Sarah Goddard looks at the background of one of the most successful property catastrophe reinsurers, Renaissance Re.
Renaissance n a renewal of interest or creativity in an area
Mediaeval Europe was a dark place indeed. Warring city states, recurring plagues, feudal barons and a dominant church were all factors in the fatalistic view that the future was in the lap of the gods. Then, in the 13th century, an Italian named Leonardo Pisano - more famously known as Fibonacci - discovered the Hindu-Arabic system of numbering while visiting Algeria, and took part his new knowledge to Europe. This transformed the numerical system from the Roman method - I, V, X, C, M, et al - to a system which enabled complex calculations to be formed, and upon which our modern form of mathematics is based. This fundamentally transformed the world of numbers; the new mathematical universe opened up huge possibilities, paving the way for Leonardo Da Vinci's futuristic machine designs to the technology of today.
But that was just the start. What the new mathematical system triggered was a new way of assessing past experience and, for the first time, predicting the future. Risk analysis became possible, and calculating future outcomes based on mathematical models derived, at least in part, from that analysis shapes the world as we know it today.
So, maybe, it was appropriate that in the wake of Hurricane Andrew, which hit the Caribbean and the US ten years ago, one of the new players in the property catastrophe market to open up shop should decide to call itself Renaissance Reinsurance.
Launched in Bermuda in 1993, Renaissance Re decided to take a scientific view of property cat - a rebirth of the sector, one might say. As John Lummis III, Executive Vice President and CFO explained, "Andrew was seen as an unprecedented loss to the insurance industry - it cost $16bn in 1992 dollars. It was a wake-up call to the industry to create a lot of sensitivity to the risk of natural catastrophes - it lent some real credence to catastrophe simulation models."
Probabilistic catastrophic modelling really started in the mid-1980s, with the work by Karen Clark first as a joint venture with broker EW Blanch and shortly afterward through the formation of Applied Insurance Research (AIR). It is this type of technology which is at the heart of Renaissance Re.
When Renaissance Re Holdings Ltd was launched in June 1993, it was the smallest of eight property catastrophe reinsurers set up in response to the wholesale withdrawal of capacity from the sector. With Warburg Pincus, US Fidelity & Guaranty and GE as initial investors, it had strong industry backing from the start and $140m in capital.
Founder Jim Stanard, an actuary with a PhD in finance and a former US F&G employee, was a strong believer in the use of models as a lynchpin for this type of underwriting, and championed the development of the company's proprietary Renaissance Exposure Management System (REMS) model, an amalgam of several commercial catastrophe models available on the market and Renaissance Re's own analytical and information systems. Another of the original executives, William (Bill) Riker, now president of Renaissance Re Holdings and chief operating officer of the holding company and Renaissance Re Ltd, has degrees in economics and mechanical engineering.
In its first full year of operation - 1994 - Renaissance Re wrote gross premiums of $273m with a combined ratio of 61.6% and a return on shareholders' equity of 44.1%. This final measure, return on shareholders' equity, is a vital element of the Renaissance Re philosophy; a superior shareholder return is paramount, and all risks presented are passed through the Renaissance Re model not just for pricing and exposure purposes, but to assess their effect on shareholder return. Net premiums written in 1994 totalled $270m, and net income came to $109.3m.
In July 1995, Renaissance Re completed its initial public offering, raising $54.3m, which was partly used to reduce company debt. Over the course of that year, gross premiums increased to $292.6m, with net income rising to $165.3m.
However, the following year saw a softening property cat market, and Renaissance Re started exercising its underwriting discretion; gross premiums written fell 7.8% to $269.9m. The increasing competitiveness of the market had led to an 8.8% decrease in premiums from renewal business, an 11.8% decrease in premiums from Renaissance Re declining to renew existing business and a 1.3% decrease in reinstatement premiums. This was offset by a 14.1% increase in premiums from new business. At the time, Chairman, President and CEO Mr Stanard commented, "The capital available in many segments of the reinsurance market clearly exceeds the opportunities to deploy it in the short run. . . Because the property catastrophe business has been among the most profitable segments of the market recently, it is the focus of much competition."
