Welcome to ‘beta prime’ – the ‘new’ style of indexing
Some bold claims have been made on behalf of fundamental indexing. "I am not suggesting they should entirely replace traditional cap weighted indices," concedes Rob Arnott, principal of US based Research Affiliates, " but our indices offer better returns for low volatility."
These claims have persuaded some pension funds in Sweden and the US, not to mention a major firm of investment consultants, Watson Wyatt, that fundamental indices offer added value. Watson Wyatt has even coined a neologism, "beta-prime" to describe this "new" style of indexing. Yet by the most generous estimates, to date, only some US$17bn (€12bn) is invested against the fundamental indices from two main providers, Research Affiliates and Global Wealth Allocation (GWA), and the smaller US-based Wisdom Tree.
This money is invested in exchange traded funds (ETFs) and other products that offer the index providers a licence-fee based income stream. These ETFs generally cost substantially more than the large, cap-weighted, pooled index tracker funds available from the likes of BGI, State Street Global Advisors and Legal & General. Tracker funds investing in large cap liquid indices, like the FTSE 100, are low cost with annual management charges reduced to a few basis points.
By contrast, the lowest cost ETF investing in a fundamental index boasts annual management costs of 18 basis points. Most of these products cost much more, with annual charges in the 40 to 60 basis point range. Cost is the chief reason why ETFs are not used as core holdings by pension funds.
Those advocating fundamental indices do so by comparing index returns, without deducting implementation costs. These costs may wipe out much, if not all, of any relative out-performance generated by fundamental indices. Rapid growth in the number of available ETFs over recent years may also help to explain why fundamental indexing has come to the fore.
ETFs have created new commercial opportunities across a wide range of index replicating strategies which were previously not commercially viable. This business is driven by burgeoning licence fees and there is no shortage of providers prepared to add fundamental index replicating ETFs to their existing product range, albeit at a price to the investor.
There is nothing really new or original about the valuation metrics used in fundamental indexing. They are already in wide use by active managers. In all cases, the index providers rely on the availability of much broader, cap weighted indexes from which they select or rank sets of equities using valuation metrics. Indeed both Research Affiliates and GWA are tied to FTSE International and use their cap weighted indices as references.
The key point about fundamental indexing is that sets of valuation metrics are applied consistently via the fundamental indices. In the case of Research Affiliates, four fundamental metrics are applied to each equity. These are the past twelve months' sales, cash flow, dividends, and the latest book value. GWA uses only three metrics. without considering dividends. Research Affiliates smooths the data produced by the application of its metrics to reach a five year arithmetic average for sales, cash flow and dividends, but not book value. GWA does not use any smoothing. To ensure consistency, each valuation metric in each set of metrics is given equal weight. The only exception to this is when a share pays no dividends, in which case Research Affiliates equally weights its remaining three metrics.
However there are some very important differences in the way that Research Affiliates and GWA construct their indices. Research Affiliates selects a sub-set of all the equities listed in the relevant cap weighted index. For instance, it selects 1,000 equities for the FTSE RAFI 1000, from the 2,546 equities that comprise the reference index, the FTSE US All Cap. By contrast, the FTSE/GWA US index, includes 723 equities, the same number as in the reference index, the FTSE Developed US.
Research Affiliates reflects the full capitalisation of a company, while GWA adjusts to reflect the free float capitalisation. These differences have major consequences for the estimated annual turnover in real portfolios using each type of benchmark.
According to State Street Global Advisors, RA indices have a likely turnover of 10 to 13%, while GWA attain 20 to 25%. Rebalancing at RA takes place only once a year; GWA every quarter. The transaction costs of running a GWA portfolio are therefore higher.
he initial claims made on behalf of fundamental indexing by Arnott and others involved an attack on the conventional wisdom that cap-weighted indexes are optimally efficient portfolios.
In 2004, Arnott, Hsu and Moore, published their paper ‘Fundamental Indexing', which is still available on the internet. This paper argues that the Capital Asset Pricing Model has misled many to champion cap-weighted indices as mean-variance optimal. It argues that cap weighted indices overweight overvalued stocks and underweight undervalued stocks. The intuitive argument in favour of cap weighting is simple; a portfolio manager with no ideas has to accept the market value put on each share. But the notion of an optimally efficient portfolio goes further; it says that return is maximised for each unit of risk in the portfolio. The use of fundamental metrics carry an implicit argument that cap weighting is less efficient than fundamental weighting.
The used of fundamental metrics has been at the core of active and particularly value orientated management for decades. "Systematising their use through a fundamental index appears to remove "subjective" elements in active value management; it ensures that the weight given to each metric is constant" argues Richard Hannam, head of global structured products at State Street Global Advisors.
One of the justifications for such an approach is an assumption that these metrics, divorced from share prices determined by the market, better replicate the fair economic value of a share than market capitalisation. This implies that the use of the relevant valuation metrics will be more successful than market capitalisation at predicting the future wealth generated by the relevant companies. On this basis, a fundamental index may consistently outperform a cap weighted one as
long as the fundamental index is selecting eligible stocks from the cap weighted index. The same effect is clearly visible when the average returns from active managers are measured against the returns from some cap weighted indices selected by those managers as their benchmarks. These claims are the subject of debate and have not been widely accepted.
Instead, we are seeing the emergence of less ambitious arguments for fundamental indexing. These accept that the relative outperformance demonstrated by fundamental indexing may be cyclical and more consistently exhibited in selected areas of an equity market. In particular, fundamentalists tend to choose fairly broad, diversified cap weighted indexes as reference indexes from which they use fundamental metrics to select undervalued stocks. There is nothing new about this.
"Active managers tend to do the same," observes Bernard Nelson, principal at Mercer Investment Consulting. The broader an index the larger the set of mis-pricing opportunities to be exploited and the easier they are to outperform. If fundamental indices also demonstrate cyclicality then they start to look very much like active management strategies.
A far stronger argument in favour of investing in cap-weighted indices is that in the long run, over twenty or thirty years, it may make very little difference whether one chooses cap-weighted or fundamental.
"The differences may be less than supposed," adds Hannam," the undervalued stocks identified by fundamental metrics end up being overvalued by the market, rewarding the investors in cap weighted indices." This leaves one quite strong argument in favour of fundamental indexes.
Instead of replacing their cap weighted counterparts, they can be used in conjunction to diversify core equity holdings along with other strategies such as enhanced indexation. This argument is very close to those put forward for combining value and growth styles in actively managed equity portfolios.
One reason why consultants may be reluctant to recommend fundamental indexing is that they regard it as nothing more than a systematisation of the valuation metrics used by value orientated managers. If investment consultants believe that they can successfully identify active managers capable of consistently outperforming fundamental indexes on a net-of-cost basis then there is no reason why they should recommend fundamental indexation strategies. The hard evidence on this is mixed and this picture may change over time if the cost of investing against fundamental indices can be reduced to the same levels as that of investing in cap-weighted tracker funds. There are other less obvious reasons why fundamental indexing may not get the attention it deserves.
For one, pension funds have other more pressing matters to consider. Defined benefit pension funds are shifting wholesale to the use of liability based benchmarks, which place an emphasis on absolute rather than relative returns.
The past few years have also seen a huge emphasis on non-core diversification into alternative investments, which are intended to display little correlation to the returns from traditional asset classes such as listed equities. Fundamental indices do not offer diversification of this kind. Instead, they offer a relatively expensive means of diversifying core equity holdings. At a time when most defined benefit schemes are reducing their exposure to core equity relative to other asset classes trustees and other fiduciaries may see no pressing reason to convert to fundamental indices.