Good Things Come In Packages For Insurers

|

Storm losses undermine CMP profits, but Chubb stands out withstellar results

|

By Susanne Sclafane

|

A small underwriting profit in 2003 was wiped away by Floridahurricane damage in 2004 for the third-largest commercial insuranceline–commercial multiple peril. However, although the line ranksbehind other liability and workers' compensation in terms of sizein the world of commercial insurance, its stature is undiminishedin the eyes of some commercial underwriters.

|

“It would be a mistake not to understand that multiple perilinsurance is the hub around which the spokes of umbrella, workers'compensation and auto are placed,” said Steven Pozzi, chiefunderwriting officer for the Chubb Commercial Insurance.

|

“If we write the multi-peril when we write those other lines, weare in a much better position to effectively address the insuranceneeds of the policyholder and improve the quality of the risk weunderwrite,” Mr. Pozzi said.

|

For the industry–and for Chubb, in particular–good things comein packages, large and small. Early estimates based on acompilation of annual net premium, loss and expense data put the2004 industry combined ratio at roughly 101–just one point overbreakeven. And in the prior year–2003–an industry aggregatecombined ratio of 99.7 revealed a slight underwriting profit.

|

Chubb's results were even better, with the Warren, N.J.-basedinsurer posting an eye-popping 69.6 combined ratio in 2004. Chubb'scombined ratio was not only lower than the loss ratios of manycompetitors, but nearly 18 points lower than the enviable 87.4combined that Chubb posted in 2003.

|

Across the industry, net written CMP premiums reached nearly$28.5 billion–up roughly 4 percent in 2004. That growth rate wasjust about half the 7.9 percent industry figure recorded in 2003,and nearly 10 percentage points below the 14.5 percent writtenpremium jump recorded in 2002.

|

In spite of declining premium growth and a 9 percent spike innet incurred losses in 2004 (fueled by rising property claims),falling loss levels in the prior two years helped keep the industryCMP loss ratio near the lowest level recorded for the seven-yearperiod reviewed by National Underwriter.

|

Although the 2004 jump in losses nearly offset the two-yeardecline recorded for 2002 and 2003, with net earned premiumssoaring nearly 33 percent over the three-year period, the industrynet pure loss ratio (excluding loss adjustment expenses) was 51.5in 2004–nearly 15 points better than the 1998 loss ratio recordedat the start of our review period.

|

To Package Or Not To Package?

|

The year 2004 marked the fourth-straight in which packageresults were better than monoline liability and property totals,according to an analysis prepared using data from NationalUnderwriter Insurance Data Services. (See Graph 1.)

|

For this comparison analysis, NU combined net premiums, lossesand expenses for the fire, allied, inland, earthquake and burglarylines to derive loss ratio and combined ratio results which werefer to as “monoline property” results. Monoline liability resultson the accompanying graphs are results for the “other liability”line of statutory filings.

|

Our comparison graphs, linked below, reveal the followingtrends:

|

o For the seven years reviewed, monoline property loss ratioswere always better–consistently lower–than package policy propertyresults. However, the gap between monoline and CMP property startedto close in 2001, when substantial property loss ratiosimprovements were recorded, whether it was written in a package ornot. (Graph 2)

|

o Monoline liability had worse loss ratios than CMP liabilityfor the most recent five years. (Graph 3)

|

o Since liability differentials are greater–and because CMPproperty results improved so much in recent years–putting propertyand liability results together, CMP loss ratios–which were worsethan weighted-average monoline loss ratios prior to 2000–are nowconsistently better than monoline loss ratios. (Graph 1)

|

o For combined ratios, the story is different. The expensecomponent of monoline combined ratios tends to be lower than CMPexpense ratios. As a result, combined ratios for package policiesare consistently worse than monoline policy combined ratios. (Graph1)

|

The shrinking gap between CMP property and monoline propertyloss ratios (Graph 2) isn't fully explained by strong net premiumgrowth for CMP property in recent years. As analysts might guessbased on intuition and buyer profiles, net written premiumgrowth–accumulating to 26 percent in 2002 and 2003–significantlylagged monoline premium growth, which was 43 percent over the sametwo years.

|

“Much of commercial multiple peril is made up of businessownerspolicies, which are available only to smaller, much less complexaccounts,” said Tom Quinn, vice president of underwriting servicesat MarketStance in Middletown, Conn. “During the post-2000 firming,pricing on this business increased substantially less than onlarger accounts, due at least in part to more pronouncedcompetition in the small commercial market,” he said.

|

In 2002 and 2003, however, net incurred property losses fellnearly 17 percent on the package policy side, while monolineproperty losses fell only about 5 percent.

