current events

Bermuda is the world's second-largest insurance market and may be changing the way risk is priced and shared



June 2, 2003

The Next Catastrophe Risk/Excess Reinsurance Market Cycle

By Andy Giffin and Rick Shaw*

Submitted and presented to the Research Round Table at the Annual Seminar of the International Insurance Society at New York, July 16, 2003.

*Andy Giffin is a Principal in Tillinghast - Towers Perrin, the management and actuarial consulting firm, based in New York.   Rick Shaw is the Manager of the Bermuda Office of Tillinghast - Towers Perrin.

You've heard the story about the frog and the boiling water.   A frog that lands in boiling water jumps out immediately.   But, a frog in water that is brought to a boil doesn't notice the increase in temperature until it is too late - he's cooked!

It can be that way with reinsurers.   A hard property insurance market following Hurricane Andrew losses was gradually eroded in the mid-1990s.   Guy Carpenter's World Rate on Line index indicates that rates on line peaked in 1993 at nearly four times the 1990 index base, dropping steadily until 1999 to about 1.5 times the 1990 base, then rising slightly in 2000 and more dramatically in 2001 and 2002 back to about 2.5 times the 1990 base.   However, falling prices for reinsurance   in the mid to late 1990s were supported by rising investment gains in equity markets.   Investment gains and the absence of significant catastrophe losses encouraged reinsurers to loosen contractual terms and conditions, to soften prices and to extend coverage to less attractive exposures.    Availability of reinsurance encouraged primary companies to extend their risk acceptance.

From 2000, investment gains declined.   Underpriced risk acceptance came back to haunt.   In 2001, 9-11 alerted the industry to new kinds and levels of losses.   Liability on old business expanded, e.g., asbestos, and new liability for corporate governance lapses exceeded expectations and pricing bases.   Despite market hardening, even before 9-11, many reinsurers, and insurers, found themselves cooked, or nearly so.

The post 9-11 market experience has been described as the aftermath of the "perfect storm", i.e., the combination of a major catastrophic event, corporate governance-related losses, accumulated rate inadequacy in both property and casualty markets and the loss of investment income support.   Given the magnitude of this shock to the system, is it more likely that the markets will now respond better than was the case with earlier systemic shocks?   Will companies take advantage of new analytical tools and exercise better discipline to avoid slipping back into overly aggressive pricing and market share grabs?   Or, will bad habits be repeated with many cooked again?

There are signs that the post-9-11 hard market, having begun prior to 9-11-01 with price hardening following the collapse of investment gains in 2000-2001, may already be losing momentum.   Benfield, the large London broker, has reported less than expected premium increases in 2002 year-end renewals, particularly in property lines.   Many primary and reinsurance prices retain an upward or stable trend today, but are current rates enough to recover from a long period of creeping inadequacy?   Will rates fall back to inadequate levels, and if so, when?

Tillinghast Survey

Tillinghast undertook an informal survey of senior executives and practitioners active in today's reinsurance markets to assess market expectations.   The object of the survey was to determine how market participants view this point in the underwriting cycle.   The current market can be characterized as, broadly, a hard market with new exposures to absorb and little investment income to cover underwriting losses.

These interviews were conducted with industry representatives in the U.S. and Bermuda.   A list of companies represented in the survey is attached as Appendix A.   Some companies included in the survey specialize in catastrophe and excess reinsurance lines; others are more broadly participating in various reinsurance and direct insurance lines.   We have not identified particular observations with any particular organization to avoid revealing proprietary information.   The following represents consensus views, with comment on the range of divergent views.

Not Just Another Cycle

There are indications that this is not just another underwriting cycle similar to those in the past.   Evidence of this includes:

•  9-11 and some recent natural disasters have raised the bar on potential maximum losses, both individually and in aggregate

•  The prolonged downturn in investment markets is forcing greater attention to underwriting profitability

•  Insurance and investment losses have forced many companies to cover significant capital requirements via capital raising, with continuing capacity limitations

•  Reserve strengthening to cover past exposures (e.g., asbestos) is cutting into results and capacity.