At the same time, though, consolidation was beginning to withdraw some capital from the market in general, as it was transferred from being used as underwriting capacity to provide funds for acquisitions. This industry consolidation helped fuel a `flight to quality', according to Mr Stanard, "which benefits Renaissance as a premier catastrophe specialist with over half a billion dollars of capital." On the flipside, it also reduced the amount of reinsurance being purchased by the newly-combined businesses. With premium rates falling between 10-15% for US business and 15-20% for non-US risks, Renaissance Re decided this was the time to diversify, but within the sector in which it had expertise, and the result of this decision was the formation of Glencoe Insurance Ltd, a Bermuda-based primary insurance company.
Glencoe was launched with a $50m capital contribution in January 1996, and in the following June, Renaissance Re sold 29.9% of Glencoe to investors in a private placement. Using Renaissance Re's underwriting, marketing and capital management facilities, Glencoe's remit was to write catastrophe exposed property business, and it received approval to write business on a non-admitted basis in 21 states in the course of its first year of operation.
Marketing joined risk modelling and capital management as the third element of Renaissance Re's strategy to become the `market of first resort'. Activity was ramped up over the course of 1996, focusing on speedy quotations, designing customised products and helping clients to interpret cat model results.
Because of its lead position within the market, the company was seeing almost all property cat reinsurance programmes, giving it strong underwriting information and allowing it to target clients which best fitted into the company's portfolio. With a potential client base of less than 500, by this time Renaissance Re already had almost half the market on its books, with around 230 clients and 400 or so reinsurance programmes with them.
There was little doubt in 1996 that the market was softening, and Renaissance Re followed its philosophy of returning excess capital to shareholders. Along with a $72m stock repurchase from its founding shareholders, the company repurchased $28m from public shareholders in January 1997, 11 months after completing a secondary stock offering and just two months after increasing its credit facility to $200m. The repurchases "allowed us to return excess capital to our shareholders in a manner that was tax efficient and accretive to earnings per share," commented Mr Stanard.
Over the course of 1997, the California Earthquake Authority (CEA) was looking for capacity for its $1.4bn excess catastrophe reinsurance programme, and approached Renaissance Re along with nine other reinsurers to provide a lead quote. Chuck Quackenbush, the California Insurance Commissioner at the time, commented that Renaissance Re was the first reinsurer to respond, "and when that happened it was like opening the floodgates," he said. "Everyone wanted to follow Renaissance's lead."
Despite opportunities such as the CEA programme, the market continued to slide in 1997. Renaissance Re maintained its stance against writing for premium income. Against a background of premium rate falls between 10% and 15% for US business - and more for international risks - compounded by particularly low loss experience, the company's gross premiums written fell 15.4% from 1996 levels, to $228.3m. This included a 17.4% reduction from programmes that Renaissance Re decided not to renew and 9.6% from lower pricing, participation levels and cover on renewed business. New business accounted for an 11.6% increase in premiums.
Much of the premium reduction was due to a fall in written retro business, which accounted for $59.5m in 1997, compared to $103.7m the previous year. At the same time, Renaissance Re decided to take advantage of the soft retro market, and increased its own cover, paying $32.5m in outwards premiums in 1997 compared to $18.3m in 1996. During the course of the year, it also wrote some aviation and marine coverage.
Consolidation among the client base continued, though Mr Stanard felt it had not harmed Renaissance Re's results. "With the security we offer, we benefit from the trend of clients seeking to deal with a smaller number of reinsurers - we are one of the survivors when cuts are made," he said. But the squeeze on the reinsurance sector fuelled Renaissance Re's expansion of its primary businesses. During 1997, Glencoe received non-admitted status in a further five states, and the insurer's gross premiums lifted from $1.5m in 1996 to $7.0m in 1997.
