Many large financial houses have introduced regional and/or world convertible indices. American brokers have been particularly active in constructing a world index, while the not unimportant European and Japanese players have so far adopted a relatively modest stance. World convertible indices are often put together with the help of sub-indices created earlier.
Understandably, given the leading role played by the US since 1980, all the indices introduced were initially American. The first institution to publish such an index was Credit Suisse First Boston in 1982, followed by Goldman Sachs (1985), Merrill Lynch (1987), Jefferies & Company (1993) and the European Warburg Dillon Read (1994).
The only house that ultimately also produced product (sub-)indices was Merrill Lynch. This met the needs of institutional investors who were increasingly making use of “convertible-like” instruments issued primarily in the US. Examples include “variants” of the mandatorily convertible bond and preference convertibles. Various international players are also increasingly creating sub-indices that draw a distinction between, for example, creditworthiness and the average conversion margin.
Today, a number of the financial institutions referred to also publish indices for Japan or the Far East. Almost all the US and European houses referred to have such sub-indices in addition to indices for Daiwa, Nikko and Nomura, for example. It is important to note that many draw a distinction for Japan between domestic and euro-loans on account of the considerable differences in marketability.
ING Barings was one of the first to launch country indices for Southeast Asia in 1995. Among the countries covered were Hong Kong, Indonesia, the Philippines, Korea and Taiwan. These sub-indices are no longer available to investors.
The European convertibles market finally followed during the 1990s. Driven by the ongoing dismantling of traditional ties in Germany, France and southern Europe among others, as well as the approaching monetary union and the impending introduction of the euro, there were a large number of issues on the European mainland in particular. In addition to western European convertible indices, EMU and Eastern European sub-indices have recently been launched by parties such as Deutsche Morgan Grenfell (1999), Goldman Sachs (1999) and Merrill Lynch (1997). The introduction of European indices often coincided with the publication of the world convertible index of the financial institution concerned, since these indices formed the final group in the series of regional introductions. Table 1 shows the various indices, while Table 2 shows the rules used in constructing the world indices. For the sake of clarity, the remainder of this article will focus primarily on the world convertible indices of Goldman Sachs, Jefferies, Merrill Lynch and Warburg Dillon Read.
Table 3 shows the main assessment criteria which must be met by a convertibles benchmark if it is to be suitable for institutional investors, regardless of the method of management (active or passive). Important characteristics such as the regional, country and sectoral allocation of the various world convertible indices are relevant here. The following section of this article tests all these elements.
In terms of number of loans, the Warburg Dillon Read index provides the largest percentage coverage of the world convertibles market. The total number of outstanding loans throughout the world is around 4,000, and the index thus covers around 10% of the total market. In terms of market capitalisation, however, given the present size of the market at around $425bn, Merrill Lynch and Warburg Dillon Read cover around 33% of the market, compared with something over 30% for Goldman and Jefferies. Tables 4, 5 and 6 compare the index-allocation at various levels with the market data.
There are strikingly large differences in terms of allocation for Europe and Japan, in particular, between the Merrill and Goldman indices on the one hand and the Warburg and Jefferies indices on the other. The extremely high weighting for Asia ex Japan in the Jefferies index is also noteworthy. It should be noted here, however, that this index focuses on the needs of a specific group of clients. The very strict liquidity standards set by Warburg Dillon Read compared with the other parties are important in this context. The fact that the Jefferies index is an equally weighted index rather than a market-weighted index is also important given that it uses a different method from the competition to construct its world convertible index. Jefferies initially compiled an index ex US, to which the US was added in 1998. Re-indexing to 1993 was then carried out. The Goldman and Merrill indices are reasonably representative as regards the regional allocation of the world convertibles market.
Only Jefferies, Merrill Lynch and Warburg Dillon Read publish a country allocation. In addition to the deviations already noted for Japan and the US, other striking differences are the percentage weightings accorded by the different parties to a large number of European countries, Hong Kong, Korea and Taiwan. The average size of each index loan is approximately $500m at Warburg; this compares with $175m for Merrill. Given the knowledge that Jefferies selects loans only from an investable universe, it is likely that this index too includes relatively large loans. The fact that large loans have been issued mainly in the European countries named explains the difference in weightings. This applies only partially to Asia and Japan. The Warburg country allocation offers the best opportunities for reproducing the country allocation of the world convertibles market in an investment portfolio.
