The agreement of General Re to its acquisition by Berkshire Hathaway may signal another fundamental change in the structure of the reinsurance market. Michael A. Smith analyses the implications.

An old friend, himself considered a sage among institutional portfolio managers, once quoted an early mentor as saying: "A paradigm is worth twenty cents." However, the new course that the General Reinsurance Corp. is about to embark on may be worth much more than two dimes. Of greater importance, it has significant implications for the rest of the reinsurance markets, and the ability of reinsurance companies to remain independent going forward. It may well be that the other publicly-traded reinsurance companies will also have to find larger partners, as did General Re.

As the second quarter of 1998 began to draw to a close, the management of General Re astounded investors, clients and competitors with its announcement that it had agreed to be acquired by the mammoth Berkshire Hathaway, run by the legendary Warren Buffett and already the owner of GEICO and National Indemnity. The initial reaction to the announcement was that the deal made good business sense from General Re's perspective, and indeed, the company's chairman, Ronald Ferguson, made the point that his company would now be in a position to take on increased risk without having to bear as much retrocession cost as had been the case historically.

From Mr Buffett's perspective, General Re will provide enormous cash flow that he can use to add to his extensive investment portfolio. Until this year, much of that cash flow had funded General Re's aggressive share repurchase programme. (See Table 1.)

But looking at the transaction again from General Re's point of view, one has to ask why the ability to retain more risk is worth giving up the company's independence. After all, the retrocession market is reported to be even more competitive than the reinsurance market, suggesting a favourable arbitrage opportunity for companies such as General Re. And on the longer term, if and when the retrocession market tightens, it is unlikely that General Re would have difficulty obtaining capacity from its so-called "pool X" and "pool Y" rosters of underwriters that it cedes to. And, in the grand scheme, General Re's net reinsurance costs have been small and declining, especially when claims recoveries are taken into consideration.

Return to fundamentals?

Perhaps Mr Ferguson was letting on to something of far greater significance when he stated that, under the umbrella of a much larger owner, he will be able to assume greater volatility in his earnings without the distraction of having to deal with such shareholder concerns as the reduced stock valuations that might result from increased volatility. It is a reasonable assumption that this statement alone may signal General Re's intention to return to one of the company's, and the industry's, very reason for existence.

One of the early lessons in an insurance career is that reinsurance serves two basic purposes:

(1) It serves as a source of capital, allowing insurance companies to write more risks or larger risks than their own capital can safely underwrite; and

(2) It allows an insurance company to smooth earnings over a period of years.

But over the past dozen years, the reinsurance industry has sharply reduced its willingness to provide insurance companies with that ability to smooth earnings, both out of necessity and by the fiat of changes in accounting rules. Indeed, one of the arguments for investing in the reinsurers on a long-term basis has been that the sector's superior earnings volatility characteristics would lead to a long term upward trend in stock valuations, while the primary market, whose earnings would show increased volatility, would see a long term decline.

The rules are about to change. General Re has recognised that the primary market has a need for the traditional products that smooth earnings. It has been obvious by the actions of companies, such as Allstate, that managements of the primary companies have been concerned about the effect of earnings volatility on stock valuations, especially at the current stage of the underwriting cycle where earnings disappointments are increasingly likely. (See Table 2.) However, the Financial Accounting Standards Board a few years ago altered the reinsurers' ability to account for some products, including this one, in a way that would be meaningful to investors focused on GAAP reports. And, as publicly-traded entities, reinsurance companies such as General Re had recognised themselves the importance of earnings stability to their own stock valuations, thus reducing their appetite for their client's volatility.

Cedants face increased exposures

The difficulty faced by the direct insurance market is that its exposure to large losses has increased dramatically over the past decade. However, it is not the mega-event, the multi-billion dollar loss, but rather high frequency but relatively low severity catastrophes that threaten to erode future earnings. Historically, even relatively small events were reinsured, but today attachment points on catastrophe treaties have moved much higher. This means that the direct insurers must bear a significantly greater amount of cost, in nominal and in relative terms, than they did previously before they receive reinsurance relief.

Putting some numbers on the equation, in 1989, when hurricane Hugo (at the time, the biggest windstorm loss in history) hit Charleston, South Carolina, virtually any event that cost more than $500 million industry-wide would have been covered in the world-wide reinsurance markets. This is why Hugo ultimately had only a relatively small net earnings impact on most insurers, and helps to explain why there was virtually no subsequent upward pressure on insurance pricing. In 1998, estimates are that it will take an event costing $5 billion, the third-worst insured loss in history, before the reinsurers participate. Because the US Congress has not outlawed catastrophes, nor has it created a funding mechanism to help the insurance industry pay for them, then insurance companies would have to pay that difference between $500 million and $5 billion.