Soon after 9-11 fresh capital was injected into the reinsurance industry.   Much of this went to new entities and recent entrants in Bermuda.   The rest replenished the capital of established players.   But, the $19 billion added in 2002, on top of the $26 billion in 2001 is not close to restoring the $200 billion capacity reduction estimated (Swiss Re and others) since 2000.  

The existing players face recovery from the overhang of heavy losses and reserve strengthening on historic business.   This puts a constraint on building new business.   The new money players enter the fray with shareholders expecting to make exceptional returns from a hard market.   All enter today's market with expectations of strong pricing and strong motivation to maintain and strengthen pricing.

The State of the Markets

Our survey sought views on the current state of reinsurance markets, particularly in the catastrophe and excess lines.   The general consensus indicated the following:

•  Property Catastrophe - most see this market as stabilized at adequate rate levels.   Some remaining rate inadequacies are observed in Latin America and Asia (other than Japan).   Additional capacity is available and there has been significant new capitalization.   Cover is generally available at adequate prices.   Some softening is observed in primary insurance markets in medium to large commercial accounts but this has not yet impacted reinsurance pricing.   Some are making measured price and term concessions to get accounts but this is not significant enough to change the current direction.   Many see pressure to reduce rates if 2003 is a light catastrophe year.

•  Casualty - significant inadequacy in reinsurance pricing remains with a continuing strengthening trend.   Reinsurers are proceeding in this market with caution, trying to revise terms to better align the interests of the ceding company and the reinsurers.   Rates for casualty umbrella coverage are still moving up.   Respondents note that in addition to the highly publicized large awards, the average of court awards is up sharply in liability lines.   Various lines show different current conditions, with workers' compensation at adequate levels and leveling off, D&O at strong rate levels, and medical malpractice suffering from the multiple crises at the primary insurance level with still a way to go to adequacy.

•  Aviation - this traditionally volatile market appears to the respondents as having overreacted to 9-11 and is now softening.   The market is considered generally less disciplined than others, due in part to the wide range of participants through pooling mechanisms and retrocessions.   Recent bad experience is consolidating this line in fewer hands, which could result in better market discipline.

•  Marine - Experienced reinsurers consider this market as lacking in adequate pricing information and modeling capabilities, and not maintaining adequate primary or reinsurance pricing.   Some are exiting.

•  Terrorism - The initial response to 9-11 was the exclusion of terrorism coverage from most reinsurance contracts.   Terrorism coverage began to reappear in 2002.   U.S. legislation reshaped much of this coverage at the primary level, eliminating rising exclusions in many areas and adding a variety of new coverage structures.   It remains to be seen how much business the primary market will do in terrorism coverage and how reinsurance pricing will evolve.   Catastrophe modeling is developing for terrorism risk.

As discussed in more detail below, the recent market turmoil has inspired a flight to quality, favoring those with strong balance sheets, higher capitalization levels, strong underwriting and pricing capabilities, and strong market channel access.

The Rise of Bermudian Technocrats

Added to the market mix is the development of some important tools to help make better underwriting, pricing and portfolio mix decisions.   Following Hurricane Andrew in 1992, new money companies entered the reinsurance markets intending to apply more sophisticated underwriting and pricing models, particularly in property catastrophe business (e.g., Renaissance Re and others).   The developed technical capabilities have demonstrated their value in either the survival of independent companies (e.g., Renaissance Re, IPC Re) or absorption of the post-1992 new money companies into larger organizations (e.g., ACE Tempest Re).  

The result is a Bermuda market dominated by technically-oriented companies that make better use of analytical tools for understanding risk and pricing levels, than is common practice in the traditional reinsurance markets.   There are about 120 actuaries based in Bermuda, most involved in underwriting.   Good risk and price analysis will tend to limit the potential for pricing to be sustained at excessive levels, but will also alert companies to points of rate inadequacy as pressures for market softening emerge.   As noted, the rise in rates on line following 9-11 has not been nearly as high as the post-Andrew movement.   It remains to be seen whether the soft market dip to come is also less pronounced.