|

“Could underwriting discipline be the difference–[explain] thecrossover between monoline and package profitability?” Diana Reitz,managing editor of FC&S Online (the National UnderwriterCompany's online coverage interpretation source), in response to ane-mail query.

|

Compared to property, an even more pronounced premium-growthdifferential is evident on the liability side in recent years, withother liability premiums soaring more than 69 percent in 2002 and2003, compared to a 20 percent jump in liability package premiums.However, towering losses for monoline liability–rising more than 50percent in 2002 and 2003, compared with only a 2 percent rise inCMP-liability losses–kept monoline liability loss ratios well aboveCMP liability in recent years.

|

“The other liability line on the statutory blank, in addition tocommercial general liability, incorporates umbrella, errors andomissions, and most notably directors and officers liability,” Mr.Quinn noted. “A sharp deterioration in D&O results doubtlesslyis a factor in the underwriting results for 'other liability,'” hesaid.

|

Reserve additions may be a factor as well. According to resultsrecently presented by analysts at New York-based Standard &Poor's, nearly $18 billion in asbestos, construction defect andD&O reserves were added to other liability from 2002-2004. Thecomparable CMP amount was $4.4 billion.

|

While the relative positions of package and monoline resultsremained the same in 2004 as they did in the two prior years,double-digit premium declines that showed up for the monolines,housing larger or more complex risks, did not appear among packageresults.

|

Other liability net written premiums fell 14 percent in 2004,and monoline property premiums dropped more than 10 percent. Incontrast, CMP premiums increased 4 percent overall–rising roughly 2percent for the property piece and 8 percent for liability.

|

An Imperfect Comparison

|

NU's monoline versus package comparison is only as good as theannual statement data that went into it–and carriers don't allnecessarily follow the same rules when they record CMP results.

|

Beyond BOP policies, “the rest of what is reported as CMP can bea hodge-podge of coverages whose only common denominator is thatthey've been brought together to qualify for a package discount,”Mr. Quinn notes. Crime coverage (employee dishonesty, forgery andalteration), for example, could be included in a package, as couldauto-related liability coverage such as non-owned autos orgaragekeepers' liability, he said.

|

Analysis By State: Florida A Wipeout

|

An analysis of direct loss ratio results by state shows thatwhile California, Michigan and Massachusetts produced the best lossratios for insurers in 2004, Florida writings punished CMP insurerswith direct losses coming in at nearly twice the level of directearned premiums.

|

The 195.4 direct loss ratio for the largest Florida writer,Zurich, looked good when compared to results for the tenth-largestFlorida CMP writer, Allstate–a direct loss ratio of 732.0. Reportsof results for the top-10 writers in California, New York andFlorida–the three largest states based on CMP premiums–areattached.

|

Chubb Leads The Profit Pack

|

In a nationwide analysis of direct premiums and loss ratios forthe 25-largest CMP writers, St. Paul Travelers leads in volume,with over 9 percent of the CMP market. The largest writer alsoproduced one of the lowest 2004 loss ratios–39.5, compared to anindustry level of 52.9.

|

Second-ranked Zurich wasn't as fortunate, with its loss ratiodeteriorating more than 36 points to 92.5, owing in part to thegroup's large Florida presence.

|

However, while Zurich's loss ratio results deteriorated innearly every state (California was a notable exception), and thecompany recorded loss ratios over 100 in 17 of them, thefourth-largest insurer, Chubb, reigned at the other end of thespectrum, reporting loss ratios under 40 in 48 out of 57territories.

|

With a countrywide CMP loss ratio of 24.2, Chubb's result fallsfar below any of its competitors. An analysis of Chubb's results byregion reveals that a single-digit loss ratio for “other alien”territories (outside the 50 states) as well as a comparatively lowFlorida marketshare explain part of the carrier's good fortune.

|

Mr. Pozzi needs no such data analysis to explain the results.According to him, Chubb simply has a different view of thebusiness.

|

The CMP package “is the core product for how we deliver[insurance to] our niche market segments,” Mr. Pozzi said, notingthat the majority of Chubb's multiple peril policies are writtenfor risks with complex exposures. Examples, he said, are lifesciences companies, technology firms, energy entities, metalworkersand law firms, which make up the bulk of Chubb's multi-perilbook.

|

Many competitors, he said, deliver commodity-type packages tothe majority of their insureds–”and the only way they can respondto specific needs of their customers is by manuscripting,” hesaid.

|

He explains that Chubb's base CMP policy–known as “Customarq”–isa specialty policy in and of itself. Special features include aproperty valuation approach that includes cost of replacement atthe same or another site and selling price of finished stock, hesaid. The policy also grants automatic limits of insurance foremerging Internet exposures, he said.

