Reinsurance companies who feel the future is grim enough to opt for run-off need to do so with a strategy or they risk a solvent run-off turning into an insolvent one. Philip Singer explains.

Currently, we have an interesting cocktail of a soft insurance market which looks as though it will continue for the foreseeable future, the only solution to which appears to be disaster on a global scale, and a hugely competitive reinsurance market where, in practice, only about a third of the world's reinsurance business is still available to more than two thirds of the world's reinsurance companies, thanks to consolidation and a flight to quality.

For many companies, the choices they face or will soon do are unpalatable. They could try to merge or have themselves taken over, but merging one problem company with another problem company merely creates a bigger problem company. Prospective partners will be very fussy and not everybody will be sufficiently attractive to be taken to the altar.

Absent a merger what else could they do? The obvious answer is to significantly increase capital, no easy task, particularly for a company which has been around long enough to have accumulated old year problems, such as asbestosis and pollution liabilities. What is more reinsurers which have been around long enough to have long tail skeletons in their closet now have to compete with newer competition, such as the big new Bermudian companies and new Lloyd's, who do not have old year problems. Given a choice, both the investor and the purchaser of reinsurance will surely prefer to place their business with the company with the fewest skeletons.

What else might a struggling company do? It could move into niche markets, either product or geographically based, or it could gamble: move into secondary and tertiary markets, write riskier business or reinsure with weaker security.

None of these strategies for survival, if that is what they could be called, provides a guarantee of success. Gamblers do not always win, niche markets can lose money as well as make it and getting out of the market has problems of its own. In the light of all the current market conditions, many reinsurers may decide that the future looks sufficiently grim to elect for run-off.

What happens next?
Let me pose a question. How many ceding companies make no bad debt provision for a company, any company, that is in run-off? What does history tell us? There are approximately 180 companies currently in run-off in the London market. Of those companies, 40 or so are known to be insolvent and they share one distinguishing feature; they all commenced a solvent run-off. Some of those run-offs lasted only a short time, such as Charter Re. Others took somewhat longer. North Atlantic took about 15 years to move from a solvent run-off to insolvency; British & Overseas took 28 years.

Of the remaining 140 companies currently in run-off, is it genuinely expected that they will all run-off solvently? Of course not, and that is why provisions are invariably set up for companies in run-off.

A company in run-off no longer has an income stream derived from premiums. It has to rely entirely on finite assets and investment income to meet all its obligations plus the cost of the run-off. A company in run-off typically relies on its investment income to defray the costs of run-off and/or uses estimated yields to discount reserves.

The trend for the last few years in both Europe and the United States has been for lower interest rates which lead to reduced investment income or an increase in discounted reserves, or both. Reduced yields may tempt some companies into making riskier investments.

As claims are paid, funds available for investment are reduced and investment income reduces correspondingly. Run-off costs, on the other hand, tend to increase over time due to wage inflation and price inflation, which pushes up both fixed and variable costs. Added to which will be claims inflation, so that protections that were adequate when business was written, can be inadequate by the time claims come to be paid.

The result of reducing income and increasing costs is that over time, investment income is likely to become insufficient to meet run-off costs. That of itself is not a problem, providing that the company is sufficiently well reserved to meet not only the costs of all claims, but also the costs of the run - off itself.

The size of the problem
The 180 or so companies currently in run-off in the London market, both solvent and insolvent, have estimated liabilities in the region of £25 billion. To that must be added the Lloyd's run-off vehicle Equitas with undiscounted liabilities of £16 billion, plus the run off of an estimated 50 subsidiaries of UK composites, and a further 50 or so branch operations in London of foreign parents. In ballpark terms, the estimated liabilities of London market run-off operations is about £50 billion.Restricting myself to London, I have speculated what the cost of administering it all might be. I started by trying to work out how many people were involved and added up the approximate staff numbers of all run-off management companies in the UK. I probably did not identify them all, but I probably identified most.Equitas in its annual report for the year to 31 March 1998 advises that it has about 740 permanent employees. Run-off management companies probably employ the best part of 4,000 people. To this must be added the staff employed on the run-offs of the composite subsidiaries and branches mentioned above, plus all the other people in the London market who are indirectly affected by run-off, including the employees of brokers, other insurance companies, market associations, etc, not counting firms of accountants, insolvency practitioners, lawyers, actuaries and various consultants. Probably at least 10,000 people in the UK are employed to a lesser or greater extent in run-off.