The three companies that publish a sector allocation are Jefferies, Merrill Lynch and Warburg Dillon Read. Considerable deviations from the market can be found at sectoral level in services (Merrill and Warburg), Materials (Merrill and Warburg), financial institutions (Merrill and Warburg), energy (Jefferies and Merrill) and consumer goods (Jefferies). This can be largely explained by the regional/country differences already referred to. It can be concluded that the three indices are not entirely representative of the world convertibles market.
It is a known fact that many institutional investors, pension funds, insurance companies, etc are only permitted to invest in securities with a high creditworthiness. An official credit rating from a generally reputed rating institute is almost always included in the investment rules. It is therefore an advantage for several parties if the benchmark can be broken down into various rating classifications. Table 7 provides a summary of the possibilities for each index.
Merrill, Goldman and Warburg offer full or only partial possibilities in this context for investors as regards allocation, calculation of the result, reports and the compilation by investors of their own index. The Goldman Sachs index was introduced only recently. Goldman has indicated that it may develop a number of relevant indices for the other regions in the future. Merrill does not have an allocation at world level based on financial quality. The conclusion is therefore that the Warburg Dillon Read index is the most complete, with the proviso that for Europe and the emerging markets in Southeast Asia, the index includes only loans with an official rating.
In addition to creditworthiness, marketability is an important element in the investment rules of many institutional investors. Marketability not only means a regular turnover at realistic prices, but also that the convertible in question has a minimum amount during and after issue. Investing in an index loan must therefore be easy. Compared with the world equity and bond markets, the world convertibles market is small. According to figures from Morgan Stanley Capital International, the size of the world equity market is currently around $23,445bn. The various indices used in the comparison all operate a minimum amount for convertibles both on issue and whilst in circulation. In practice, there is a danger of running into certain liquidity limits. One positive factor is that more and more multinational and transnational companies are issuing large – and presumably more liquid – convertible loans. In our view, it would be advisable to tighten up the liquidity criteria somewhat in the future. Ultimately the Goldman and Warburg indices come out best on account of the criteria they employ.
When investing in convertibles it is also important that an active duration policy can be pursued. Given the hybrid nature of the instrument, a change in short-term and/or long-term interest rates can have implications for the price. It is known that the Goldman Sachs Japanese sub-index has a duration requirement of at least five years. No special requirements have been set for the other indices. All indices do however have rules relating to the (early) redemption of loans. In a nutshell, very short durations do not occur. Ultimately only Merrill offers the possibility of breaking down the regional sub-index for Japan by duration (and then only for convertibles issued in Japan and denominated in Japanese yen). Jefferies does this for the world index, but only on request.
As we saw earlier, there are considerable differences in the sectoral structure of the two world indices for which a sectoral allocation can be analysed. The Merrill, Warburg and Jefferies indices can be broken down into 12, 20 and 25 sectors compared with the eight indicated in Table 6. It is also important to note that, for the US sub-index, Merrill allows an allocation to be made by company size (market capitalisation) and type of company. Jefferies does this for the world index, but only on request. Of importance is also the fact that Warburg has the possibility to take more than one convertible of one debtor into the global index.
Reinvestment of coupons
All indices considered here are calculated on the basis of total return. In addition, Jefferies and Goldman can supply a price index figure to interested parties. Current interest rates are included on a daily basis by all parties. Released interest is reinvested in the loans concerned.
Merrill and Goldman update their indices once every six months, while Warburg carries out updates every quarter. This does not apply for Jefferies. In general, changes are implemented at the end of the month, with the exception of additions under the Goldman Sachs ‘fast add’ rule. Jefferies adds new loans on the coupon date. The conventional rules are taken as a basis for the settlement of the loans included in the indices. Exceptions to this are Goldman and Warburg, which do not apply the T+3 rule, but the T-rule. The remaining construction rules shown in Table 2 indicate that there are clear procedures for the other indices regarding inclusion, deletion, etc. Merrill, Goldman, Warburg and Jefferies present their indices in a range of currencies (multi-currency), making it easier to use these indices as a benchmark.
Valuation and index level
All indices included in the study are available on a daily basis. All indices take the market price or issue price as a basis. Reference was made earlier to the liquidity requirements set, which can be regarded as realistic for the convertibles market. Non-liquid loans have been or are excluded as far as possible. All loans in the various indices meet the set liquidity requirements.
Table 8 accords a score to the four world indices, both for each individual element and for the index as a whole.
The table shows that the Merrill Lynch index ultimately achieves the best score. However, the Goldman Sachs index was introduced only recently. We will have to wait and see whether this financial house ultimately proves capable of a comparable performance.
Jacques Grubben is head of research and Hans van Summeren manager of convertibles at Insinger Asset Management in the Netherlands. A fuller version of the study is available from email@example.com