Alternative markets in a changing insurance environment

ALTERNATIVE markets are not so alternative any more. By many estimates, they account for well over a third of commercial insurance premiums. The raison d'etre of alternative markets is to enable qualified insureds to assume more control of, and responsibility for, their loss exposures. Some agents, brokers and program administrators–particularly those with underwriting expertise and a good book of business–also get involved in the alternative markets as a way to earn additional revenue, although the opportunity is not risk-free.

Alternative markets take on many forms, including captives, rent-a-captives, self-insurance plans and pools, risk retention groups and purchasing groups. For this special report, we've contacted seven people who are involved with these mechanisms to obtain their perspectives on how they are being affected by everything from the softening market to New York Attorney General Eliot Spitzer, as well as what the future holds for them. Before reading their comments, however, it might be helpful to briefly review the characteristics of single-parent captives, group captives and rent-a-captives, the main alternative-market mechanisms discussed in this report.

An overview of captives

Agents and brokers sometimes help large accounts create single-parent captives, which the account uses to cover its own risks, while using reinsurance to cap its losses. Agents and brokers also are involved in various kinds of group captives. One is an agency-owned captive, in which the agency puts up the formation capital, accepts a certain amount of risk and owns any underwriting profits and investment income.

There also are member-owned captives, which agents and brokers may sponsor without taking on any risk or sharing in any profits. Instead, they are compensated by commission and by fees for any services they may provide. Meanwhile, the captive's risks and rewards are shared by a (usually homogenous) group of midsize clients, who also put up the capital to form the group captive. The required one-time stock purchase generally ranges from $30,000 to $50,000. The captive members also pay annual premiums, as determined by individual underwriting.

In a typical group captive, a captive facility is retained to design and manage the captive. It arranges for a fronting insurer, generally an admitted or nonadmitted U.S. domiciled insurer, to issue policies to the captive's members. Premiums, less the fronting company's fees and various expenses (commissions, residual-market loads, taxes, “boards and bureaus,” etc.) are then ceded to the group captive, which functions as the fronting company's reinsurer and may be domiciled offshore or, increasingly, in states like Vermont and Hawaii. The fronting company usually obtains reinsurance above the group-captive layer, up to the limits of all the policies the fronting insurer issues.

The protection layers of a group captive can be structured in various ways. One common approach is for each member to have a deductible or self-insured retention for each line of coverage provided by the group captive. Above this initial retention would be a “dividend” pool, with a size calculated to cover all members' expected losses. Those members whose losses turn out to be lower than the projected “loss pick” benefit accordingly. Above the dividend pool is a risk-sharing pool, designed to provide a limited amount of catastrophe coverage. Most of the risk-taking by captive members takes place in this layer. Reinsurance then protects captive members from further losses.

Rent-a-captives are specialized captives, usually owned by one entity, that “rents” individual “cells” of the captive to agents, brokers or program administrators, or directly to insureds. One advantage of rent-a-captives is that they do not require users to contribute formation capital, although users pay a fee to the captive manager that typically runs between 1% and 2% of the gross written premiums placed in the facility. Rent-a-captives also are much easier and quicker to get into (30 days or less, compared with months for an owned captive).

A rent-a-captive uses a fronting insurer, in the manner previously described. It cedes the net written premium, less expenses, to the rent-a- captive cell, which acts as the fronting carrier's reinsurer. The amount ceded, typically 50% to 60% of the gross premiums, becomes the cell captive's loss fund. Aggregate excess insurance is purchased to protect the cell captive from high claims frequency, but its attachment point is set well above the amount of money in the loss fund. The difference is the “gap,” which is the layer of insurance for which the agent, broker, program administrator or insureds will be responsible.

Under the rules of statutory accounting, to which U.S. insurers must adhere, the rent-a-captive is considered a nonadmitted, unapproved reinsurer. (The same holds true for group captives.) Thus, while the insurer will show a liability on its balance sheet for the potential losses represented by the gap, it cannot show the rent-a-captive's contractual obligation to cover those losses as a corresponding asset. Therefore, the rent-a-captive user will have to post collateral for the gap, typically in the form of a bank letter of credit. This fact has a number of implications, particularly for agents, brokers and program administrators assuming risk via a captive, as discussed in some of the commentaries that follow.

John Shea
Tangram Program Managers and Insurance Services

John Shea, president of Tangram Programs and Insurance Services, in Petaluma, Calif., has a dual perspective on the alternative marketplace. The bulk of Tangram's revenue comes from eight programs it operates as a typical program administrator. Four of the programs use rent-a-captives, making him a user of alternative-market facilities.

Tangram also provides services to agents, brokers and program administrators interested in setting up their own alternative-market facilities. Those services include consulting (on captive feasibility and formation), arranging fronting services and procuring reinsurance. Tangram can pro-vide claims and loss-control services through Advance Risk Technologies, a TPA that Shea also serves as president. Tangram assists in the formation of group captives or can refer agents and brokers to rent-a-captives that it has worked with.

