Dances with alpha
OUR INVESTMENT industry is currently characterised by an obsessive, compulsive mantra that generating alpha is the central purpose of investing. Producing alpha is certainly sensible and vital to your business if you’re an asset manager or hedge fund manager but is it what the client really needs?
We have spent two years debating and soul-searching at CIO level with the largest pension funds, sovereign wealth funds, asset managers, hedge funds and consultants in Europe, Britain and Scandinavia. Our intensive efforts and research have concluded what long-term investors really, really want from their investments are three key components for investment success, being: • An excess return ( eg above inflation, liabilities, benchmark, etc). •A positive
after costs. •A return without
“excessive risk”. (See illustration 1).
Note that the word alpha or skill-based excess return doesn’t feature in the above conclusion. We need to investigate each of those components further but we must appreciate that “value for the client” is represented by a triangle of needs, not just alpha.
When generating an “excess return” our industry needs to realise there are two ways to accomplish that. The first and most obvious is through skill-based investing (ie, active investing), where we hire active managers to deliver an excess return (ie, an alpha return).
The second and often underappreciated source of excess returns is excess risk. In fact, the very cornerstone of modern finance with its many derivatives (eg, CAPM, APT and new-age riskfactor models) all are based on the single belief that the majority of excess returns originate from those excess risks and not alpha. After all, the correct definition of alpha is: “The excess, residual return that remains after all returns from beta risk-premia have been accounted for.” (See illustration 2).
Only once we realise the majority of excess returns originate from excess risks and not alpha can we even attempt to use the word “benchmark”. That’s often poorly understood by active managers as well as investors. Alpha isn’t the return above a benchmark ‒ it’s the residual return that remains after the returns from ALL excess risk-premia have been accounted for. As a consequence, to measure alpha we need to know what our betas are! In other words, every active manager needs to have a risk-factor model that he’s managing against, whether he wants one or not.
A well-defined risk factor model defines explicitly which betas or risk-premia the fund is exposed to. Without such a model we can’t even measure or define alpha. The problem comes in when we ask active alpha managers to explain their risk-factor model. If they don’t have one, or don’t understand what you’re talking about, it’s very difficult for them to justify or even claim alpha. Soon, if active fund managers don’t in advance and explicitly define their risk factors, a risk-factor model will be imposed on them (see article in this supplement on “Returns-based style analysis”).
Without a properly defined risk factor model, alpha could mean anything. It’s sad but true that much excess return we term alpha is actually not “real” or “pure” alpha. Our concern is that the alpha industry is currently suffering from an identity crisis and that the term “alpha” is rapidly being lost in translation. Too many fickle interpretations of alpha have created a serious and awkward “translation fallacy” ‒ where what we think alpha is, or what we want alpha to be, is moving away from an objective, scientific and consistent definition of alpha.
Beating benchmarks without skill
Given the above, is it possible to beat a common benchmark without skill? We’ll now demonstrate two ways to achieve that. The first benchmark we’ll use is a wellknown type of strategic asset allocation or balanced fund approach, where the benchmark consists of the return calculated from a 60% weighting in equities, 30% in bonds and 10% in cash.
Much research has been conducted into the equity risk premium (ie, the excess return above the risk-free rate we obtain by taking on equity risk). If we assume the equity risk premium is around 5%/year then the easiest way to outperform our strategic allocation (60%/30%/10%) over the medium to long term is to go overweight equities, because the equity risk premium is higher than the bond risk premium (2% to 3% above cash returns).
So simply by taking on a static higher equity exposure (eg, 75% in equities) we can outperform. That’s not a skill-based decision but merely an excess return by overweighting the riskier beta or asset class. The more risky portfolio naturally should provide us with an excess return over the medium to long term if basic financial theory is sound. That’s clearly not an alpha return but a simple beta return.