What to do about Japan?" This question has long represented one of the major dilemmas facing strategic and tactical asset allocation decision-makers. It is sure to generate as many different perspectives as there are decision-makers.

Since its peak in December 1989, when the Nikkei 225 just failed to reach 40,000, the market has consistently and conspicuously underperformed other major markets.

However, a persuasive argument can still be made for having some exposure to Japanese equities from the perspective of risk diversification. Three main factors are relevant.

The first is the size and liquidity of the market. At the end of 1989, Japan was comfortably the world's largest regional market in terms of total capitalisation size, and third by GDP weighting. Even on a country-by-country basis, Japan ranked first by market capitalisation, and second by GDP weighting (see Table 1).

Although the Nikkei 225 is now nearly 60% lower than in 1989, the Japanese stock market still has a critical mass that makes it difficult to ignore. It has now fallen to third position according to both the market capitalisation and GDP criteria on a regional basis, but still ranks second on a county-by-country basis according to both criteria. Furthermore, Japan ranks third behind the US and Germany in terms of liquidity.

Secondly, the Japanese market has low or negative correlation with the US and European markets, meaning that it can be used to offset the risks of a sudden fall in the latter (see Table 2).

On the downside is the question of the volatility of the market. Analysis of the standard deviation of each market shows that Tokyo's has seen considerably higher volatility than in New York and European countries.

Assuming the above logic is sufficiently compelling for an exposure to the Japanese market to be considered appropriate, the next question is to decide on the level of exposure.

In doing so, it is clearly essential to look at the outlook for the Japanesemarket not only on its own merits, but also in the light of all the other markets, and over different time horizons. Thus, while Japan now looks relatively unattractive compared with Europe and the US in terms of price/earnings, price/book value, price/cashflow and overall risk/ return, this may be tempered by the belief that the continued poor fundamentals of the market have been largely discounted and that, subsequent to the government taking appropriate fiscal measures, the market will recover to the psychologically important 20,000 level in the second half of 1998.

Looking at the history of the Japanese market can also be instructive (see chart). An analysis of the recent historical trend of the Nikkei 225 shows that since the collapse of the market in early 1990, it has moved in a fairly tight range of between 14,500 and 22,500. At the bottom end, 15,000 is widely regarded to be the level at which the government will be forced to provide support. For almost half of the period since 1989 the index has traded between 19,000 and 20,000. Consequently, at current levels, it can be considered to be particularly low relative to its recent history, although the upside has also been quite limited.

Whether, and if so to what extent, to be overweight or underweight in a particular sector will depend on:

q the degree of tracking error permitted within the portfolio;

q the management style or orientation (eg indexed or active, value or growth; large or small cap); and

q the short and long-term characteristics of that sector.

One of the major aspects of the market recently has been the existence of the 'two-tier' phenomenon, under which there has been a clear demarcation between the strongest performing stocks, having characteristics of strong international competitiveness, high technology, and high innovation, and the others. As a result, stocks in the high technology and automobile sectors have done particularly well. This trend is expected to continue for some time, although periodic anomalies will be apparent.

As regards the small cap effect, the recurrence of this will depend upon the speed and scale of the economic recovery. The Daiwa Institute of Research envisages low economic growth of 0.5-1% in 1998, well below what is necessary for the small caps to come to the fore.

Francis Paxton is head of marketing at DICAM (UK)"