Benchmarking: The Multi-Asset Class Exception

Ask any active fund manager about his benchmark, and he will have a ready-made answer. US equity managers will find it convenient to benchmark themselves against the S&P 500. European equity managers will likely invoke the Eurostoxx 50 as their main market gauge. Even fixed-income managers have plenty of indices to choose from, regardless of their location. The call is trickier for diversified fund managers, or for investors such as pension funds taking strategic exposure to financial markets as a whole, as opposed to a single asset class. In the absence of multi-asset class market indices, arbitrary benchmark choices have been made. The understanding of the performance generated by these investors is deeply affected.

A quick look at the benchmarks used by these investors reveals a particular inclination for the 50/50 or 60/40 constant mix allocation in equity and fixed income. This is partly the legacy of the twenty years up to the 21st century, when markets featured a somewhat similar asset split. The evolution of financial markets over the last ten years has made such benchmark choices little indicative of the market composition at any point in time. New multi-indexation techniques taking into account yet to be launched stock and bonds indices would allow doing it much more.

MARKET PORTFOLIO

Both fund managers and investors are at a loss when referring to the market portfolio. Even if under the CAPM theory the market portfolio is made up of an asset mix based on the market capitalization of equities and bonds, diversified fund managers are generally benchmarked against a constant mix portfolio. As such, they express their bets with respect to this constant mix benchmark, which turns out to be itself and active bet with respect to the market.

It is easy to fault the absence of multi-asset class indices for this apparent carelessness. But given the wealth of data on equity and fixed income markets enabling the construction of accurate multi-asset class indices, the investment management industry now has little excuse for perpetuating such behavior.

Given the wealth of market data now available, more realistic benchmark indices should be elaborated to benchmark diversified funds and asset allocators.

Thierry Roncalli, Head of Quantitative Research, Lyxor Asset Management

WILD SWINGS

A look at the stock/bond market portfolio, which accounts for almost all of the performance of long-term investors, shows that the market portfolio has been subject to wild swings over the last ten years. This applies to most regional markets and has direct implication in terms of performance assessment.

For instance, the weight of the equity market with respect to the entire market capitalization for sovereign bonds and equities in the US reached 89% in September 2000, whereas it was 55% in December 1987. There are also wide differences between countries. During the nineties, the equity weight increased in the US whereas it decreased for Japan. The split is also very different in Germany and France from the US and the UK. In March 2012, the equity weight was 68% for US, 32% for Japan, 51% for Germany, 54% for France and 66% for the UK. If investment grade bonds are taken into account, the breakdown is 50/50 in the US and 35/65 in the eurozone. At the beginning of 2000s, these figures were respectively 70/30 and 65/35.

These results show that it is impossible to characterize the market portfolio by fixed weights, as diversified managers and pension funds have long become used to. Doing so implies that a manager with a supposedly neutral market exposure by sticking to its 50/50 or 40/60 allocation is in fact taking a bullish or bearish view on the market, depending on the real asset split of the market. It also means that the alpha that some diversified managers may claim to have generated might be nothing else than pure market beta.

For instance, for a pension fund, a 60/40 asset mix policy was a negative bet on equity in 1999, but a positive bet today.

Thierry Roncalli, Head of Quantitative Research, Lyxor Asset Management

A BETTER UNDERSTANDING OF PERFORMANCE

Given the wealth of market data now available, more realistic benchmark indices should be elaborated to benchmark diversified funds and asset allocators in general. The implications in terms of performance analysis would be considerable.

The performance analysis of a risk parity strategy allocating amongst equity and fixed-income of developed countries using various benchmark is a case in point. For the period 1999-2011, benchmarked against the 60/40 portfolio, the annualized alpha and the tracking error volatility of this risk parity strategy come at 72 bps and 7.80%, with an information ratio at 9.2%. Benchmarked against a 30/70 portfolio, which is more representative of the traditional portfolio of large institutional investors in the eurozone, the same information ratio is 5.4%. By now, it should be obvious that a 50/50 would also result in a different information ratio. In order to avoid these arbitrary choices of benchmark, the most rational choice is to use the market portfolio. In this case, the information ratio is equal to 17.75%, highlighting the value created by this risk parity strategy.

As well, multi-asset class indices would help to better assess the performance and bets of long-term investors such as pension funds or sovereign wealth funds. For instance, for a pension fund, a 60/40 asset mix policy was a negative bet on equity in 1999, but a positive bet today.

Today, it is unthinkable to manage an equity or a fixedincome portfolio without a reference to a single asset class benchmark. It is now time for index providers to launch similar indexes representing the stock/bond market portfolio, even if this enterprise that is likely to be fraught with difficulties given the split between fixed income and equity indices providers