Most insurance industry participants and Wall Street analysts point to the fact that statutory surplus in the insurance industry has grown by 130% since 1989, while premiums have only increased 33%, using this as evidence that the industry is over-capitalised. (See chart 1.) However, the industry's exposure to relatively frequent events has increased by as much as 900%, based on that difference in attachment points between $500 million and $5 billion.

Looking at the problem from another direction, underwriting norms used to guide insurers to limit risk exposure on any single event to some small amount, often 1%-2% of statutory capital. This event could be a factory explosion, a hurricane in Florida, or any of a myriad of potential disasters. Underwriting guidelines would be in place that would enforce these norms:

* Do not insure two large buildings contiguous to one another. If one burns, the blaze could spread to the other, and we do not want to be on both risks;

* Limit exposure in southern Florida to a [specified] number of homes per square mile;

* Buy reinsurance.

In fact, all of these guidelines worked through 1989, evidenced by the relatively small net losses reported by companies that had substantial gross loss exposure in South Carolina. Most companies with large losses attributed to hurricane Hugo had those losses narrowed to within a range of 0.5%-1.5% of surplus, thanks to the world-wide reinsurance market.

Today, the story is much different. No longer can insurers look to their reinsurance partners for help with the smaller catastrophe losses, the occasional $1 billion event. Even quota-share treaties have loss ratio caps and occurrence limits that insulate the reinsurers from high frequency/low severity catastrophes. As a result, there has been a noticeable increase in earnings volatility in the property lines of insurance since 1989, as measured by the standard deviation around the mean loss ratio. With greater frequency, investors can anticipate situations such as the first quarter of 1996, when eight separate events aggregated up to $3.6 billion of insured losses and caused every primary company from Allstate through USF&G to apologise to shareholders for poor earnings comparisons "because of catastrophes", while the Bermuda based catastrophe reinsurance specialists reported record low loss ratios.

Perhaps even more telling, actuaries today generally agree that their quantifiable "probability of ruin" is higher in the primary insurance sector than it is in the reinsurance market, representing a reversal of the historical position that has only occurred since 1993.

Conclusion

The conclusion has to be that significant changes in the reinsurance markets, in terms of product availability and the economic level of pricing, have forced insurance companies into significant changes in capital allocation and operating leverage. While the premium-to-surplus ratio may be at historic lows, they are at these levels by necessity, simply because premiums today do not define risk exposure to the extent that they once did. To increase operating leverage will increase an insurer's risk exposure and its earnings volatility; and any increase in earnings volatility will put pressure on the company's stock valuations in terms of price/earnings multiples, over time. (See Table 3). This is the issue General Re intends to address for its clients.

General Re may not be the leader in this movement, as the company has indicated that it is actually following the lead of its major competitors.

If the reinsurance underwriter, in this case General Re, is to provide an earnings-smoothing product that is meaningful to a primary market that in turn is increasingly sensitive to the needs of investors and the demands of the rating agencies such as A.M. Best, then the reinsurer by definition will be assuming its client's volatility. But General Re has recognised that, if this is the case then its own premium stock valuation will be endangered, and it therefore can no longer operate so visibly in the public domain - the only solution is to find a large enough cocoon where its volatility will be less visible. Berkshire Hathaway - friend, client and business partner - was an obvious choice.

General Re, the largest reinsurance company in the US market and third largest in the world, is in a commanding position to dominate the sector. Perhaps only Employers Re (owned by General Electric) and the combination of Munich and American Re, can offer the same capabilities without having to factor in shareholder concerns. These three companies, all direct writers, already control about 40% of the premiums reported by companies listed in the Reinsurance Association of America (RAA) survey, and the ability to offer capacity for the more volatile risks puts them at a significant advantage over the rest of the market. (See Table 4.)

The other publicly-traded reinsurance companies have two choices: they can either attempt to go it alone as independent shareholder-owned vehicles and accept the consequences of their market actions; or they can do what General Re did and find much bigger partners where their earnings volatility will not be so noticeable. If they attempt to go it alone, they face certain pressure on their stock valuations: either volatility increases, which will scare away investors; or if they maintain the current market posture and eschew the riskier business, their customers will go away. This should not be a difficult decision!

So who are the candidates? Already, a number of the Bermuda based property catastrophe reinsurers have been swallowed up by larger companies on the island that are attempting to diversify themselves. But the large Bermuda based casualty insurers have also indicated an interest in becoming more active in the standard lines reinsurance market in the US, and the logical move would be to buy one of the large brokered market names; it may well be time for Everest Re and NAC Re to give some thought to this. Finally, Transatlantic Holdings probably will prove that one really can go home again, and return to the fold of the American International Group that already owns 49% of the outstanding shares.

Michael A. Smith is managing director, Bear, Stearns & Co Inc, New York. Tel: (1) 212 272 9465. Fax: (1) 212 272 5304. E-mail: masmith@bear.com.