What then is the actual cost of administering London market run-off? How much is spent on salaries, national insurance contributions, pension plans, medical insurance, accommodation, light, heat, council tax, the cost of infrastructure, furniture, electronic equipment, consumables, paper, etc, services, gas, electricity, water, telephone, consultants, actuaries, auditors, lawyers, etc?

The cost of run-off per annum is unlikely to be less and probably significantly more than £500 million a year, and that is before paying a single claim or having regard for the costs of litigation.

I have also tried to form some idea of the assets currently under management by companies in run-off by extracting information from the accounts of a number of companies in run-off chosen at random. On average they hold investments equivalent to one third of liabilities.

Assuming an average rate of return of 6%, companies in run-off have perhaps £1 billion in investment income to cover costs of run-off of not less than £500 million. Some companies will achieve a better rate of return than 6% but the selection of companies in run-off referred to above have average investments yields of about 6%. It seems likely that the rule of thumb that the cost of run-off will be met by investment income looks to be right, especially in the early days of a run-off, but that has to be qualified with regards to the costs of litigation which each company bears.

On the present figures investment yields would have to halve before incomes overall match the cost of run-off, that said, the EU model is already at about 3%. As claims are paid, funds available for investment reduce so that over the course of a run-off the point may eventually be reached where income can no longer cover the cost of run-off.

I am familiar with a mainstream LMX company. Most of its transactions are processed electronically through the London Processing Centre (LPC). In the last five years it has averaged 22,000 transactions a year. The average cost of processing a transaction through the LPC is £6 not including the costs incurred in raising the entry nor of producing the information that enables the processing to take place and produce the entry, nor the costs to cedants and retrocessionaires on either end of the transaction or their brokers. It takes a lot of activity to create even a single entry, and every entry leads to several others.

The liabilities of the company are of the order of £550 million and it averages 22,000 transactions per annum. Averaged out against its liabilities the company processes one transaction for every £25,000 of its liabilities. Apply that average across the market in any one year, and £50 billion of liabilities would generate approximately 2 million transactions. If the cost of run-off to the London market is £500 million, or more, then the average cost of a single transaction is of the order of £250. So why all this concentration on transaction costs? Because those costs have to be paid for. It also means that a great deal of time, effort and capital, intellectual capital included, is devoted to the business of run-off instead of new business.

There was a time when underwriters, particularly Lloyd's underwriters could argue that having paid the broker his brokerage, the broker did all the work and kept all the records. Unfortunately the days when the underwriter could rely on the broker to do everything for no further consideration are long gone.One broker has suggested that something like 60% of the transactions its handles annually relate to clients who are no longer writing new business; in other words they are in run-off. Therefore, 60% of their transactional activities have to be paid for out of brokerage that was earned years ago and which has long been taken to profit and spent.

Some companies both in run-off and still actively underwriting have recognised the problem of transaction costs and have set materiality limits below which they will not raise outwards closings. The Equitas annual report describes their approach to saving money by reducing transaction costs. Equitas is no longer issuing collection notes unless reinsurance collection have reached an aggregate threshold of $25,000 per reinsurer, a process that has reduced the number of collection notes raised by nearly 60% while deferring collection activity on less than 10% of reinsurance recoveries. Cost savings significantly exceed the cashflow impact of deferring low value collections. In addition, it has ceased processing reinsurance transactions on a daily basis for 150 syndicates whose reinsurance programmes have a value of less than £10 million and now only process losses attaching to those syndicates' reinsurance programmes bi-annually or annually.

