Investors Rebuilding Bermuda Capital

While diversified companies such as ACE, Max Re and Arch Capital emerged with positive results (net income) through nine months:

o Rosemont put its business in runoff.

o Quanta announced plans to curtail its writings in property businesses.

o Three catastrophe specialists–Montpelier Re, PXRE and IPC Re–watched rating agencies drop their once top-tier financial strength ratings, even as they raised capital to replenish lost funds.

Hurricane losses–totaling more than $8 billion for the Bermuda market overall–did not stop investors from putting nearly $6 billion of new capital into 14 existing Bermuda companies, according to a report from Moody's Investors Service.

Investors eyeing opportunities in hardening property-catastrophe reinsurance, retrocessional and energy markets also poured funds into at least seven new start-ups, adding more than $5 billion in new capacity, Moody's reported.

Body Blows

"This loss hit us square in our core business, and as such, we expect the most attractive opportunities will be in our core business," said Kip Operling, chief financial office of Montpelier, during the company's earnings conference call, reporting a third-quarter net loss of $875.1 million.

Chief Executive Officer Anthony Taylor characterized Katrina as a unique storm with multiple perils causing extraordinary commercial losses not seen in Hurricanes Hugo or Andrew, in addition to large offshore marine and energy losses.

"All in all, this event was an imperfect storm for us, with a correlation factor between our books much greater than we would have expected," he said. "Our industry has been taught an ugly lesson, and risk perceptions have changed overnight. After a $50 billion-plus [industry] loss from Katrina, followed by a close shave from Rita, we must now accept that $100 billion-plus wind seasons, or even single events, are more likely than previously thought."

Hurricane Katrina, together with losses from other third-quarter catastrophes, produced $972 million in losses for Montpelier, prompting a two-notch downgrade from Fitch Ratings–to "triple-B" from "A-minus." The Chicago-based rating agency said that infrequent large losses are expected from property-catastrophe reinsurers, but the magnitude of Monpelier's losses–representing 66 percent of its June 30 equity–revealed risk concentration far beyond Fitch's expectations.

Although Montpelier raised $600 million in a September public offering, A.M. Best also cut its rating (to "A-minus") as did Moody's (to Baa1). James Eck, assistant vice president for Moody's, cited concerns about the company's risk management, noting that Montpelier had returned capital to shareholders with a $390 million special dividend earlier this year.

Montpelier said the dividend was predicated on a reduced risk profile. The loss numbers "suggest they didn't really pull back," he said.

"I think a lot of companies will be reassessing their exposures," Mr. Eck said. "Retrocessional capacity is going to shrink and become much more expensive. For a lot of companies, the way to control risk is going to be writing less business."

Both Mr. Taylor and Jeffrey Radke, CEO of PXRE, said their companies would reduce exposures. Mr. Radke (whose company suffered cuts to "A-minus" from Best and Standard & Poor's, and to "triple-B" from Fitch) said PXRE would shed risk-excess contracts and catastrophe protection on direct and facultative business. (Risk-excess contracts cover losses above a specified retention for each risk involved in each occurrence.)

"Our customers' and our own expectations of damageability" for large multi-location property risks covered by these contracts "were simply wrong….The [catastrophe] models were at their worst in this area–and the business showed ill effects of steadily worsening underwriting conditions since 2002," he said, vowing to shrink PXRE's already small market share.

Mr. Taylor said Montpelier would reduce aggregate catastrophe exposure, buy more traditional retro coverage, and partner with capital providers in non-traditional reinsurance vehicles. (Pre-Katrina, Montpelier already partnered with a hedge fund manager, co-founding a Cayman Islands reinsurer.)

Mr. Taylor told investors that rating agencies are "raising the bar" in evaluating capital charges for catastrophe risks, and Mr. Radke said recent downgrades weren't votes against claims-paying abilities but against monoline property-catastrophe business models. Still, the two said they wouldn't change their stripes to diversify into long-tail lines.

A Viable Model?

The executives, repeatedly asked if a monoline property model was still viable, answered affirmatively–as did rating agency representatives. "We do not have to set aside capital for casualty loss-reserve risk, litigation uncertainties, or myriad other risks we avoid with our focused strategy," said Mr. Radke.

