Over the 20 years the average PE ratio of the index was 11.3 times, the highest was 19, the lowest seven and the present figure 9.7 (assuming our 1998 earnings forecast is correct). For only one-fifth of the period was the PE ratio lower than it is now. This makes the market look relatively cheap. But it does not look quite so cheap when earnings growth is taken into account.

For most of the 20 years the ups and downs of earnings growth tied in closely with the Hang Seng PE ratio except for the last two years.

The slowdown last year clearly took investors by surprise. They expected rising earnings and paid record prices to participate in them. They got falling earnings. The fact that PE ratios and earnings growth diverged so sharply whereas they had previously been so closely matched certainly suggests that something unusual has happened. The market has never before got it this wrong.

However, another way of assessing value may prove useful here. Instead of using a PE ratio (price divided by earnings), use an earnings yield (earnings as a percentage of price). Interest rates are expressed this way and so, of more importance, are returns on Hong Kong's other big financial asset, property.

Now compare this earnings yield of the Hang Seng Index with the yield on grade A commercial property (annual rents as a percentage of capital value). The result is a very close match although note again that the second chart has two Y-axes. Property yields go up and down with earnings yields but are actually lower because big investors will pay more for property when buying it directly than through the stock market.

What this suggests is that the pricing of stocks does not so much rely on changes in earnings growth as the pricing of other assets in HK dollars.

What it comes down to is that the stock market remains efficient at pricing itself relative to the fundamental forces driving it and that it probably stands at about the level it should be just now.