The recent inquiries into accounting practices at AIG have caused a great deal of press comment about finite reinsurance, not all of it accurate, and so we thought that our readers might be interested in a little more background to this type of transaction from the old greys who make up TGP. We will not, incidentally, be looking at the question of how AIG managed its smooth upwards progression of earnings year on year, in what is a notoriously cyclical business, nor the analysts’ lack of comment on this phenomenon.

At the syndicate’s closing date two major variables remained: provision for IBNR and likely future investment income on established reserves. The former was traditionally dealt with by reinsuring into the next underwriting year (known as ‘reinsurance to close’). But whereas a regular insurance company could wait to see how the investment income turned out and reflect this in its ongoing trading figures, Lloyd’s syndicates were forced to declare a fixed point result, and few, if any, had the investment expertise to manage the difficult matching problems that would arise with the running off of each discrete year. Of course, the syndicate underwriter could choose to deal with this by discounting the reserves for reinsurance to close by anticipated investment income, but this would push added uncertainty onto the reinsuring year.

A way around these difficulties was to arrange what was called a ‘time and distance treaty’ with a reinsurer. In practice the syndicate’s reserved losses would be transferred to a reinsurer for a premium of something less than the total amount of the reserves. This premium would recognise the expected future investment income that the premium would produce, as well as the reinsurer’s profit requirements. The difference between the amount of the reserves transferred and the premium paid to the reinsurer was thus made available as profit for payment to the syndicate members. A feature of such arrangements was that the maximum amount recoverable from the reinsurer could not exceed the stated amount of the reserves transferred to it. This was because reinsurers would not pick up the under-reserving risk, which had to be dealt with as part of the reinsurance to close.

It was sometimes argued that this type of reinsurance was ‘risk free’, but it was not. In calculating the premium the reinsurer would take a view on the likely timing of claims payments and organise a matching investment programme. If actual payments were made more quickly than anticipated the reinsurer would not receive the amount of investment income anticipated. A problem could also arise if sufficiently long-dated securities were not available to cover the whole anticipated programme period, leaving the reinsurer exposed if coupon rates for later issues turned out to be less than expected when the premium was fixed (known as the reinvestment risk). In time reinsurers would seek to mitigate this exposure by specifying, in the treaty wording, the actual timing and amount of payments to be made, irrespective of the actual underlying claims payments pattern.

It will be noted, in the case of a time and distance treaty, that arrangement of the reinsurance permitted the freeing up of investment income for profit with no ongoing risk to the cedant. It was also observed that the ability of incorporated insurers to run reserves forward from one year to the next made similar arrangements unnecessary for them.

But accounting regulations vary across jurisdictions, some would allow insurers to discount reserves and others would not. An insurer not allowed to discount reserves could achieve the same effect via a finite risk reinsurance programme in a similar way to that described for Lloyd’s, and thus present its results in a more favourable light. Of course, if the reinsurer’s accounting treatment matched that of the cedant, it would book a loss in the year that the business was written, because the losses transferred to it would exceed the premium paid, and the true profits would emerge only much later. Writing a number of such deals, year after year, would risk damaging the reinsurer’s balance sheet to the extent that it could be forced out of business. One response to this was to locate such reinsurers in jurisdictions that would allow the recognition of the ultimate profit at a much earlier stage.

As transactions became more sophisticated multi-year contracts were developed. Such a deal might purport to take a large volume of liabilities off a cedant at a comparatively modest premium, producing a substantial profit for the cedant in the year in which the reinsurance was arranged. But the treaty would also require the cedant to continue to pay premiums to the reinsurer until the reinsurer had received all its money back plus a profit. Not only would such arrangements boost the profits of the cedant but, more dangerously, they could also make its solvency position appear far better than it actually was, or even make a technically insolvent insurer appear solvent.

Again there was scope for an accounting mismatch, because the cedant might book only the first year’s premium against the whole recovery, making no provision for subsequent years, whereas the reinsurer would book the full premium due over the multi-year period. It is not too difficult to imagine how the skilful use of such arrangements could mask volatility in the results of an insurer, as long as the industry downcycle roughly matched the previous upcycle.

Next Page