Donald Thorpe, senior director for Fitch in Chicago, said it's fine to be a monoline insurer as long as the company avoids risk concentration. Mortgage guaranty companies, for example, write in all 50 states to avoid concentration. Monoline property-catastrophe writers, too, can be successful if they don't have concentrations in a single peril or geographic area, he said.

"As long as people want to live in Florida, on Long Island, in the Gulf, or in California, insurers will have to buy catastrophe reinsurance," said Jim Bryce, CEO of IPC Re. "It's the one product that is going to be sold," no matter what happens.

"We've specialized in that product for over 12 years [and] built up geographic diversification that's unparalleled," he continued. After 9/11 and Katrina, "we're still in business. We're still viable," he said.

What will rating agencies want to see from property-catastrophe reinsurers to affirm ratings–or even upgrade them?

"Better underwriting and better risk management," said Robert DeRose, assistant vice president for A.M. Best in Oldwick, N.J., noting that Katrina was a good test case to evaluate the understanding that company managers have of their risks.

But wasn't Katrina too unique to hold out as a test case? Not according to Mr. DeRose.

"Look at the Gulf Coast. Look at how densely populated that area was. Look at the proximity of the offshore oil rigs and the underground pipelines. Don't you think a prudent underwriter writing property and marine, who knows the way hurricanes track, should know it's feasible for one event to impact offshore and onshore exposures?"

"The exposure was there. Everybody knew it, and no one managed it," he said.

Mr. Eck said reinsurers using the monoline property-catastrophe model "can make a lot of money if nothing happens, but it doesn't appear, over the long term, to be creating adequate returns for the volatility."

"That said, we're only really looking from Hurricane Andrew until now. A hundred years from now, maybe we'll have the answer," he added.

Did Cat Models Fail?

Even data collected over 100-year timeframes fails to provide good answers to some key questions.

Catastrophe models, until now, have been based on 100-to-150 years of past hurricane data, and the assumption is that the next few years will be similar to the last 100, William Riker, president of RenaissanceRe Holdings, told NU.

The assumption is no longer valid, explained Mr. Riker, the principal architect of RenRe's proprietary REMS model (Renaissance Exposure Management System), and Ian Branagan, head of Group Risk Modeling.

Research in the last five years shows that hurricane activity has "decadal cyclicality," said Mr. Riker. "There's enough evidence that we're in a higher-frequency period for hurricane activity than the last 100 years of data would predict," he said.

Mr. Branagan said research in the wider scientific community supports this. The research indicates "we're in a cyclical warm phase of the sea surface temperatures of the Atlantic basin. We've been in this phase since 1995, and it's likely to last for some time," he said, adding that this means hurricane activity will be well above historical averages over the medium term.

Mr. Branagan said RenaissanceRe has now deliberately built assumptions about increased hurricane activity into its models.

"You have to have a healthy skepticism for the accuracy of [model] results," said Mr. Riker, adding that "people in our business started to look at model results as facts," instead of viewing models as tools to assess risk.

Hurricane Katrina also demonstrated there is a lot of risk to which insurance policies respond that just isn't captured in models, he said, pointing to flood losses as an example. "You can choose to ignore that, or you can build in conservatism" to account for unmodeled exposures, he explained.

Mr. Branagan said that going forward, users need to ask themselves tough questions about how they are coding commercial data. "To get good results out of a model, you need to be quite complete in your data capture," Mr. Riker agreed.

For example, "if you have a large commercial account with 3,000 locations, users might [typically] only capture the 100 largest locations–or the large locations in coastal areas," he said, noting that more locations need to be captured.

Opportunity Knocks

Like RenRe, third-party vendors are recalibrating models to reflect higher storm frequencies. With rating agencies adjusting capital adequacy models as well, supply-demand imbalances will vault property reinsurance prices skyward, executives say.

Mr. Taylor, for one, said reinsurers with existing large retro programs renewable on Jan. 1 could face large holes in their programs for 2006, at a time when rating agencies have increased the capital required to support catastrophe exposures, he said.

Meanwhile, the primary markets, "having taken huge hits from Katrina [that] exhausted their reinsurance programs, will be looking for increased reinsurance–and will also be subject to new capital adequacy rules from rating agencies." Unlike reinsurers, they won't be allowed to reduce the scope of coverage because of regulatory constraints.