Complementing Glencoe's access to the excess and surplus lines markets in catastrophe-prone states, Renaissance Re launched DeSoto Insurance Co in September 1997. With an initial capitalisation of $10m, DeSoto started writing insurance business in January 1998, as the first special purpose homeowners' insurance company licensed by the Florida Department of Insurance. At its launch, DeSoto assumed about 12,000 policies from the Florida Residential Property and Casualty Joint Underwriting Association (JUA), with in force premiums of around $10m. As with all diversification, Renaissance Re's priority was to produce an attractive return on capital, alongside developing competitive advantage and keeping sufficient resources focused on the core business.
A salient lesson
The end of 1997 also saw the ill-fated purchase of Nobel Insurance Co, a specialist in low value dwelling homeowners' insurance, for $56.1m. Nobel, a licensed insurer in all 50 states, also wrote commercial property, casualty and surety risks for specialised industries - in particular the explosives industry - and its casualty business was `substantially' reinsured by American Re-Insurance Co and Inter-Ocean Reinsurance Co. In addition, Nobel included an independent loss adjuster, IAS/CAT Crew. Renaissance Re had been a reinsurer on Nobel's programme, and financed the acquisition with $35m raised from a syndicate of banks, as well as a $15m revolving credit facility.
The Nobel purchase was completed in June 1998, and between September and December of that year, Renaissance Re put $9m of capital into the company, as well as providing it with a $8.9m limited recourse loan to liquidate Nobel Ltd. But in the fourth quarter of 1998, Renaissance Re recorded a $40.1m post-tax charge, comprising $29.6m adverse development on Nobel Insurance Co's casualty and surety books, a goodwill write-down of $6.6m and other related costs of $3.9m.
As a result, Renaissance Re decided to exit Nobel's businesses, and in 1999, Nobel completed the reinsurance of its casualty and surety business, and found third parties to take over its bail and low-value dwelling books. It also sold off IAS/CAT Crew to its management team. Renaissance Re kept ownership of Nobel, however, along with the licenses. It also benefited from a retroactive reinsurance contract bought by Nobel at the time of the acquisition that provided $38.0m cover for adverse developments on its pre-October 1997 casualty book. The Nobel purchase provided a salient lesson for Renaissance Re, which subsequently would look at start-ups rather than acquisitions as a way of diversifying and using capital efficiently.
While the Nobel purchase was still fresh, and the market still soft, Renaissance Re continued its programme of repurchasing stock. Over the course of 1997, the company repurchased 1.5m shares - 6.4% of the outstanding shares at the beginning of the year - for $53.5m. This followed hard on the heels of a stock repurchase in December 1996, when the company picked up 2.1m shares for $79.1m, and during the course of 1997 Renaissance Re issued $100m of securities to retire bank debt. In addition, secondary offerings by the founding shareholders increased the public ownership of Renaissance Re from 26% at the beginning of the year to 54% by year end.
After a quiet catastrophe year in 1997, the following year proved much more difficult for the industry as a whole. January ice storms in Canada, spring storms in the mid-west, Hurricane Georges in the Caribbean, Hurricane Mitch in Central America, floods in the UK, riots in Indonesia, floods in China and Typhoon Vicki in Japan all contributed to one of the worst loss years on record and the third largest property cat year in the US to that date. The abnormally high catastrophe experience did not, however, cause Renaissance Re undue problems. In fact, the company recorded an operating return on equity of 19%, although the $40m charge associated with Nobel reduced the return on equity to 13% for the year. At the same time as the cats were beginning to seriously hurt some of the reinsurance players, an investment market unwilling to put its faith in the re/insurance sector was providing a barrier to new start ups.
Following on from two years of reduction in gross premiums, Renaissance Re booked an increase of 18.5% in gross premiums for 1998, up to $270.5m. This was, however, due to premiums from its direct insurance operations; Nobel contributed $30.9m in premiums, while DeSoto, which started writing in January 1998, contributed $26.7m. Gross reinsurance premiums fell 6.3% to $207.2m, from $221.2m in 1997. This fall was from a 16.4% drop in premiums from non-renewed business, and a 14% drop from lower pricing, changed participation levels and altered coverage. Offsetting this was a 24.1% increase in premiums from new business.
