Scientific Beta

Financial Times: "In recent years, new forms of equity indices have proliferated and not a day goes by without a new benchmark being offered to investors. These new forms of indices are often presented as sources of outperformance, but it is clear that the alpha they provide is not a free lunch. There are two major sources of related risk."

Financial Times 18/06/2012

(...) In recent years, new forms of equity indices have proliferated and not a day goes by without a new benchmark being offered to investors. These new forms of indices are often presented as sources of outperformance, but it is clear that the alpha they provide is not a free lunch. There are two major sources of related risk.

First, risks that stem from a more pronounced “structural” exposure to risk factors, which, through their associated premia, lead to outperformance of cap-weighted indices over the long term, but which, in certain conditions, can negatively affect the performance of these new indices.

For example, by de-concentrating the benchmark of stocks, numerous alternative weighting schemes seriously increase liquidity or size risks compared to cap-weighted indices that are solely invested in the largest and most liquid capitalisations. As an illustration, the small-cap bias pays off in the long run, but it can turn out to be dramatically penalising in some periods. In the same way, weighting stocks according to their economic size, without reference to market price, leads to the creation of sector and style biases that can end up being dangerous. The recent sovereign debt crisis has clearly shown the limitations of these approaches that over-weighted large and highly profitable financial corporations.

Second, every weighting scheme, whether it is qualitative or quantitative, corresponds to a choice of model and therefore contains model risk. Even if financial research allows one to hope for a certain robustness in the models used, the fact remains that every index construction method relies on assumptions and meets what academics modestly refer to as conditions of optimality, but what practitioners understand as limitations to the promise of outperformance.

Ultimately, every new form of index has a serious likelihood of underperforming its cap-weighted counterparts. Recent research shows that the main alternative indices have considerable relative drawdowns (peak to trough moves) with regard to their cap-weighted counterparts. These drawdowns can be long (more than two years) and significant (more than 13 per cent).

To the best of our knowledge, no alternative form of index is immune to a significant risk of negative underperformance. This does not condemn the new weighting schemes, because over the long term most of them do better than the cap-weighted indices, but it should lead the promoters of these indices and investors to think about best practice in the area of documentation and the management of the relative risks.