Trevor Petch looks at why classic theories explaining the movement of insurance shares are not working.

Bond yields go up, and share prices go down. It is a rule, because when interest rates rise, the “equity risk premium” implied in a company's price/earnings ratio falls. But from time to time, you can hear the sand grinding in the gears of the market mechanism.

For most of 1997 and 1998, stock markets were rising and interest rates falling at the same time. Among the main beneficiaries were insurance companies. When interest rates fall, bond prices rise, and up go insurance company shares, because (as any fool knows) insurance companies, and especially life insurance companies, are geared bond investment trusts. And when equity markets rise, up go insurance company shares, because (as any fool knows) insurance companies, and especially life insurance companies, are geared equity investment trusts.

And so when interest rates rise, and equity markets fall, insurance shares fall, and fall again, and that, more or less, is what explains the dismal under-performance of the insurance sector in 1999. But in the case of insurance company shares, there is a second handful of sand in the works. The interest rates up/share price down equation works theoretically because of the assumption that the risk premium attached to investing in equities remains constant. For that to be the case, the effect of the interest rate rise on earnings would have to be neutral, which of course it is not (and that is one of the reasons why the equity risk premium is not as constant as the Lex column of the Financial Times would have you believe).

For an industrial company, it may not matter very much: the increased cost of servicing debt will have a negative impact on earnings, and for valuations more elaborate than the simple price earnings ratio, it is usual to take earnings before interest charges. For insurance companies, it is a different matter. Interest rates at the beginning of 1999 were at levels barely imagined by actuaries of the 1970s and 1980s, leading to well publicised concerns in the United Kingdom and elsewhere over the servicing of life policies carrying what are now unrealistic guarantees.

Non-life companies
They are a problem for non-life companies too. Say a 10% increase in premium rates was sufficient to bring a non-life portfolio into underwriting equilibrium in 1995. The correction required at the beginning of 1999 was significantly higher. With a significantly lower yield on short-term bonds, the investment income allocated to the technical account is significantly less.

In current circumstances, therefore, a modest increase in interest rates is positive for the earnings of insurance companies and, therefore, for their price/earnings ratios. Shouldn't their share prices be going up? They might, if they were valued on the basis of their p/e ratios. By and large they are not, or not outside the United States at any rate, and at least the best diversified multinational insurance and financial services groups (which have earnings which are predictable with a high degree of probability) should be.

In the UK and Europe, the market looks more closely at multiples to net asset value or embedded value, and hence the double whammy which comes from falling bonds and falling shares - coupled, in the case of the UK composites, with lingering poor sentiment after the poor underwriting results in 1998. This brings us to a further anomaly: that there is really no such thing as a “UK composite sector” any more, because Royal & SunAlliance is the only one.

What is in the sector with it is the UK-listed half of a global insurance and financial services group (Allied Zurich), and a European life and pensions company which retains a significant interest in non-life insurance (CGU). The idea that the performance of these three companies will have any long-term resemblance is implausible, especially considering that the other two major UK non-life insurers, Norwich Union and AXA Sun Life & Provincial, are both classified as life companies and are benchmarked differently.

It will not be long before CGU could demand reclassification as a life company, which ought to be worth, in theory, about a pound on its share price. One reason will be the additional revenues from the merger of Dutch subsidiary Delta Lloyd with health insurer NUTS/OHRA. The other will be something else the market has yet to take into account: the phenomenal success of the new Commercial Union pension fund in Poland.

By the end of the year, the whole Polish non-agricultural workforce under 30, and a large proportion of those aged between 30 and 50, will have signed up into new compulsory, contributory personal pension funds. For the rest of their working lives, 9% of their salaries will be diverted into these funds, with attractive management fees to be earned. At least a quarter of this eight to nine million people will be members of the CGU pension fund - by no means insignificant, since Poland has a population the size of Spain, and will be a member of the European Union by the early years of the next decade.

Trevor Petch is an insurance analyst with Robert Fleming Securities Ltd.