Again, Renaissance Re took advantage of the soft market conditions, increasing its ceded retro cover to $47.7m in premiums, from $27.7m the previous year.
During the course of 1998, Renaissance Re repurchased more than 1m shares, in two programmes, for $42.7m.
A new type of venture
Although property cat reinsurance business in general was showing pretty disappointing pricing levels, Renaissance Re entered into a joint venture with State Farm Mutual Automobile Insurance Co, the largest US insurance company, at the beginning of 1999. The resulting company, Top Layer Re, was a new type of company. Writing high level, non-US property cat reinsurance business, Top Layer Re is 50% owned by each partner, with Renaissance Re originally providing $600,000 in capital and a $50m letter of credit. It is responsible for the underwriting, marketing and administration of Top Layer Re, while State Farm provides about $3bn of backing, primarily through reinsurance.
Diversification continued in 1998 with the launch of a structured products facility which completed eight transactions during the course of the year, including catastrophe and weather-linked securities and derivatives, generating about $3m in fees and risk premiums for the year. In the fourth quarter of 1998, Renaissance Re recovered $7.5m from a non-indemnity catastrophe index transaction.
The company also extended its core business in 1998 by opening Renaissance Reinsurance of Europe in Dublin, to serve its European client base and extend its marketing reach.
A bad loss year
The following year was one of the worst on record for catastrophe losses, with estimated losses topping $24bn. Where it differed from many other bad loss years was in the attrition of a number of large losses, including the European windstorms Anatol, Lothar and Martin right at the year-end. During the course of the year, Renaissance Re's gross premiums written leapt to $351.3m, up from £270.5m in 1998. Net premiums rose to $221.1m in 1999, from $204.9m the previous year, despite a drop in premiums from both Glencoe ($5.0m in 1999 compared to $5.6m in 1998) and DeSoto ($14.3m in 1999 compared to $26.7m the previous year). Reinsurance premiums however, rose to $282.3m from $207.2m in 1998. This increase was a result of the contraction in reinsurance capacity in the wider market following the 1998 and 1999 losses. Despite a total of nine significant worldwide catastrophe events, including Sydney hailstorms, typhoon Bart in Japan, earthquakes in Taiwan and Turkey, and Midwestern hailstorms, the company reported net incurred losses of $64.4m for the year, up from $42.4m in 1998.
During the course of the year, Renaissance Re renegotiated its revolving credit facility, increasing the commitment from $200m to $300m. It also increased its borrowings from $50m at the end of 1998 to $200m by the end of 1999. These funds were drawn to increase Renaissance Re's liquidity, and to be available to operating subsidiaries following a large loss event, when capacity may become scarce and business opportunities could arise. Renaissance Re continued its stock repurchase programme in 1999, taking back almost 2.23m shares over the year for $80.1m. Following on from the success of the Top Layer Re formation, Renaissance
Re entered an agreement with Bermuda insurance and reinsurance group Overseas Partners to form OPCat. Wholly owned by Overseas Partners and capitalised with $450m in equity, OPCat had Renaissance Re as its sole underwriting manager, through the newly formed Renaissance Underwriting Managers Ltd, writing companion layers to the Renaissance Re programmes and rewarding Renaissance Re through fee income. Overseas Partners' withdrawal from active business earlier this year has led to Renaissance Re absorbing the OPCat business into its own books.
The European windstorms of December 1999 provided the push towards a hardening market long-awaited in the property cat sector. "The recent upward swing in the pricing cycle is unusual in that it is not triggered by a capacity crunch: there is still more than enough capacity available at prices well below those of the early 1990s," commented Mr Stanard in early 2001. "This, combined with the fact that 2000 was a year of unusually low cat losses, has moderated the price increases and much of the business in the market remains under-priced."
He continued: "The upward price pressure continued through the January 2001 renewal season. There are currently no signs of a move towards price reductions, but unless insurance or capital market losses pull a significant amount of capital out of the industry, future price increases will be relatively modest."
Interestingly enough, a year later, following the huge - and still unquantified - losses of September 11, insurance and reinsurance capacity remains available, but the price increases have, indeed, come through.
