To calculate the expected risk and re-turn ratios of the Structured Alpha model’s three layers, Watson Wyatt has developed the ‘Structured Alpha Curve’ which considers each of the components looking at their tracking error against their added return over a benchmark established following an asset liability study of the pension fund. The aim is to create a balance between the risk and return elements.
The passive layer forms the foundation of the model, starting the curve with no added return but with a small amount of tracking error. It is the primary component in the model based on the belief that active managers seldom consistently outperform the market . While Tim Hodgson at Watson Wyatt notes that some will naturally outperform and others will underperform, transaction costs always reduce performance. “So on that basis, you can say, why not just index? And therefore, Structured Alpha has this underpinning passive layer which provides very low risk relative to your benchmark, very cheaply, with very low management risk.” The passive element can also be useful when swapping stocks with the active component, making the transition a cheaper one.
The active layer takes into consideration both active balanced managers and active specialists, the latter option being a clear favourite with Watson Wyatt in terms of adding value. The combined elements should be able to produce active returns of 50-70 basis points with a tracking error of 250-300 basis points, according to the model.
The satellite layer incorporates the smaller specialist managers who are typically running concentrated portfolios or investing in areas such as alternative investments, and while Watson Wyatt would admittedly like to increase the size of this layer, pension funds are likely only to dabble in this area. The risk levels reflect this, with a tracking error of around 600 basis points against an added return of around 250 points. Levels of diversification are decided by what Hodgson refers to as the fund’s ‘risk budget’.