Major innovations in weighting methodologies for equity and bond indices

What is the purpose of indices in asset management ?

Indices are, more than anything else, a barometer of the financial markets, such as the CAC 40 of French shares. Indices represent most of the different types of asset classes. They are used as building blocks for asset allocation. For example, at CPR AM, we make use of a model of multi-asset allocation strategies which cover twenty or so asset classes, such as : stocks of different geographical zones, government debts of developed or emerging countries, highly rated or high yield corporate debt, raw materials, currencies.....

Indices are not only benchmarks for index funds, the ETF (Exchange Trading Fund), but also most of the actively-managed funds..

The benchmark of a fund allows a measurement of the fund’s over performance, its relative risk (tracking-error), its performance attribution and its likely variable management fees

What is the main methodology of weighting these indices ?

Today, most stock or bond indices are weighted by market capitalisation. These indices best represent these types of assets. They do not generate high transaction costs, as the weighting is derived from the price of the securities. They are therefore, by construction, very sensitive to price bubbles during the formation period of the bubble as well as during the bubble burst periods This phenomenon was seen from the equity indices at the time of the technological bubble burst in the early 2000s and more recently, the bond indices in the eurozone.

The painful experiences resulting from bubble bursts have been the reason for new methods of weighting and have also shown the sector instability of classic indices as well as the instability of their valuation ratios and their concentration on a few possible securities

Christian Lopez

Why are we researching other methodologies of construction?

The painful experiences resulting from bubble bursts have been the reason for new methods of weighting. These events have also shown the sector instability (cf. sectoral distribution of the 500 largest european caps) of classic indices as well as the instability of their valuation ratios and their possible concentration on a few possible securities (cf. weighting of ten larger caps of MSCI Europe). More generally, investors started to look for indices which are more robust to known extreme risks (rare events of great intensity). At the same time, certain investors started to question the efficiency of classical indices both on the short and long term.

All was in place for the race to find the "best solution"....

What are the main innovations thus far ?

Today, we can identify four main families of alternative solutions.

The first family of solutions to the problem of bubbles consists of weighting the indices by micro or macro economic fundamentals. For example, the share indices FTSE - RAFE are constructed from four micro factors: (sales, cash flows, net assets, dividends) and bond indices of the Eurozone EuroMTS AAA- Macro-Weighted, weigh the eurozone countries according to their ratio of debt to GDP, their ratio of current account to GDP, quarterly growth of the GDP and long-term interest rates. These solutions are not always robust enough to withstand extreme risk

The second family of solutions to problems of robustness to extreme risks, stability and concentration consists of equally weighting the indices, either in price, in risk, or in risk contribution (Risk Parity). MSCI offers the first two for shares whilst the FTSE will provide the third. These solutions are, however, very sensitive to the world of investment indices.

A family of solutions to the problem of efficiency in the short term in very depressed markets brings together the minimal volatility/variance indices. The weightings are obtained by minimising the volatility under certain conditions. These indices depend on the conditions and the risk model used. For example, the share index MSCI - BARRA (cf. MDCI Europe in euro and MSCI Europe in minimal volatility).

A family of solutions to the problem of efficiency in the long term brings together the maximum diversification indices. The weighting is obtained by maximising a measure of diversification under certain conditions. FTSE provides two fairly close indices. (FRSE-EDHEC and FTSE-TOBAM)

How does one chose a methodology of weighting indices ?

There is no methodology which provides, at the same time, the five following properties: representative, efficiency at both short and long term, robustness to extreme risk, and low transaction costs. The indices weighted by capitalisation will still be around in the future, as they provide at least two essential properties : they are representative of the markets and they generate low transaction costs.

Nevertheless, if the investor could waive the property of representativity, interesting solutions are already available to index providers : certain minimal volatility/variance indices are efficient in the very depressed markets, such as the share market of the summer of 2011, the indices weighted by fundamentals or to the maximal diversification can prove to be efficient in the long term, certain equally weighted indices in risk contribution (Risk Parity) are more robust to extreme risk....

We believe that, in general, the choice of methodology must depend upon market conditions past or anticipated : adopt a minimal volatility index in a depressed market, otherwise a fundamental index or maximum diversification index could prove an efficient solution in the short and the long-term.