If I had been warned in 1997 that I would lose £12,000 each month for the next five years, I would simply not have believed it possible. But the unbelievable happened; and this is the price that 2,500 Names including me are paying for investing in Lloyd's of London from 1997 to 2001.

Some will have lost much more and others less, but a deficit of £725,000 is broadly the result that an average underwriting portfolio of about £1m would have produced over these unfortunate years. Many are asking how a market that was rescued from the brink of extinction in the mid-1990s could have gone off the rails again so soon?

The answer is by no means all to do with September 11 2001. The huge claims stemming from the world's costliest insurance loss had a big impact on premium rates, and everyone knows that insurance now costs more. If anything September 11, which cost Lloyd's an estimated £2bn, rescued a market that had fallen asleep at the wheel.

The reality is that the September 11 claims are less than one-third of Lloyd's total losses since 1997, which will exceed £7bn. Most claims have arisen from failed reinsurance (in spite of the regulatory systems Lloyd's has in place to ensure the robustness of reinsurance cover) and from US Directors and Officers and Professional Indemnity business.

Lloyd's has in effect been underwriting Nasdaq and New York stock exchange losses. It seems as though every American who lost money on a flotation has made a successful claim against his broker. A sizeable portion of the pre-September 11 losses also came from a string of disasters.

Lloyd's is reporting its accounts on an annual basis. It took the entire September 11 hit in one go and announced an aggregate loss for 2001 of £3.11bn. But the underlying accounts are still settled on a three-year basis. This means that, while the headline annual figures going forward are expected to show healthy profits, Names and corporate investors will be paying off their losses up to and including September 11 at least until 2004, and in all probability for far longer. A large part of the September 11 losses have been called and many Names have gone into debt to cope with the double burden of meeting them alongside the current "normal" losses.

The evidence of Lloyd's failure to understand risk is well illustrated by Lloyd's own forecasts - which are often inaccurate. If you cannot forecast your business, how can you know it? Forecasts come in three flavours - best case, worst case and mid-point - and are produced quarterly by each syndicate. It is now common practice for investors trying to work out what they have really lost to take the worst case, multiply it by two and then add some. My 1999 forecast, for example, started on a mid-point loss of some 6 per cent and then moved quarterly to 18 per cent. Even though the year closed on a three-year accounting cycle in 2002 it is still spewing out losses.

The 2000 forecast has likewise gone from an initial 5 per cent loss to 22 per cent (admittedly knocked out hard by the aviation losses from September 11, which fell into the 2000 year of account and by deterioration on earlier years) and still has some months to run. Few yet dare think about 2001. Lloyd's is forecasting a loss of some 15 per cent, but who can believe that?

Names have grown used to regarding Lloyd's forecasts as bad jokes, which they would be if they did not drop through the letter box as misleadingly optimistic notification of the next quarterly demand for payment.

For many, falling Equity markets have made it still more painful. Down at syndicate level the picture is often more disturbing. Back in 2000, one syndicate was forecasting a loss of about 15 per cent for the 1999 account. Hit by reinsurance failure, it is now in run-off and forecasting a loss of 140 per cent. This means that it lost more than three times the amount of capital supporting it.

While a majority of well-run syndicates have avoided such excesses, this is by no means the only one where something went badly wrong. However carefully they traded, most investors - either corporate or private - will have ended up on one or maybe more. Lloyd's used to conduct big loss reviews in such atrocious cases, but it has apparently lost the appetite for internal inquiries. Those left to pick up the tab think this is shameful because how else, we ask, will lessons be learned? Perhaps the problem is that any loss review would show up the failure of Lloyd's own regulatory ability.

Insurance is a high risk and a cyclical business, but the inescapable conclusion from this five-year run of losses is that, in spite of all its unique concentration of expertise and excellence, Lloyd's still has people working in it who do not know how to assess and hedge risk. And the central controls are simply not effective.

Sax Riley, former chairman, set up his strategy group in part to tackle this weakness and the result is a Franchise Board that is supposed to weed out bad underwriting before it happens. Lord Levene, the new chairman, has a new team and a hard hitter, Rolf Tolle, as franchise performance director. But I believe that if he wants to restore credibility he should dispose of Nick Prettejohn, chief executive.

Whatever his merits, Mr Prettejohn cannot escape from the facts that he inherited a market with a clean slate, freed from the nightmare of long-tail asbestos claims, and yet has subsequently presided over one of the worst performances in Lloyd's long history. Apart from failing to curtail the losses, he is blamed for being soft on corporate investors - the limited companies that moved in to provide capital alongside traditional unlimited liability Names after the 1994 restructuring.

Corporate-owned syndicates have produced some of the worst results, and some backers have simply walked away from their obligations, leaving a gaping hole of £500m in the central fund, which could yet deepen.

Lloyd's has a huge challenge ahead if it wants to persuade Names, or for that matter any investor, to trade through another soft market. The number of Names has dwindled steadily from some 20,000 at the end of the 1980s to today's hard core of 2,500, who account for about a fifth of the market's capacity.

Some are trading in the hope of recouping part of their losses and cashing in both their capacity and the value of their joint ownership of the society. Lloyd's Names are in effect co-owners of the world's second largest commercial insurer. Optimists put a price of up to 50p in the pound on their capacity - but realists know that with Lloyd's there is almost always a sting in the tail.

Next week: Lloyd's defends itself against Brian Mooney's attack.