Nearly five years ago–back in the July 23, 2001 edition ofNational Underwriter–we stepped into an actuarial time machine andforecast the “Top 10 News Stories of 2005,” as viewed from theperspective of the property-casualty insurance industry. In thisfollow-up piece, we boldly examine the accuracy of our forecastsand their implications, as well as peek ahead to spotlight issuesthe industry might be facing five years down the road.

1. Underwriting Cycle Heads South Again:

We might just as courageously have predicted the sun wouldcontinue to rise and set, given that the industry seems determinedtime and again to use its competitive pricing tools to erodehard-won profit margins.

Prior to Hurricane Katrina, 2005 showed a repeat of thispattern. After an industrywide underwriting profit in 2004, pricelevels slipped by 3 percent for commercial lines in the second andthird quarters of 2005, compared to the corresponding periods oneyear earlier, according to the Tillinghast commercial lines pricemonitoring survey, CLIPS.

However, over the past five years, we have seen encouragingdevelopments related to several underwriting cycle implications wediscussed in our 2001 article.

o Price Monitoring. Back in 2001, we warned of the dangers toinsurers that establish their pricing based primarily oncompetitors' actions, or adjust loss-reserve adequacy to soften thedeterioration of reported underwriting results. We have beennoticing, however, that many insurers are introducing andstrengthening price-monitoring tools and paying greater attentionto the results.

We also see an increasing number of executives, boards ofdirectors and audit committees that insist on receiving unvarnishedbriefings from loss-reserving actuaries. But more work is neededwithin companies to create full feedback loops that allow–orforce–managements and underwriters to learn from the results ofpast actions, and to prompt them to make better decisions the nextgo-around.

o Segmentation. While insurance is a business of averages, noone policyholder is “average.” In 2001, we observed the developmentof various techniques for refining market segmentation andindividual account analysis, and predicted that insurers usingthese tools in a disciplined way would outperform competitors.

This trend has taken hold in a major way in personal auto, wherethe use of predictive modeling is pervasive and the search foradditional explanatory variables continues. Innovative companiescontinue to lead the pack–witness Progressive having increased itspersonal auto marketshare from 2.6 percent in 1995 to 4.7 percentin 2000 and 7.3 percent in 2004.

The use of new data elements and analytical tools to definesegments for targeted marketing, pricing and underwriting has alsomigrated to homeowners. We are seeing an increased level ofpredictive modeling exploration in commercial lines as well,beginning with Main Street accounts and moving to larger accountsizes and specialty coverages.

2. Mini-Cats Cost Insurers $100 Billion:

We correctly predicted “an unusually large number of tropicalstorm systems,” but we doubt anyone expected that those naminghurricanes would need to dip into their supply of Greek letters tolabel the final few storms of 2005–and, of course, we won't try tocategorize Katrina, Rita or Wilma as mini-anything.

Fortunately, other predicted catastrophes did not occur, and itappears the $100 billion cost estimate for 2005 was pessimistic.Still, Katrina's $34 billion-or-so insurer price tag indicates that$100 billion in catastrophe claims in a single year in the UnitedStates is by no means outside the range of probability. Superimposethe Asian tsunami or the devastating October earthquake in Pakistanto a heavily-populated U.S. city or coastline, and the cost wouldbe beyond counting.

We earn mixed scores on predictions of industryimplications:

o Predicted Solvency Threat Didn't Materialize. While manyinsurers took sizeable earnings blows from the storms, the numberof insurers facing an immediate capital crisis has been mercifullysmall. Certainly risk management techniques adopted by insurers inthe 13 years since Hurricane Andrew must be credited with helpingprepare some companies.

o Public Outcry Muted. We aren't seeing restricted marketsgiving rise to a predicted outcry for full coverage of catastropheperils at affordable prices, but pending litigation in Mississippiillustrates a worst-case scenario–retroactive coverage of catperils at no additional cost. Plus, we do expect property insurancecost and availability will be one challenge of the Gulf Coastrebuilding effort.

o Catastrophe reinsurance market tightens. Nearly weekly, a newcatastrophe reinsurer springs up in an offshore domicile,indicating there will be plenty of reinsurance capacity, butperhaps not at prices or terms primary insurers would like.

o Securitization lags. A predicted increase in securitizationdeals has not developed as rapidly as we anticipated, perhapsbecause the deals are complex to structure and traditionalreinsurance has remained available.

3. Internet Sales Of Personal Lines Insurance Dominate YouthfulMarket:

Our 2001 story included a prediction that by 2006 the Internetwould account for nearly 80 percent of the young-adult market andalmost 40 percent of new and renewal transactions in the personallines market overall.

Although the exact figures quoted in the story have not beenachieved, Forrester Research estimates that nearly one-quarter ofauto insurance buyers apply for their policy online, and aboutone-fifth of the remainder conduct at least some comparisonshopping or other information-gathering online.

Some of our anticipated implications were on target–among them,faster cycle-times for product innovation and price changes, aswell as growing price competition as more shoppers make onlinecomparisons. But we inaccurately predicted that online shopperswould become increasingly aware of jurisdictional regulatorydifferences and limits on insurer flexibility to price, classifyand underwrite business in some areas.

4. Information-Gathering Devices Produce Flood Of Data:

The 2001 article anticipated that the smart chips routinelyinstalled in cars, appliances, phones and other devices wouldproduce information on usage patterns, and would ultimately gainacceptance in pricing and underwriting–especially with the lure ofpremium discounts coaxing low-risk consumers to allow insurers touse data collected from these sources.

Technological innovation continues to point in thisdirection–onboard computers signaling the location of automobilesor appliances that “ask” the home office for maintenance when theyneed it.