|

In addition, Customarq has customized contracts for nicheclasses like life sciences accounts, he said. Marketing material onChubb's Web site, for example, notes that spoilage of abiotechnology company's cell cultures during a power outage wouldbe covered due to a broad property definition that includesresearch and development property, together with a “change incontrolled environment” contract.

|

“The thing is not to confuse the fact that simply because it's apackage product, that it is non-sophisticated and therefore grantedto anyone who comes in looking for insurance,” Mr. Pozzi said.

|

Once you have developed a highly sophisticated package product,you must treat it with a great deal of care on the underwriting andpricing fronts, he continued–”because you're granting wider-rangingcoverages than most of the competition would as a baseproduct.”

|

Explaining Chubb's dramatic turnaround in underwriting resultsin 2002, Mr. Pozzi admits “the market helped,” because the climatewas right for Chubb to more adequately price exposures.

|

In the late 90s, Chubb–and the industry as a whole–”gave up moreon the underwriting than we should have,” falling into the trap ofcompeting on price. Responding before most competitors, Mr. Pozzisaid his company culled those accounts that could not beappropriately priced.

|

(According to NU data, Chubb still managed to grow net CMPpremiums 52 percent in 2002, while industry CMP premiums rose lessthan 15 percent.)

|

AIG Impact

|

With every data compilation comes decision-making, and NU'sdecision to show industrywide net combined ratios in our graphscarries with it one inherent problem created by recent events.

|

While net combined ratios (including the impact of cededreinsurance) give a better sense (than direct ratios) ofunderwriting results that contribute to bottom-line net income, thefact that net premium and loss information for the largest membersof American International Group have not yet been filed may have adistorting impact on the overall 2004 combined ratio result of101.0 estimated based on carrier results reported through May2005.

|

While the directional impact of the missing AIG insurers can'tbe determined, we note that in prior years the AIG companieslowered overall industry ratios. In 2003, with $0.5 billion in CMPnet premiums, AIG moved the industry underwriting result from aloss to a profit. Excluding AIG, the 2003 combined ratio would havebeen 100.1, rather than the 99.7 reported for the industry inaggregate.

|

Given AIG's volume in this line, restatements of AIG results forany of the prior years will similarly swing the overall industryresults.

|

|

SIDEBAR with Data Insights logo

|

Flag: Editor's Note

|

Head: About the Data

|

Source: NAIC Annual Statement Database viaNational Underwriter Insurance Data Services/Highline Data.

|

National Underwriter Insurance Data Services (formerly ThomsonFinancial Insurance Solutions and Sheshunoff Information Services)is part of Highline Data LLC, the data affiliate of Highline Media,parent company of this magazine.

|

For information about NUIDS' products, contact Chris Rogers at617-441-5976.

|

Graphs: Line graphs for this article wereprepared using information from the Insurance Expense Exhibit, PartII.

|

Industry aggregate figures calculated by NUIDS used for 2003 andprior; 2004 ratios are preliminary estimates based on totalpremiums, losses, expenses reported by 3,121 individual insurancecompanies as of mid-May. The largest members of AmericanInternational Group have not filed net data and are not included inthe 2004 total.

|

Pure loss ratios exclude loss adjustment expenses; combinedratios include LAE, underwriting expenses and dividends.

|

All ratios are net of reinsurance.

|

To calculate weighted average monoline CMP ratios shown on Graph1, the percentages of property and liability premiums inindustrywide CMP figures were used as weights and applied tomonoline property and monoline liability ratios in each year.

|

Charts: The tables for this article wereprepared using direct premium and loss information from individualstate pages of insurer annual statements.

|

Direct premium growth figures for individual insurance groupsreflect organic growth only. In other words, for any group thatmade an acquisition in 2004, premiums for the acquired company wereincluded in both 2003 and 2004 to calculate growth percentages.

|

Direct data by state as reported to regulators by AmericanInternational Group is included in the charts.

|

Art: Tearing ThePackage Open

|

A National Underwriter analysis reveals that industry commercialpackage loss ratios fell below loss ratios for individual liabilityand property coverages in recent years, reversing the positions for1998-2000. While package policy property results are consistentlyworse than monoline property results in all seven years, ashrinking property gap and a wide gap between CMP liability andmonoline liability loss ratios explain the turnaround.

|

Quotebox, with Pozzi mug

|

“It would be unfair not to give [commercial multiple peril] itscredit. It is the hub of an account.”