Executives predict that some reinsurers will reduce coverage. In particular, those that sell risk-excess covers will rewrite contracts to exclude coverage for losses from natural perils. Buyers of these covers will be forced to go to the property-catastrophe market, where capacity will be constrained and prices high, said Russell Fletcher, Montpelier's chief underwriting officer.

On property-catastrophe reinsurance business, he predicted that loss-impacted programs in the United States could see price hikes of 50 percent–or more than 100 percent in some cases. Loss-free programs would see 20-to-25 percent jumps, he said.

Others avoided the question of how high prices might go. Mr. Radke, for example, when asked if he thought retrocessional prices might increase several fold, conceded that 200-to-300 percent jumps were possible "in some spots." However, "I'm uncomfortable throwing out percentages," he said. "The reinsurance industry is better off letting the market work," without talking about guesstimates that "can be used against us."

"The market is clearly moving beyond original expectations, and I'm quite content to not speculate and, for some clients, create a ceiling where we didn't want to."

Will downgraded property-catastrophe reinsurers be able to capitalize on this market? Executives at all three downgraded property-catastrophe reinsurers–Montpelier, PXRE and IPC–predicted they still would see the same programs in spite of downgrades, but that participations might be reduced. (IPC still has an "A" rating from A.M. Best, having seen its rating reduced from "A-plus" earlier this month.)

"Will it be crucial to have at least an A-minus?" one analyst asked Mr. Taylor. "Clearly, having a rating with a 'B' in front would make things difficult, but it does rather depend on what happens to the whole market. If the whole market went 'B'-rated, then it would be a good rating."

Investors Pony Up

While reports of Bermuda start-ups are surfacing regularly, executives and analysts said their entrance would not dampen the market opportunities. "The capital that known start-ups are bringing in is insignificant in comparison to capital that was lost," said Mr. DeRose.

Mr. Eck said it is questionable whether new companies will be fully up and running for Jan 1 renewals. Those start-ups taking business from existing companies–such as vehicles being put in place for Chubb and XL Capital–might have a little bit of an advantage over pure start-ups, he added.

Executives and analysts agreed that limited availability of intellectual capital is a significant disadvantage for start-ups, but the Lloyd's market companies probably have experienced underwriters they can assign to new ventures, said Fitch's Mr. Thorpe, noting that Bermuda interest from Lloyd's is a notable feature of the class of 2005.

He speculated that a lower-cost structure in Bermuda might be attracting this interest, while Mr. Eck said the draw might be the recognition that Bermuda is becoming the center for property-catastrophe exposures. For Lloyd's participants, Bermuda vehicles might simply be being viewed as a way to access customers who might decide to make only one trip–to Bermuda–to get reinsurance capacity, without going to London, he said.

In addition to new capacity entering through start-ups, $6 billion of capital invested in existing players replaced roughly 71 percent of third-quarter catastrophe losses, according to Moody's.

"Companies have had a good track record, after catastrophes, of being able to recapitalize and keep going," Mr. Eck said. However, "from a credit perspective, it's difficult to look at that strategy as being viable in the long term," he cautioned, noting that Moody's does not assume that bailouts will always be forthcoming.

"It will be interesting to see whether investors become a little bit more discerning going forward," he said. Hurricane Wilma could test that, he said.

Infographic–Don, could pick up shot of chick hatching from Bermuda report of a year or two ago…..(4/29/02)

Flag: Roll Call

Head: New Companies, Some Old Faces

Don Kramer, former vice chairman of ACE Ltd., is just one familiar name on the roster of new Bermuda ventures. In a recent report, Moody's Investors Service included this partial list of start-ups:

o A new venture incorporating runoff company Rosemont Re. (Mr. Kramer heads the investor group buying Rosemont's infrastructure and renewal rights.)

o Amlin Bermuda, formed by Lloyd's insurer Amlin Plc

o Validus Holdings, formed by a consortium of private equity companies

o Lancashire, which will target energy and marine business

o A new venture to which XL will cede portions of its property-catastrophe and retrocessional business

o New Castle Reinsurance Company Ltd., formed by hedge fund Citadel Investment Group. Christopher McKeown, former CEO of ACE Tempest Re, will head up New Castle.

o Harbor Point Ltd, formed by Chubb and private equity firm Stone Point Capital to acquire the ongoing business of Chubb's reinsurance operations. John Berger, CEO of Chubb Re, will lead Harbor Point Ltd.