The Prudential, the largest household insurer with 2 million homes on its books, took in pounds 248m in premiums for buildings and contents insurance in 1992. It made a profit of pounds 33m on the buildings side but lost pounds 2m on the contents business.
So it is not surprising that premiums are rising. But they are not rising uniformly.
Increasingly sophisticated analysis of claims is allowing insurers to make more refined judgements about the level of claims they can expect from different areas and different people.
The post-code related rating system was the first step. Then there were discounts for belonging to a Neighbourhood Watch Scheme, and money off for having a burglar alarm.
As the rating gets more and more individual, the premiums will match the risk more closely. High-risk homes will pay dearly, while those likely to see out the year unscathed should pay moderate premiums.
The problem has been that refined data has not only been difficult to collect, but necessarily historical. Times change.
Rates for the obvious inner-city areas are already high enough to compensate for the greater risk of burglaries. But mid-sized places, such as Cheltenham, Reading, Leicester and Bradford, are now being targeted and rates there are rocketing to catch up.
Apart from the obvious advice to check at least half a dozen companies to find the best rate - and to note whether terms such as the sum covered are comparable - homeowners should try to find companies offering high excesses. The more of each claim you are willing to pay, the cheaper the premiums - particularly as this cuts out all the tiny claims.
You need contents insurance to protect yourself in case of a real disaster, but you do not need to claim for every scratched table top.
BUILDING societies really must be more careful about the way they display interest rates.
The Nationwide kicks off with an awful muddle in its latest rate sheet.
For a start, the Tessa rates show a net rate. The whole point of Tessas is that they are tax free - the only time a saver will receive interest net of tax is if the capital is withdrawn before the fifth anniversary. If the interest alone is taken, it is still paid out without tax being deducted and without any tax liability.
The Tessa rates are banded into five tiers. The problem here is that no one could possibly earn the highest rate of 7.1 per cent for sums of pounds 8,400 to pounds 9,000.
The Tessa rules permit savers to invest a maximum of pounds 3,000 in the first year and pounds 1,800 in following years, with an overall maximum of pounds 9,000. As Tessas are now in their third year, this means no one can possibly have invested more than pounds 6,600.
Nationwide says the top tier is included to give people an idea of how the tiers are structured.
But if it wanted to show the top theoretical rate, it should also have shown this was currently impossible to obtain - by colour coding, a footnote or some other clue.
UNIT trusts show performance figures with income reinvested to give a picture of the total return from a fund. This makes it possible to compare a trust with a high income and another with a tiny income, assuming that dividends are immediately reinvested.
Now the FT-SE indices are following suit - from July, a daily total return figure will be given for all the main indices. The returns will be calculated gross and will assume that all dividends and tax credits are reinvested on the ex-dividend date, although investors will normally suffer a delay.
Currently the FT All- Share Index has a dividend yield of 3.97 per cent, and the FT-SE 100 of 4.04 per cent. This will be factored into the new index calculations.
Ignoring this essential part of the return from owning shares has enabled those packaged bonds offering guarantees in line with the index after a set time to deliver returns that ignore the dividend income from shares. This has been the hidden cost that investors paid for taking up these guaranteed packages.
The challenge will be to see whether anything can be devised to match up to the new total-return indices at an acceptable cost.