|

Steven Pozzi, CUO, Chubb Commercial

|

Flag: Fast Facts

|

Head: CMP Commission Ratios

|

Among the top 25 writers of CMP based on direct writtenpremiums, ratios of direct commissions to direct premiums variedfrom a low of 3.8 percent to a high of 21 percent.

|

The top five ratios were:

|

o 21.0 for Maguire Group (Philadelphia Consolidated)

|

o 20.7 percent, Erie Insurance

|

o 20.5 percent, Hartford Fire & Casualty

|

o 20.3 percent, White Mountains

|

o 19.5 percent, Harleysville

|

For Data Insights Charts

|

Source: NAIC Annual Statement Database via NationalUnderwriter Insurance Data Services/Highline Data. For information,contact Chris Rogers at 617-441-5976

|

Net Pure Loss Ratios and Combined Ratios

|

PROPERTY AND LIABILITY COMBINED

|

Package vs. Wted. Monoline

|

Flag: Graph 1: Expense Ratio Impact

|

What CMP writers gained in terms of a loss ratio advantage in2001-2004, they gave up on expenses. Higher package expense ratioshave kept monoline combined ratios below package combined ratios.While monoline liability and property expense ratios averaged 25and 28 in the last three years, the three-year average expenseratio for CMP was 32.

|

Net Pure Loss Ratios

|

PROPERTY

|

Package vs. Monoline

|

Flag: Graph 2: Gap Closes On Loss Ratios

|

The industry loss ratio for the property portion of CMP was only3.6 points higher than loss ratio for monoline property in 2004,closing a gap that was more than 10 points in 1999 and 2000.

|

Net Pure Loss Ratios

|

LIABILITY

|

Package vs. Monoline

|

Flag: Graph 3: Advantage Package

|

The industry loss ratio for the liability portion of CMPaveraged 17 points better than the industry loss ratio for theother liability line over the last three years. Reserve issues andunfavorable results for D&O embedded in the other liabilityline are likely contributors to difference.

|

Net Combined Ratios

|

CMP TOTAL

|

Industry vs. Chubb

|

Flag: Graph 4: 30 Points OfProfit

|

“The thing is not to confuse the fact that simply because it's apackage product that that means non-sophisticated and thereforegranted to anyone who comes in looking for insurance,” said Chubb'sSteven Pozzi, explaining the carrier's better-than-averageresults.

|

Top 25 Chart:

|

Head: Direct CMP–Top 25 Writers ($000)

|

Flag: Storm Damage

|

For the top 25 CMP writers (based on 2004 direct writtenpremiums), direct loss ratios ranged from a low of 24.2 for Chubbto a high of 132.4 for Allstate in 2004. The percentage of premiumsthat an individual insurer wrote in Florida was one key factordetermining where its loss ratio fell within the range.

|

50 (or 57 state) chart:

|

Head: Direct CMP Results By State ($000)

|

Flag: CaliforniaDreaming

|

CMP insurers continued to like California in 2004, writing 13percent of their premiums in the Golden State. While the attractionseemed well-founded, with the state producing one of the lowest2004 CMP loss ratios within national borders, Florida writingspunished CMP insurers with direct losses coming in at nearly twicethe level of direct earned premiums.

|

Footnote: The loss ratio results for “Other Alien”territories are distorted by negative incurred losses reported byACE Ltd., the third largest “Other Alien” writer, in 2004. RemovingACE from the totals produces direct losses ratios of 32.0 and 41.4for 2004 and 2003, respectively.

|

California chart:

|

Head: Direct CMP Results for State ofCalifornia ($000)

|

Flag: Results Uneven

|

While the industry reported its best loss ratio result in thestate where insurers wrote the most CMP business, individualresults varied widely in that state–California–in 2004. Zurich, thelargest California CMP writer reported a 48.4 direct loss ratio,nearly 38 points higher than sixth-ranked Chubb.

|

New York chart:

|

Head: Direct CMP Results for State ofNew York ($000)

|

Flag:EmpireState

|

The names of three regional insurers–Greater New York,Harleysville, and Tower–joined national writers among the top 10players in New York state. But neither group had a clear advantagefrom a loss ratio perspective. National writers reported thehighest and lowest loss ratios in 2004.

|

Florida chart:

|

Head: Direct CMP Results for State ofFlorida ($000)

|

Flag: Read 'Em AndWeep

|

Devastating hurricanes that blew through Florida last year madeit necessary to widen the loss ratio column to include three digitsbefore the decimal point for eight of the top 10 insurers. Indeed,the 195.4 direct loss ratio for top-writer Zurich looked good whencompared Allstate's results of 732.0. Chubb whose results (notshown) were better than competitors in nearly every state, manageda 95.2 loss ratio in Florida on $42.4 million of direct written CMPpremiums.