are usually given names that include the adjective ‘diversified’. Other mixes are also on the rise. Multiple alpha strategies are also mixed – into ‘multi strategy’ funds. And, in keeping with the portable alpha concept, a broad-based mix of active returns and diversified market returns can lead to a portfolio with better overall characteristics.

Beta to alpha and back again

These different betas may not always be easy for investors to access. For example, the data required may not be widely available, or implementation may require portfolio gearing. Historically, active strategies may have involved seeking to outperform the market by putting a consistent bias in their portfolios towards a particular source of return (eg. a ‘value’ tilt). Where this style is easy to quantify and implement, competition will naturally drive the fees for implementing such a strategy towards indexation levels.

As new instruments or techniques become available, the range of narrow market exposures that can be easily implemented continues to expand. For example, by using derivatives it has been possible to construct strategies that access the premium available for market volatility, without being dependent on the level of the market. As the cost of these instruments falls and increasing numbers of managers are able to deliver the strategies, the skill level falls and fees follow suit. These strategies, which would previously be classified as alpha, thus begin to make the transition towards beta.

As the range of possible beta strategies increases, investors are better able to analyse how outperformance is being generated. The transparency is clearly in their, and the market’s, interest. The lesson for active managers is of course that they need to continue to innovate if they are to outperform the market, and earn the fees that they would like to command.