Alternative Risk Transfer (noun, abbr: ART). Like non-life insurance, ART is best described by saying what it isn't. Indeed, ART is sometimes described as `any non-traditional risk solution' (although this inherently equivocal definition has the whiff of marketing about it, and thus invites only cautious cynicism).

To complicate matters, ART moves with the times. When captives were new they were ART. Now they are the accepted way for reinsurers to pilfer insurers' largest clients - and are alternative no more. Likewise, loss portfolio transfers were ART when they were new. Now most are simply bad decisions, at least from the point of view of the transferee. Thus for most insurers with a portfolio stuffed full of asbestos exposure, an LPT is no longer an alternative.

So what is ART? Aon's seemingly extinct CatEPut construct definitely qualifies. They guaranteed the reinsured that a specified number of its shares would be purchased for a fixed sum if the put option was triggered by a loss. Aon broked at least three of the things, including one between La Salle Re (since dismembered by Trenwick), Allianz and Swiss Re. The latter's recent unsuccessful attempt to wriggle out of its commitment to buy $55m worth of equity in troubled Trenwick is understandable, but the Swiss had no alternative.

For insurers too, the CatEPut was something less than an alternative solution (there's that marketing jargon again...), since traditional reinsurers (Centre Solutions and Swiss, the same old markets) were the only apparent sellers of the options. US insurer Horace Mann bought a three-year CatEPut from Centre in 1997. Then, in September 2002, it did another contingent equity deal with Swiss Re, this time with a conventional quota-share option thrown in. Why were there no alternative transferees? The pot-of-gold money managers known jealously to reinsurers as `the capital markets' didn't bite.

So-called `Act of God' bonds did get some attention from the capital markets. Alternatively known as `cat bonds' (and derisively as `pussycat bonds' due to their sadly predictable failure to revolutionise the reinsurance sector), most big reinsurers have some cat bond retro from this seemingly alternative source. However, the mantra of uncorrelated risk has not set the investment community on fire (even though cat bonds issued by USAA's special purpose vehicle Residential Re in 1997 were said to be the only decent asset owned by defunct Long Term Capital Management). The simple fact that USAA's $476m market-priming deal has not been exceeded to date hints that cat bonds are no revolutionary alternative. The capital market's trillions were not put on the table, so a large chunk of the sputtering cat bond market has been sold to traditional reinsurers, usually at prices exceeding open-market cover.

ART can be stretched to include all kinds of nifty deals. Remember the Chicago Board of Trade's indexed catastrophe futures? Like cat bonds, they were supposed to send insurance risk to the capital markets (although they seem to have done nothing at all). However, some visionary reinsurers developed a particular appetite for accumulating capital market risk on their own balance sheets. This child of convergence (cleverly dubbed `insuratisation') encompasses whims such as credit enhancement deals, residual value insurance, weather swaps and the now notorious film finance cover.

Insuratisation has certainly proved a welcome alternative for bankers, who briefly could offload their most fearsome risks onto (sometimes naïve) insurance risk markets. Alas for reinsurers, it has proved only an alternative source of loss. Witness Centre Solution's pre-Christmas withdrawal from the credit wrapper market (accompanied by a pledge to focus on `core insurance-based products'). Consider Swiss Re's third quarter announcement of a SFr379m loss from its Financial Services Business Group (of which SFr338m was lost by its `Credit Solutions' division).

The real alternatives, it seems, are few and far between. Finite reinsurance is ART, but it is rarely positioned as an alternative. Often such contracts are presented in this way: "You have no alternative. Buy some finite, or we're off your programme." Nor, for reinsurers, has finite re lived up to its agonisingly non-descriptive nomenclature. Some of the biggest finite writers are seeing portfolio losses that are not yet stopping. The business has proved less ART and more junk science.

Happily, the developing realisation that ART is singularly unattractive is actually very good news. Reinsurers can forget the alternatives, and go out and write some good old-fashioned reinsurance treaties. They're clamouring for it.