During the course of 2000, Renaissance Re managed $397m in catastrophe premium, from the Top Layer Re and OPCat facilities and its own business. Top Layer Re's premium income reached $25m in 2000, up from $4m the previous year, while OPCat's first operating year in 2000 booked premium income of $55m.
The continued development of the REMS system, along with the strong capitalisation and superior results of Renaissance Re, led AM Best to upgrade the company to A+. The hardening retro market in 2000 led Renaissance Re to lock into multi-year agreements with its reinsurers, while it increased its specialty reinsurance lines business to $38m in 2000, from $3m in 1999. Accident and health, finite and satellite business were added to the book during the course of the year, and the company entered an alliance with Chubb Re to screen the business initially, reciprocated by a small participation in the Renaissance Re cat reinsurance book.
In the first half of 2000, Renaissance Re repurchased a further $25m of its stock, while Warburg Pincus distributed the remaining shares it held to its limited partners and GE sold 1m of its shares into the market.
During the course of 2000, Renaissance Re repaid $200m of its bank debt, leaving only $50m outstanding with a $310m revolving credit facility available in the event of a large cat event. Such measures proved useful a year later. Come September 11 and the huge subsequent dislocation in the markets, and Renaissance Re was immediately able to act. Firstly, it increased the capital and surplus of direct writer Glencoe to $100m, and on October 9 announced the formation of DaVinci Re Ltd, a new side-by-side reinsurer initially capitalised with $500m, including $200m from State Farm, $100m from Renaissance Re itself and the balance from other investors. Later, Bermudian reinsurer Max Re purchased $75m of DaVinci Re shares, and Renaissance Re provided $100m in bridge debt for the new operation. Renaissance Re announced its expected gross loss from the World Trade Center tragedy would be $150m, with a net loss of around $50m, and in October 2001 it announced the issue of up to $400m in securities.
In January this year, the company committed a further $100m to Glencoe, and a month later it filed a shelf registration with the Securities and Exchange Commission (SEC) to offer up to $500m in securities, when just $64m was still available from the previous offering. In April, the $100m bridge loan to DaVinci Re was replaced with a third party credit facility led by Citibank NA.
All these measures are tackling the substantially hardened re/insurance markets, particularly for catastrophe cover. "The worldwide insurance markets were already in turmoil from the growing underwriting losses due to years of inadequate pricing when September 11 threw them into chaos," said Mr Stanard. "The torrent of red ink accelerated upward pricing pressures that already existed in the reinsurance markets. However, this hard market cycle differs from the hard markets of 1985 and 1993 in one important respect: there is sufficient capacity available for most types of business - as long as the insured is willing to pay the market price." In fact, this has led to a flight to quality as buyers have demanded higher rated reinsurers and stronger balance sheets behind their programmes - to Renaissance Re's benefit.
With the extra funds raised through its $233m common stock offering and $145m preferred stock offering, along with the new capital in DaVinci Re and the additional $135m in Glencoe, Renaissance Re looks set to substantially increase its influence on the sector. According to Mr Stanard, the company is aiming to both increase its market share in the property cat reinsurance business and add to its specialty reinsurance portfolio. "The market turmoil has finally presented opportunities at attractive pricing that we have patiently been awaiting over the past several years," he commented. "We expect to write over $100m of premium in 2002 from aviation, catastrophe exposed workers' compensation coverage, property risk, satellite and finite reinsurance, among other lines." At the same time, he anticipates Glencoe will increase its primary book from $13m in premiums for 2001 to more than $100m in premiums this year.
"The right time to diversify is in a market like the current one, in which competitors are shedding business in order to get back to their core competencies," he said. "We have, today, the staff, management, vehicles and client relationships in place to move quickly and smoothly to build our base of business."
This will no doubt be assisted by the rating upgrades Renaissance Re has received in the wake of September 11, and could herald the birth of a new wave in the interest in the creative area which is Renaissance Re.
By Sarah Goddard
Sarah Goddard is the editor of Global Reinsurance.