Some argue that technically-oriented Bermuda has changed the rules of reinsurance.   Historically, many reinsurers felt secure in accepting slips representing small portions of large exposures.   Theoretically, a broad selection of such slips might be expected to produce a near market average result.   However, if Bermuda companies have been successful in identifying the better risks and better matching price and exposure, they may be leaving the rest of the market with less than average results.

Potential for Stable Pricing and Reduced Cyclicality

The speed of the inevitable progression to softer markets depends heavily on whether the new underwriting and portfolio management tools are used with high levels of underwriting skill and strong discipline.   The survey probed the potential for this to happen, along with the possibility that traditional market factors will, nevertheless, make the next cycle much like the last few.

Time to Rethink the Operating Fundamentals

The reinsurance (and insurance) business involves four primary operating requirements:

•  Marketing

•  Risk Management

•  Investment

•  Administration.

To effectively dampen the tendency for markets to soften, these operating areas must leverage new approaches available in the market to better understand risks and risk patterns, to price adequately and to manage operations efficiently.

Risk management must account for new sources of potential loss, from terrorism to corporate malfeasance.   It must account for the higher potential loss from natural disasters.   It must account for the aggregation of risk across individual risks and lines of business.   In a competitive world, it must also make the most of the benefits of correlation of risks and diversity of business lines to remain price competitive.

In marketing reinsurance, brokers and direct writing representatives need to be more vigilant in supporting the underwriting process and more efficient in collecting underwriting information and completing claim payments.   The often heavy cost of accounting for business through multiple brokers and complex arrangements with retrocessionaires needs to be rationalized.

Lax record-keeping and administrative inefficiencies lead to delay and errors, a cost that current market conditions will no longer tolerate.

Investment returns, once a cover for rising combined ratios, cannot now be counted on.   Underwriting needs to stand on its own and produce profits to cover investment losses and provide capital for growth.

It may be that the hard market carries on a bit longer, that investment markets rally, and that the worst fears about potential catastrophic losses do not materialize.   Who can afford to bet on such good fortune?   Reinsurers, and insurers, need to rethink their approaches to the business to address the challenges as we see them now.

Updating the Operating Fundamentals


The best underwriting capabilities are of little use without access to high quality risks.   Some reinsurers prefer to rely on brokers to increase their ability to maintain an optimal portfolio of risks.   Broker intermediation avoids the kind of closeness to customers that makes reinsurers susceptible to accommodating ceding company needs even where the reinsurer's portfolio analysis encourages risk rejection.

Other reinsurers prefer the close relationships with ceding company through direct writing, as well as broker produced business.   They see advantages in understanding   the customer well enough over time to evaluate the underlying risk.   This also provides continuity in the flow of business.

Reinsurers, and insurers, are becoming more aggressive about demanding underwriting information to understand and control ongoing exposures.   Among other things, 9-11 illustrated the importance of knowing where covered employees and property are located.   Proximity to potential terrorist targets is getting more attention.   Additional information is needed for more intensive modeling analysis, as well as to provide improved underwriter understanding of the underlying risks.   Increased consistency in ceding company submissions is improving comparative risk analysis.

Collection of underwriting information often depends on brokers and direct marketing representatives.   Many face a need to upgrade this process, making data collection more accurate, complete, consistent and timely.   Efforts have been made to use the internet for this process.   These attempts have not taken hold for lack of a consistent standard for ceding company submissions.   Each underwriter has variations on data requirements and format needs.   The limited number of transactions also hinders such standardization efforts.

Further, it is increasingly important to control marketing and sales costs.   Although hard markets provide temporary reprieve from some of the cost pressure, it is expected that we will soon see price shaving and a growing demand for greater efficiency to keep costs in line.

          Risk Management

Reinsurers must be able to assess the various characteristics of risk and price the portion of risk they assume.   Reinsurers concentrate on the least predictable and most volatile portions of risk.   This demands a high level of skill and the best applications of experience data, technology and risk theory.

Today's risk analysis includes some robust new tools.   Sophisticated catastrophe models help to assess some of the largest risks with the least direct experience.   Other quantification tools have upgraded our understanding of exposure, even where the bases for determining exposure have changed (e.g., new estimates of asbestos exposure).   A twenty-year evolution in developing these tools has been accelerated by the increased demand for understanding an increasingly complex combination of risks.   Actuaries are becoming an integral part of the underwriting process.

Sophisticated risk models are now a basic requirement in the property catastrophe business.   Modeling is less prevalent in casualty lines where the analysis is more difficult to apply.

Analysis now includes more extensive coordination of risk taking by members of company groups, tracking group-wide risk acceptance to avoid too much exposure to particular kinds and locations of risks.   This has been accomplished by integrating underwriting systems within groups and monitoring overall risk portfolios for the groups.   Prior practices of maintaining multiple underwriting silos are being broken down.

These new tools are being combined into models of company financial dynamics to understand how a risk portfolio in aggregate can be expected to impact profitability and capital adequacy.    Such models can help management understand how risk influences profitability and provide guidance for determining the amount and types of risk that can be supported by available capital.   This analysis can help determine what risks to retain and what to reinsure with others.

A good understanding of portfolio risk, and risk/capital interaction, provides the basis for assessing the impact of prospective new business.   Some of the property catastrophe reinsurers have developed sophisticated underwriting tools to determine how a new piece of business will change portfolio risk overall.   This is more difficult for multiline reinsurers but establishing a direct link between underwriting new risks and portfolio risk management needs to be strengthened by all.

The developing technology, combined with experienced underwriting, can aid the determination of appropriate rates.   However, such tools do not guarantee improved results.   During the market softening in the mid to late 1990s, many of these tools were available but did not prevent managements from accepting risks at prices below indicated levels.   There is some expectation that reinsurers are paying more attention to the modeling results, at least while they are supported by a hard market environment.

There is also the risk that modeled results will be relied on too heavily, without full understanding of the models' limitations.   Model results are the outcome of assumptions and model algorithms.   They are only as good as the underlying assumption validity and the ability of the algorithms to simulate expected market behavior.   It is important to consider not just the end results but the patterns reflected in the process of producing the end results.   Good models, supported by the accumulated experience of a skilled underwriter, can improve understanding of risk.   Actions based on such models need to consider their relative predictability and, particularly, the potential for experiencing outlying adverse results.   In the end, sound business judgement will determine whether advances in modeling tools will be used effectively to dampen cycle amplitude.


In the   mid to late-1990s, investment returns offset underwriting losses, particularly in long-tail lines.   Equity market declines since 2000 have eliminated most of this offset and reduced asset portfolios supporting capacity.   For some, this problem has been compounded as assets supporting pension funds have suffered losses in value and earnings, requiring additional pension fund contributions.   Certain reserve deficits have added to the strain.

Today's investment markets require that reinsurers and insurers price for underwriting profit in short-tail lines and much reduced reliance on investment income in long-tail lines.   Yet, this does not mean that the influence of investments should be ignored.

To achieve the kind of stable profitability demanded by investors today, it is important that reinsurers better understand the risks and potential returns from investments, along with expected results of their insurance risk portfolios, in combination.   Financial models that incorporate risk performance should also measure the expected result of the combination of liabilities and assets in the company's balance sheet.


Both the insurance and reinsurance businesses have a long legacy of informality in contracting and record keeping.   Many improvements have been made over the years, but much still needs to be done.   Failure to finalize major contracts until well after coverage is in force, and even after unprecedented catastrophe losses, continues to pose serious risks (e.g., World Trade Center).   Lack of centralized data maintenance can result in serious errors and extensive delay in collection of premium and claim payment, particularly when several brokers, reinsurers and retrocessionaires are involved in coverage.   A complete, current picture of liability and exposure requires improvement in this area.

Pressure to maintain better data flows, for underwriting and other functions, is encouraging companies to integrate data systems.   Historically, systems for underwriting, claims and policy administration have often been independent.   Primary insurers have made considerable progress toward integration.   Reinsurers have lagged behind the integration trend.

Various attempts have been made to address the problem of multiple administration systems of the insurers, reinsurers and brokers involved in reinsurance programs.   Pooling, syndicating and retroceding business can result in many individual companies involved in particular risks and claims.   From time to time, efforts are made to consolidate information flows in facilities developed through cooperation of brokers and others.   These have generally fallen apart as individual organizations assess what is gained and given up in such cooperation.   Many pride themselves on high quality claims and administrative processes that they consider competitive differentiators.   The potential accuracy and cost advantages of pooled data have yet to overcome preferences for retaining individual participant control.

Some of the administrative complexity arising from dispersing risk through multiple layers of reinsurance and on through retrocessions is being reduced as the appetite for slips representing small bits of largely unknown risks declines.   Consolidation and the increasing size of reinsurers will result in the distribution of risk in larger shares with fewer layers of retrocession.   This will help to reduce the overall cost of the system.

Product Trends

Response to 9-11 included new restrictions on coverage (e.g., terrorism exclusions).   The reduction in capacity due to 9-11 losses, plus reserve strengthening for asbestos and other outstanding exposures, has inspired reinsurers to restrict the types and nature of coverage.   For example, liberal writing of proportional covers to cover lower levels of catastrophe exposures has been replaced by catastrophe exclusions and replacement of proportional covers with lower layers of excess of loss coverage.   Continuing proportional covers have been subject to loss corridors and caps.

Part of the fallout from concern for proper accounting for risk has been increasedscrutiny of finite and other financing structures for requisite risk transfer.   Finite risk mechanisms face a more challenging market as a result.  

This comes at a time when there is even greater need for creativity in risk finance design to cover risks more efficiently.   For example, many Caribbean jurisdictions face the constant threat of windstorm and earthquake catastrophes.   Rather than annual reinsurance arrangements, strongly influenced by current world market conditions that may have little to do with Caribbean risks, long-term combinations of financing and risk transfer could offer mechanisms   for more efficient smoothing of results.   Small company size often requires extensive reinsurance protection, both to avoid an unmanageable hit to loss ratios in a particular year and to avoid having losses in excess of the top end of coverage.

Although some new finance/risk product designs may be developed to address this market need, there are a number of practical restrictions.   Assuming the arrangement passes muster on the issue of risk transfer, such arrangements are generally limited to three years to gain reserve credit under accounting rules.   Also, reinsurers are reluctant to do longer-term arrangements simply because it is so difficult to anticipate market conditions and related social and political impacts.   The solution may come in a series of related three-year arrangements.

Another innovation that holds potential for improving efficiency in the spread of risk is risk securitization.   Catastrophe bonds have had limited success due mainly to their complexity, and the cost and time required to complete registration processes.   The financial markets do not seem ready to make this a significant part of their product line.   However, some insurers and reinsurers have used these mechanisms as part of their overall risk financing programs.

Operating Silos to an Integrated Team

The new analytics have not only focused on underwriting individual risks and monitoring developing risk portfolios.   Financial models help companies track operating cash flows that encompass all operations, including underwriting, loss experience, administrative expenses and investment results.   These models automate profitability results and return calculations.   They provide improved means for seeing how the whole operation works together, from a financial perspective.

Traditionally, insurance and reinsurance operations have tended to function as separate silos.   Each operating area has maintained separate data management systems.   Underwriting and claims operations have tended to have difficulty comparing trends using separate systems.   Particularly in some reinsurance operations, the recent trend has been to integrate the operations and information systems for all operating areas.   The financial models provide a picture of how the whole operation performs together.   More integrated operating systems and procedures provide the means for responding to the financial model results with actions that lead to improved financial results.

Reinsurers are developing line of business teams around this integrated approach.   The teams work together to improve business line results, but also help to improve the quality of the individual functions.   Thus, for example, claims experience is brought to bear more effectively on underwriting decisions.

The Changing Role of Brokers

Brokers are expanding their services to ceding companies to help them organize data in response to expanding data requirements. They are also more active in providing catastrophe modeling support and either doing, or contracting for, reinsurance optimization analysis.

Thus, the broker role is no longer just a marketing/trader role.   Brokers increasingly facilitate and complete underwriting analysis to help ceding companies present their needs to the reinsurers and to help reinsurers refine their risk analysis and product responses.   The demand for this assistance varies, with the U.S. and Bermuda markets tending to be somewhat ahead of London and Europe.

When it comes to the more complex coverage structures (e.g., finite) brokers get a mixed review from our respondents.   Some specialist brokers are very good at identifying client needs and suggesting alternatives to reinsurers.   Other try but don't suggest practical programs.   In these cases the finite specialists prefer to deal directly with the client, with a broker introduction, to design appropriate solutions.

Capital Allocation, Investor Interest and Transparency

Enron, WorldCom and other celebrated cases of the failure of corporate governance to control management practices have heightened awareness of the need to monitor business performance in general.   Adverse results in reinsurance markets have encouraged suppliers of new money to establish clearer guidelines for how capital is to be applied.   Managements are now held to a higher standard of understanding the risks they are insuring.   Better underwriting tools, supported by more complete information on risks, are responsive to these requirements.

Newer companies, without legacy exposures and with a combination of strong underwriting tools and experienced management, are particularly attractive to investors.   But, even such companies need a clear strategy of the risks they intend to insure and need to establish a record of maintaining underwriting discipline within such plans.

A Note on Industry Structure

Through the cycles of the 1980s, 1990s and into the 21st Century, the reinsurance industry has lost many small and several larger competitors.    At the same time, significant new money has entered the market (much of it in Bermuda).   Our survey provided fairly consistent views that the reinsurance industry would continue to concentrate, with a continuation of large multi-line insurers exiting the markets (e.g., Nationwide, St. Paul, Hartford, Kemper, AXA (reduced US business)).  

There has also been some consolidation through acquisition as some of the new reinsurers in the post-Andrew new-entry class combined with other organizations and other combinations (e.g., Ace Tempest Re, Endurance Specialty acquisition of LaSalle Re, Axis Specialty acquisition of Kemper D&O renewals).   However, several of the long-standing reinsurers, with legacy exposure, have not been acquired.   New players appear to prefer to avoid legacy exposure and to grow by acquiring business in the marketplace.

Lloyd's reported an 18% growth in capacity to $23.2 billion at the beginning of 2003.   But, the number of syndicates trading declined by 15 in 2002 to 71, mostly the result of mergers. Some of the new money coming into the market post-9-11 has been added to Lloyd's market participants, but some of those participants have also added capital to new or existing reinsurance vehicles outside the Lloyd's system.    Lloyd's appears to be losing ground as a reinsurance center compared to Bermuda's growth.

The more successful new companies have tended to have very clear strategies, often   fairly narrow in scope.   Some have focused solely on property catastrophe and then added some diversification in lines with very similar operating requirements.   It appears that trying to mix reinsurance with various primary insurance lines is a particular challenge.

The need for spread of risk requires a significant number of competitors.   Reinsurance is not likely to be dominated by a hand-full of global players anytime soon.   Indeed, concentration of business in the largest companies has declined.   But, the minimum size has increased dramatically.   Whereas some of the early Bermuda entrants started with capital of $25 to $50 million in the early 1980s, and by the late 1980s capital of $250 million was considered quite large, many now believe that the minimum capitalization is $500 million or even $1.0 billion.

Bermuda v. the World

The Bermuda market has grown dramatically with the influx of new capital post-Andrew and post-9-11.   Most of these recent entrants see high levels of modeling capability, demands for more thorough underwriting information, application of high levels of underwriting skill and capital deployment discipline as fundamental to their strategies.   Each has a somewhat different selection of markets they participate in, but most have a very clear sense of where they participate and where they don't.

The physical limitations of the island have produced another common outcome.   The inability to bring in large numbers of people have forced the Bermuda players to select staff carefully, bringing in high levels of skill and significant versatility.   This has helped to maintain low overhead cost levels.

The Bermuda market also has the advantage of operating with tax advantages, in a jurisdiction with a strong reputation for stability and reliability.   Because of the recent entry of many of the companies, the Bermuda industry has relatively limited exposure to past   liability.

The U.S., London, European and other world markets have introduced many of the advances in underwriting technology, new efficiencies and capital allocation discipline.   London retains a much larger infrastructure of reinsurer staff and related practitioners.   But many of these reinsurers also carry significant legacy exposures and vestiges of less efficient and effective practices of the past.   As a geographic market, Bermuda appears to have an edge at least for now, in part based on skill but also based on preferred market circumstances.

Market Stabilization or Just Another Cycle

The consensus of our survey was that despite the many improvements that could limit the potential for descent into a deep and long soft market, there is still significant risk that irrationality could prevail and the cycles of the past could be repeated.

We are perhaps at a point in history where forces are lined up to limit the depth and duration of a soft market.   Catastrophe and underwriting modeling tools, better underwriting information, improved underwriting skills, investor constraint on application of capital, improved claims management and administration, tighter controls on overhead costs, improved profitability analysis and other factors provide an impressive array of tools for understanding pricing requirements, avoiding excessive risk taking, and maintaining efficient operations.   What prevents the maintenance of market discipline to stay at adequate price levels?   Aren't reinsurers going to be persuaded by the success of those whom they see as having maintained risk selection discipline throughout the last cycle (e.g., Renaissance Re)?

Many see a longer stable pricing period in part simply due to the depth of the hole faced by those with legacy exposures.   It is difficult to go back to a market share expansion game if your capital has been depleted.   But, established players are going to the equity markets for additional capital and generally being successful.   Thus, this constraint may not last long.

The most obvious reason for continued market volatility is the unpredictable nature of the underlying risks.   High severity, low frequency exposures result in some years with minimal losses and other years with extreme losses that can wipe out years of accumulated profits.   The best analysis will not remove inherent volatility in the experience.

Another of the countervailing factors is year-to-year risk patterns.   For example, respondents have noted that if 2003 has limited catastrophe experience, pressure will build for rate reductions.   Pricing indications may not change significantly, but ceding companies are reluctant to pay current high prices against the prospect of large catastrophe losses in future years.

Another factor is available capacity.   New money seems to be available so long as pricing is strong.   Too much capacity will tend to push prices down as the added capacity seeks a market.   As companies attempt to keep their capacity in active use, there will be increasing temptation to accept lower expected returns rather than have idle capacity.

Of course, it is likely that the inevitable softening of the current market will harden again on the heels of some major event or series of events, as occurred post-Hurricane Andrew and post-9-11.   This introduces a new factor into the equation.   It might be simply a matter of an increase in the expected scale of potential losses.   Or it may be new areas of exposure or new combinations of known exposures.   It is difficult to anticipate the impact of such a change in exposure conditions.  

As noted in the Guy Carpenter World Rate on Line Index, the jump in prices in 2001-2002 has been considerably more restrained than in 1992-93.   This has been attributed, in part, to better reinsurer understanding of the underlying risks leading to less panic response in pricing to an unprecedented loss.

Even if hard market pricing continues, that provides no protection against poor underwriting decisions.   There is concern that the flood of new capital in the market post-9-11 is encouraging some to pursue business too aggressively in order to put the masses of new capital to work quickly.   Particularly in the long-tail lines, it is too soon to determine if current "hard market" pricing is adequate to cover the ultimate losses.

Many see the winners and losers in the current cycle as divided between those with a clear, logical market strategy and those who lack a clear focus.   Those who stick to lines of business they know, and who have similar competitive success requirements, have a better chance to set appropriate limits on risk selection and price and term concessions.  

Some see the emerging Bermuda model as offering advantages.   Certainly tax advantages and the requirement to manage business with carefully selected staff provides a good start.   Heavy reliance on strong analytics and high underwriting data demands provide a good information base for maintaining discipline.   Further, although the Bermuda companies have a range of line of business focuses, they tend to define those focuses carefully and relatively narrowly.

