In recent years, there has been a flurry of mergers and acquisitions across the world as markets become increasingly global. This extremely fast-changing scene has had a profound impact on investors’ portfolios – particularly their equity.
For an equity portfolio, the traditional benchmarks slice the world markets into a domestic zone and other regional components; indices are then used for each area.
However, this can create a relatively high individual stock risk due to the narrowing of individual equity markets. This is due to the typical investor’s home bias and the proliferation of ‘national heavy weights’ – successful companies that have often grown into multinationals but are still listed on their (original) home country stock exchange. The largest stock of a national European stock market typically represents some 15% to 30% of the market capitalisation (as measured by FTSE World country markets at end February 2000), with some more extreme examples such as Ericsson in Sweden (almost 50%) or Nokia in Finland (more than 70%).
Home bias also means that a heavier weight than capitalisation on world market is given to those multinationals listed in the domestic market. For example, a 50% exposure to the Euroland market (and balance spread by market capitalisation across other markets) means that 5.4 times more is invested in a FTSE multinational if it is listed on one of the Euroland stock exchanges rather than another stock exchange.
Under this benchmark, Nokia would have 13 times more weight than Motorola (2.7% as against 0.2%) while Nokia’s market capitalisation is only about twice Motorola’s.
This is rather an arbitrary over-weighting as multinationals of the same sector that happen to be quoted on different stock exchanges actually compete in the same markets across the globe.
Mergers and acquisitions can also cause disruptions in stock weightings within a portfolio using regional index framework. Using the same benchmark example, Telefonica and AT&T represent 1.1% and 0.4%, respectively, of the total (equity) benchmark. If these two companies were to merge, a US listing of the combined group would mean the new company becomes 0.6% of the benchmark. Under a Europe listing, the new company would become 3.1% of the benchmark.
The growing proportion of multinationals on stock exchanges implies that investing in a stock exchange means investing in the local economy less and less. This is a fact to be aware of, particularly if a bias towards domestic equities is adopted for liability matching purposes.
Fresh look
Recognising these issues, we conducted in-depth research with investment group Barclays Global Investors on different solutions. This was accompanied by a consultation process among the consulting and asset management industry and culminated with the launch of a new family of indices by FTSE as from 1 October 1999.
The basic idea is to strip out the multinationals from the regional indices (FTSE World indices) and group them in a separate asset class. The newly redefined regional indices are ‘local’ indices, composed of the remaining companies. A company is classified as a multinational if it generates more than 30% of its sales from outside its region (Europe, North America, Far East and the rest of the world). The local indices already created are for Europe, Euroland, UK, Europe ex-UK, US, Japan, Pacific ex-Japan and the world. Chart 1 shows the proportion of multinationals in different markets.
The traditional and new approaches are illustrated in Chart 2. Under the former approach, we assume a 50% exposure to Euroland, the balance being spread by market capitalisation. By way of illustration, we also show what the portfolio becomes if a same exposure is kept to local markets (ex-multinationals) and a multinational asset class is created (after rebalancing weightings in proportion to market capitalisations).
The removal of multinationals from local indices can substantially reduce the portfolio concentration when the domestic market is relatively narrow and the home bias high. An example is the UK market, where the average fund has a 75% bias towards UK equities. Applying a similar reallocation to the above makes the concentration in the largest ten stocks drop from 28% to 15% of the total (equity) portfolio.
For some Euroland investors, a first move to avoid concentration problems has been a trend towards considering Euroland (or even Europe) as the domestic zone. In the case of a relatively high bias towards Euroland, say 75%, the concentration in the largest ten stocks would drop from 20% to 17%. The smaller drop than in the UK case is due to fewer Euroland heavy weights being multinationals. However, this can change very fast in today’s world. The multinationals approach is naturally designed to cope in a smooth way with fast economic (vs. political) globalisation.
The rather artificial over-weighting of multinationals listed in the investor’s domestic zone, due to home bias, no longer happens as all multinationals are attributed a weighting proportional to their market value in the multinational asset class. Mergers and acquisitions among multinationals no longer cause any arbitrary change in their weighting either.
Finally, the creation of new local indices makes it clearer which economies are actually invested in it; it should also enable a better match of liabilities when this is deemed important.
Experience to date
The new approach has a significant impact on fund managers. In particular, managing a mandate relating to that new multinational asset category requires global research capabilities (as opposed to research compartmentalised by region). The selection of appropriate managers is therefore as key as ever. Our extensive manager research showed us, however, that there are a number of asset managers well placed to fulfil such mandates.
The multinational index family has generally received a very positive reaction from institutional investors. Since its launch last October, a number of pension funds, totalling assets of some E30bn, have decided to adopt a benchmark based on this approach. The trend is very likely to continue.
Note: Stock market figures quoted are as of 31 December 1999, except if mentioned otherwise.
Luc Berchem is a senior consultant at Callan Bacon & Woodrow in London:
Luc.Berchem@bwsubsidiary.com

Topics