Even data on when, where and how people interact with Web sitesis being gathered and analyzed. The ability to track activity isbecoming pervasive–and invasive, with involuntary and unknowingdownloads of spyware to our computers. The acceptance ofinformation sharing, especially among young adults, is growing.

We anticipated several implications of this trend, most of whichstill ring true but are somewhat delayed. For example, there hasn'tyet been much insurer activity aimed at using this new data toestablish rating classes. But we understand some insurers areoffering time-sensitive coupons and other incentives to capturebusiness of online shoppers–based, in part, on shoppers' behaviorin navigating insurer Web sites.

5. Stat Data To Be Abbreviated:

The predicted story: Several large U.S. insurers rally around aproposal that the National Council on Compensation Insurance andthe Insurance Services Office dramatically reduce the variety andvolume of statistical data collected.

While pressure is on for statistical agents to justify theefficacy of, and streamline, various data collection processes,this pressure has not produced a wholesale reduction in the varietyand volume of statistical data that is collected. Several trends,in fact, point in the direction of increased data collection:

o Increased regulatory attention on data quality/integrity.Regulators show greater interest in real-time information on costtrends and market behaviors.

o Sophisticated market segmentation research presentsopportunities. Whether stat agents do research themselves or allowcompany members to manipulate online databases, the stat agents andtheir clients will need more information, not less.

6. Solvency Regulation Moves To Washington; Rating Freedom ForMiddle-Market Accounts:

While the story of federal solvency regulation has not unfoldedexactly as we might have predicted, we have been watching severalrelated national trends:

o Sarbanes-Oxley, enacted in response to several significantfinancial reporting scandals, is creating marked changes in the waymany insurers–and their professional advisors–gather, interpret andreport financial data. If these changes ultimately result in moreaccurate, timely, complete and meaningful information, the abilityof states, federal regulators and the market itself to monitor thesolvency of insurers would indeed be enhanced.

o Enterprise risk management is emerging strongly as a soundbusiness management tool, notwithstanding that many companiesinitially adopted some elements of ERM primarily as aSarbanes-Oxley compliance measure. The impact of ERM is notparticularly to enhance the ability of outsiders to evaluate thesolvency of an entity but for management to have better informationand control regarding key business decisions and actions that willaffect future performance and solvency.

o Rating agencies increasingly are serving as de factoregulators of financial strength and performance for much of themarketplace. Today, few insurers make major strategic or financialmoves without considering how the agencies will respond.

The other half of the predicted story focused on deregulationfor commercial accounts with over $100,000 in premium.

We perceive that market forces, rather than state rateregulation, are indeed the primary determinant of prices for thisbusiness. However, the driver of this trend has been themarketplace itself, rather than regulatory change. And, into thefuture, predictive modeling and segmentation will foster stillfurther competitive innovations in commercial lines insurance.

7. Banks Integrating P-C Products Into Service Offerings:

This prediction anticipated that banks would have by noweffectively incorporated the distribution of personal linesproducts of partnering insurers into their suites of products andservices.

In fact, we do not yet observe a broad change in theeffectiveness of such partnering, cross-selling and integrationinitiatives with p-c insurers, although banks continue to expandinto the insurance brokerage business.

8. Jury Awards Skyrocket:

Back in July 2001, we predicted that in 2005, class-action suitsdirected at auto, tire, computer, artificial sweetener, airline andinsurance firms would total $100 billion.

Tillinghast's 2005 tort cost study indicated that the cost ofthe tort system continues to grow at a faster rate than the U.S.economy, but that the growth rate had moderated somewhat fromhistorical trends. While tort reform remains an issue in front ofCongress, headline stories have been more localized than in the2001 predictions (for example, medical malpractice crises invarious states).

While this remains more of an evolution than a sudden story,direct implications for p-c insurers align with ourexpectations:

o Continued demand for liability insurance, and continuedchallenges in the underwriting, pricing and reserving for liabilitycoverage.

o Cost of defense rising faster than the cost of settlements.Leading insurers have restructured relationships with defensefirms–revising financial arrangements and becoming more involved inthe control of cases.

o ERM heads in a positive direction. We anticipated that theincrease of tort issues would have implications for the growingtrend of corporate entities to embrace a more holistic,enterprisewide understanding and management of risk. For example,we predicted corporations would demand risk-transfer mechanismsthat more completely address the consequences of class-actioncases, such as loss of reputation and marketshare.

This is a relatively difficult trend to observe and measure, butwe perceive that a considerable proportion of ERM energy is headedin a different and positive direction–toward identifying andaddressing circumstances that give rise to problems in the firstplace, not just to financing and handling problems when they dooccur.

9. Plaintiff Law Firms Positioned In Latin America andJapan:

We are not yet observing a sea change in the legal climates inother jurisdictions toward a U.S.-style tort system. If and whensuch changes begin to occur, they will produce both opportunitiesand challenges for the p-c insurance industry.

10. Acquisitions Hit Record Pace; Outstripped ByDe-Acquisitions:

We have not seen a significant predicted shift in the number orcreativity of divestitures, spin-offs or breakups of parts ofbusinesses that no longer fit insurers' strategies, although somemajor insurers have taken exactly these steps.

The world has indeed changed in the past five years. Some of thechanges were very predictable, and some more surprising.

We will be able to make the same statement five years fromtoday. The challenge for insurance company management teams is tobe able to answer “yes” to the following question: “Are ourdecision-making, strategic, operational, risk management andresponse systems better equipped today to anticipate, monitor andaddress emerging external and internal trends, changes and problemsthan they were five years ago?”

Companies that can answer this question in the affirmativeimprove the odds of surviving and thriving despite theuncertainties ahead.