Some of the Bermuda companies are seeing a trend for brokers to seek markets in Bermuda ahead of London and other European sources, which may improve the quality of business bound in Bermuda.

In reinsurance, problems arise from a combination of ignoring data, not interpreting data appropriately, and not applying basic insurance theory.   If a particular risk, with a high attachment point, has been underwritten for three years at a premium level, properly determined, in the vicinity of, say, $500,000, and no losses have been experienced, there will inevitably be pressure to reduce rates.   Recent (last three year) experience would indicate that a rate of $300,000 would still lead to significant profit on the individual account.  

Such an attitude, applied again and again, in a competitive environment where companies seek increased market share, leads inevitably to an under-priced market.   Risks cannot be considered in isolation, insurance is a mechanism for pooling risks because individual losses are unpredictable.   The above approach uses the median or mean of a distribution as an intuitive proxy for appropriate rate level.   But, the skewed shape of loss distribution curves mean that risks have to be priced by reference not only to the mean but also higher moments of the assumed underlying distribution.   It is the viability of premiums for a portfolio of risks over time that matters, not the adequacy of rates for individual contracts.

It is likely that a softer market will be brought about by unsophisticated underwriters offering rates that appear viable in isolation, but result in under-funded portfolios.   This is just cashflow underwriting.  

A major change experienced in reinsurance in the last fifteen years is the rise of Bermuda to become the third largest market in the world.   Bermudian-based underwriters require higher standards of information from cedants; more intensive and sophisticated risk and portfolio analysis is carried out in Bermuda.   This enables Bermuda to select and price risks appropriately.   This means that as the market softens, underwriters relying on intuition and gentlemen's agreements will not have the whole market to participate in, as once was the case.   Lax underwriting based on broad participation to achieve market share will no longer achieve average market returns.    Bermudian-based companies, and others exercising pricing discipline, will have that share of the business that is priced appropriately, the business remaining being subject to adverse selection but still searching for lower rates.   There is plenty of anecdotal evidence of underwriters from the London market generally undercutting the Bermudian market, or not pricing risks appropriately.

Despite significant progress in developing underwriting and pricing tools, much remains to be done.   For example, the market is a long way from assessing aggregate exposure in casualty lines of business effectively.   Efficient ways of determining capital requirements are in their nascency.  

Further, the synthesis of financial and insurance structures (eg. catastrophe bonds), has not taken off as expected.   The main impediments to such products seem to be industry orientation and jargon; the financial world and the insurance world speak different languages, and no-one has yet been able to bridge the communication gap effectively.   But, despite the language differences, the finance world and the insurance world are really both in the same business, buying and selling capital and risk.   Unitizing risk, and sharing risks in ways that unbundle the components of risk and finance, will promote market efficiency from which we all gain, except those determined to preserve vested interests in an inefficient market.  

The theoretical paradigms underpinning insurance practice also need development.   Insurance is essentially a non-linear game, but linear models tend to be used.   There has been a development toward generalized linear models and some bootstrapping techniques, but one hundred years or so of mathematical theory is being ignored when it really matters.   It is hard to bring academic development into the business world, but it was noticeable after 9-11 that some attempts were being made to model terrorist activity using some ideas from non-linear game theory.   The actuarial profession needs to deepen and modernize the mathematical components of its curriculum, and not stop at teaching Monte Carlo techniques.

So, we have an impressive array of tools to fend off pressures to let market softening force pricing below indicated minimums.   And, there is the prospect of continuing improvement of these tools.   However, the outcome of the current cycle will depend on whether company managers use the tools effectively to maintain needed discipline.   Some will likely respond to the "boiling water" of price reductions and leap out of the pot in time.   Others will likely be cooked!

Appendix A

Individuals surveyed for this paper were associated with one or more units of the following companies and groups:


ACE Ltd.


Allied World Assurance

Axis Specialty

Chubb Re

Endurance Re

Fairfax/Odyssey Re

General Re

Max Re

Montpelier Re

Platinum Re

Partner Re

Renaissance Re

Swiss Re

XL Capital



Back <<<