Equitas observes that reinsurance collection notes representing just 4% of transactions by number account for more than 95% of the total value of reinsurance to be recovered. By concentrating their energies on the high value items they achieve significant economies and without any adverse effect on cashflow.

Was Murphy right?
Elsewhere in the Equitas annual report, there is mention of the run-off taking perhaps 40 years. A similar figure has been quoted for the KWELM run-off; others are, no doubt, in a similar position. The venerable Murphy opined: “What can go wrong will go wrong”. Actuaries confirm that if you take a long enough period and/or a large enough number of events, then something is bound to go wrong.In run-off things can go wrong over time. The longer the time, the greater the opportunity for things to go wrong, and the greater the number of things that can go wrong. There are some obvious reasons. Adverse loss development over time is one of them, problems occasioned by reinsurers themselves becoming insolvent being another.

Actuaries also say that on average the ultimate outcome of a long tailed class of business generally deteriorates over time. This deterioration can occur for several reasons including:
1. The initial reserves are set based on limited knowledge.
2. Improving knowledge decreases uncertainty but allows the quantification of the uncertainty.
3. Improving medical knowledge allows the better detection and identification of medical conditions and also provides techniques that are able to directly link the identified conditions with the underlying causes.
4. Social inflation has meant that the perception of “acceptable levels of compensation” has increased over time.
5. Lawyers need to be paid!
In one case, a company wrote a small book of US casualty business in 1965 and 1966. By 1976, there having been no claims at all, the business was clearly hugely profitable. They probably had a party; they certainly destroyed all the records. The first claim came in 1982, by which time the company was in run-off. By 1996 having run off all its business other than the rump of the US casualty business and the parent having put in substantial additional capital, the company became insolvent.

The danger for any run-off that continues more than a few years is that what can go wrong will go wrong, and every time a reinsurance company becomes insolvent its cedants will have an insolvent reinsurer. A company does not need to have too may insolvent reinsurers “protecting” its portfolio before it has a real problem on its hands.

Does run-off really make sense?
This suggests that any company deciding to go into run-off, and for that matter any company in run-off, must have a clear strategy for managing the run-off. That strategy must recognise a number of factors, not the least of which is that run-off is very different from active underwriting. Another problem is that of cash flow; with no premium income to speak of, the company will be dependent upon its finite assets and its investment income.

The strategy ought to include some means of ensuring that either the retrocessionaires will pay claims on a timely basis, as and when claims arise, or that steps are taken to convert the future non-income producing asset represented by those retrocessions into a present income producing asset.

Costs need to be addressed and the strategy should be designed to drive costs down. If investment income is likely to go down in the future, then it is essential that costs be reduced as well and the sooner the better. The strategy must ensure that the run-off is concluded as quickly as possible. The shorter the length of the run-off, the lower the costs will be, and the fewer the things that will go wrong.

More run-offs ahead
The predictions are that many more companies worldwide will go into run-off. There are simply too many companies chasing too little business at too low a price. Quite frankly, it makes far more sense for a company to retire early and in good order than stay on in an adverse market to the point where it becomes so financially weakened that its prospects for continued solvency come into question. Every company in run-off represents an additional risk for its cedants and they need to find some way of assuring themselves that their reinsurer will still be there when needed.

Retiring in good order so that the shareholders get a return of their capital makes more sense than leaving things too late, but any company going into run-off must understand that life for them has changed fundamentally and it is essential that they should have a strategy; a strategy for a successful run-off which recognises the realities and doesn't simply assume that because the weather looks fine today it will not deteriorate tomorrow. There has to be a clear recognition that run-off is very different from active underwriting.

Most companies that have gone into run-off do not have, or did not have, a clear strategy and the proof of the pudding is in the eating. How many have successfully concluded their run-off? How many have paid all their claims in full, closed their operation and left a surplus for their shareholders? The answer can be counted on a very few fingers!

Philip Singer is a partner, financial advisory services, PricewaterhouseCoopers. Tel: +44 (0) 171 583 5000; fax: +44 (0) 171 212 6316; e-mail: