The reinsurance sector has changed enormously over the past decade and more, since the last major upturn in underwriting that commenced in 1984. In the ensuing years, the industry has seen significant shrinkage, from 86 active professional reinsurance companies in 1982, to only 28 listed by the Reinsurance Association of America today - and this number is expected to continue to decline.
Coupled with the shrinkage in the population, market leadership also took over, where it became more important for cedants to focus on doing business with a small number of large and well-capitalised reinsurers, rather than the other way around as it had been pre-1984. Because of this, reinsurance underwriters were able to have their way, imposing a risk-excess model on the market more broadly, in place of the proportional form that had been prevalent, thereby gaining more direct control over their own underwriting and pricing.
Using a term that we take credit for coining in 1987, the reinsurance sector had de-coupled itself from the primary market's underwriting cycle. The net result was an extraordinary run of out-performance within the insurance industry for the reinsurers. From 1986 on, the sector's underwriting results were consistently superior to total property-casualty insurance industry results, except in 1992 that was marked by hurricane Andrew, and 1995. Putting this performance in context, one has to go all the way back to 1971 to find a year when the reinsurance combined ratio came in lower than the total industry average.
Not only did the sector's earnings out-perform within the insurance industry, but the stocks of those companies that were publicly-traded also led, trading at consistently superior valuations. The direct writers General Re and American Re both traded in the vicinity of 2.0-2.5 times book value, and the smaller names such as NAC Re, Everest Re and Transatlantic Holdings hovered at 2.0-times book.
Somewhere along the way, it all went wrong.
In 1999, reinsurance industry results took a turn for the worse. The 28 professional reinsurers standing at year-end reported an aggregate combined ratio that soared to 114.6% from the 104.5% recorded by 32 professional reinsurers in 1998. This was in sharp contrast to the 107.9% turned in by the entire insurance industry for the year (we can debate whether or not those are the “real” numbers, but for the time being they will have to do for a benchmark). And investment income did not help, falling 1.5% from the year earlier. As a result, the return on average equity fell to a paltry 1.7% from 9.1% the year before.
Results have remained poor into the new millennium, based on the data through the first quarter of 2000.
According to the Reinsurance Association of America, 28 professional reinsurers (down from 31 in the 1999 first quarter) reported an astounding 59% drop in net income despite an 11% increase in written premium. The key contributors to the decline were an increase in underwriting losses of 156%, with the combined ratio at 112.3%, up from 103.8%; and a 46% drop in realised investment gains. Investment income was nearly flat, up only 1.3%. By our estimate, the annualised operating (excluding investment gains) return on average equity amounted to a measly 2.6%.
One might conclude that the group would be better off investing its investors' capital in certificates of deposit (CDs) at the local bank, rather than trying to underwrite reinsurance.
The biggest factor behind the poor first quarter results was the 25% increase in reported incurred claims, resulting in a ten-point increase in the loss ratio. The group's expense ratio declined by 1.4%, but we do not read too much into this, as the decline probably reflects a reduction in profit-sharing commissions paid to the ceding underwriters: no profits, no commissions.
Ten of the professional reinsurers reported a net operating loss. Among them, the US operations of the French reinsurer SCOR reported a $30 million loss related to the reversal of $38 million of premium (about 6% of premiums reported for all of 1999) that had been over-estimated in previous months. But in addition to SCOR, American Re, and Swiss Re America, two of the six largest reinsurance underwriters in the United States reported deep operating losses.
Everest Re and Transatlantic Holdings, the two largest publicly-traded “pure plays,” each turned in superior results, both in terms of underwriting performance and return on equity. Strong returns were also reported by General Re (mostly because of strong investment income, because its underwriting results were terrible), Odyssey Re, Trenwick, and NAC Re.
As with the underwriting performance of the reinsurance sector in general, the reinsurance stock valuations have also fallen from their previous perch. Transatlantic Holdings has traded at premium valuations, more probably a reflection of its majority ownership by American International Group. But lagging far behind are Everest Re and the remainder of the group. And even Transatlantic Holdings' valuation trails the multiples that investors had long been used to seeing among reinsurance stocks.
In the past, we have been able to make a clear distinction between the “haves” and the “have nots,” identifying sub-sectors where underwriting performance was generally good or especially bad. Not this time. No matter where we look - among the four direct writers, within the brokered market, big company, small company - the results are generally bad. Even the bright lights in the darkness are thinly sprinkled throughout the group.
The direct writers - that we once viewed as the leaders of the pack and nearly immune to the horrors - neatly demonstrated their own ability to get it all wrong. Two of the four recorded deep underwriting losses, and in aggregate the direct writers earned a slim 1.2% annualised return on equity (ROE) and scored an astoundingly high 114.6% combined ratio.
In comparison to the direct writers, the brokered market reported a combined ratio of 109.5%, with an annualised ROE of 4.2%.
Including the four direct writers, the 14 largest professional reinsurers, one-half of the population, accounted for 85% of written premiums and 91% of the aggregate underwriting loss, compiling a combined ratio of 113.1%. The bottom half of the crowd wrote at a 107.3% combined ratio, earnings an annualised ROE of 4.2%.
Where has the leadership gone? Probably the same place that the discipline once exhibited by ceding underwriters has disappeared to. While the 14 largest professional reinsurers reported a 10.1% increase in written premiums (mostly because of the direct writers, and that because of an acquisition that doubled Swiss Re America's premium), the smaller half of the sector recorded a 21.1% increase in premiums.
So what has happened? In our estimation, a combination of two factors has occurred. First, the year 1999 was a difficult one for the insurance industry with respect to catastrophe losses. No single event amounted to much, but the problem was the accumulation of a number of comparatively small events, just at a time when the reinsurance industry was stepping up to help insurers address the potential volatility this peril presents to the industry going forward. But perhaps the bigger influence was the decline in discipline of the cedants as they searched for revenue and profit opportunities, some of which were designed to take advantage of their reinsurance capacity.
What we have seen in the past two or three years is a surge in so-called “program” business. This is business aimed at a target market and often underwritten by a control source outside of the risk-bearing insurance company's direct operation. The managing general agent (MGA), once viewed as the industry's pariah, is back! And very often, these programs, many targeting the non-standard auto market and workers compensation, have been backed by cheap quota-share reinsurance.
It became disconcerting a few years ago when word began to spread that a number of the smaller brokered market reinsurers were offering attractive terms on non-standard auto deals that were being accepted. More recently, the eruption of the so-called Unicover workers compensation pools has highlighted the dangers present in this business model, where everyone was hurt (except the MGA).
Some may view Wall Street to be at the heart of the problem. Insurance is a long term business, and the appropriate business model is to withhold capacity during the down-cycle phase of the industry's cycles. But Wall Street has a much shorter time horizon that does not fit the long term nature of the insurance business. This leads analysts to contradict themselves with comments such as “XYZ is a great company that is maintaining its discipline in a competitive market - but its growth remains below expectations, so we are neutral on the stock.”
The problem develops when insurance companies, operating in an environment where industry-wide growth is negative in comparison with the economy, begin to search more desperately for ways to please Wall Street. Growth in this environment is expensive - companies must either buy other companies at a premium, or they must under-price the business they write. And this is where the MGA armed with cheap reinsurance capacity comes in.
For an insurance company, the underwriter's salary is a fixed cost. A company that writes little business feels the urge to write more, in order to cover that fixed cost. However, by allowing the MGA to produce and underwrite the business, the insurance company transfers the fixed expense to its producer, converting this to a variable cost. The problem is that, under this model, the insurance company should probably spend more to maintain an oversite presence in the MGA's office to make sure that risk selection and pricing standards are being adhered to.
Of course, if cheap quota-share reinsurance is available, who cares about risk-selection or pricing? We recently asked one underwriter that specialises in program business, what the gross loss ratio is, before reinsurance, of this segment of his business. His answer was that he did not know - he books this business at the industry average but believes the loss ratio is much higher because the book is heavily weighted toward the problematic workers compensation market. But, in his view, the actual loss ratio really does not matter, because the book is mostly reinsured.
This is a business model that is based on “sticking it” to a partner, and it is not sustainable. Either the partner will figure it out and leave, usually after huge losses have been rung up, or the partner will eventually be bankrupted. In the latter case, the ceding insurance company finds that it has simply converted unacceptable underwriting risk to an unacceptable credit risk on the asset side of its balance sheet.
Based on the results posted recently, it would appear that the reinsurance industry is rapidly approaching the “crunch,” where capacity starts to drop off and prices rise. We may tongue-in-cheek suggest that the underwriters would be better off investing in CDs, but some will take the hint and leave voluntarily, while others will find themselves captive to a regulatory takeover in the sort of merger and acquisition activity that investors do not ordinarily think about.
In either event, we believe that the population will drop sharply in the year or two ahead, with a concomitant rise in pricing. But in the short term, earnings results will continue to bear the weight of poor pricing and poor underwriting decisions made in recent years.