Shea said he has been involved with captives since the early 1980s, when few “Main Street” brokers had any understanding of, or interest in, captives. Today, he says, “Almost any broker has at least a working knowledge of how captives and rent-a-captives work.”

Interest is particularly high among larger producers, he said. “I would venture to say that half of the large, regional brokers, and certainly all of the alphabet houses, have somebody on staff who is an alternative-risk expert.”

Why do brokers form group captives or cells in rent-a-captives? The chance to share in underwriting and investment profit certainly is a motivator, Shea said, but the desire for control is even more of a driver. By “control,” Shea said he meant not just the ability to determine such factors as pricing and risk selection but also to control access to a proprietary market. “You wind up with an exclusive market that other brokers are not going to access,” Shea said, “and that gives the broker an edge in the market.”

Alternative-market programs tend to be formed for risks that are underserved or not well understood by standard insurers, Shea said. By creating a captive, a broker often can deliver a better or less expensive product than can be found in the open market, he said. Of course, a key attraction for insureds is the prospect of enjoying long-term stability and predictability in their insurance costs, rather than seeing those expenses fluctuate with the underwriting cycle.

The first step in assessing the feasibility of a proposed alternative-market program is determining whether it has “critical mass,” Shea said. It should have $3 million to $5 million of premium and, preferably, a homogeneous book of business. “You can put together captives for a heterogeneous book, but it's much more difficult,” he said. The program also must have an acceptable projected underwriting profit–at least 5% to 10%, Shea said.

The information checklist for captive formation is fairly straightforward, Shea said. “You need five years of premium and exposure history, an overview of the exposures and a good marketing plan,” he said. “That's really about it; it's not rocket science.”

Of course, the agent or broker also must be prepared to invest in the program. In the formation of a group captive, the broker and each insured typically share the initial capitalization costs. But the real investment in either a group captive or rent-a-captive is collateralizing the “gap,” Shea said. If you were to have a $10 million (gross premium) program, he said, anywhere from 25% to 40% will be needed to cover expenses, including commissions program administrators pay to retail agents. The remaining 60% (assuming expenses are 40%) is the captive's loss fund. Reinsurance kicks in if the fund is depleted, but not until the broker (and sometimes captive members) has absorbed some losses–the gap.

Shea said the amount of capital required to fund the gap typically is anywhere from 20% to 50% of the captive's gross premiums–for example, $2 million to $5 million for a $10 million program. Brokers sometimes do not realize that the gap is not a one-time investment, but has to be collateralized each year (although as losses are developed for each year, the collateral posted for it can be drawn down.)

Contrary to conventional wisdom, Shea said more captives and rent-a-captive programs are formed in a soft market than a hard one. A big reason is that reinsurance usually becomes more plentiful. That's certainly the case now, he said. “I'm suddenly getting calls from my friends in the reinsurance market, looking to help me put together captive programs.”

Insurers also are more interested in offering fronting services, Shea said. Standard insurers with idle capacity–even those that previously had not been in the alternative-risk business–see fronting as a means for expanding their market share, he said.

All this does present a challenge to agents and brokers marketing an alternative-risk program, Shea acknowledged. “When the client wants it the most, there is such a huge demand that it's almost impossible to put together a captive program that makes financial sense,” he said. “When the market is soft, on the other hand, there isn't that much demand from the standpoint of lowering premiums.”

Tom Kozal
Arch Insurance Group

Arch Insurance Group's alternative markets division works with a variety of captives, all of which are initiated by agents and brokers, according to Tom Kozal, senior vice president. It has established single-parent and group captives in which the agents or brokers have no ownership, as well as group captives in which they do. Arch, however, does not work with a captive that is entirely owned by an agent, broker or MGA. “We always want the ultimate insured to be involved,” Kozal said, adding that in the wake of investigations by Eliot Spitzer and others, Arch would require that insureds acknowledge in writing that they are aware that the agent is involved in the captive ownership.

For group captives, Arch does not “give out the pen,” Kozal said, which in essence means it likely would not form captives for program administrators, although Arch does have a division that works with program administrators on deals not involving captives.

Kozal said his typical captive client is a midsize regional broker interested in putting a number of long-time, preferred insureds into an alternative-market arrangement. You might have nine insureds and the broker, Kozal said, who each would own 10% of the captive's stock. “We have several opportunities like that on our desk right now,” he said. In such arrangements, brokers can provide a group of superior clients with lower-cost coverage while the brokers also share in any underwriting profit and interest income.

Kozal said some of his business comes to him via captive “wholesalers.” These are organizations that assist agents and brokers who are interested in creating alternative-market programs but don't have the necessary expertise. They include independent operations like Innovative Risk Management and Garnett Captive Services, and subsidiaries of larger entities like Gallagher Captives Services (owned by Arthur J. Gallagher) and Innovative Captive Services (owned by Holmes Murphy & Associates).

Whether approached by a retailer or a captive wholesaler, Arch typically functions as the issuing insurer and also provides aggregate stop-loss insurance. For those programs that desire to use a rent-a-captive, Arch owns two facilities, based in Bermuda. Kozal added, however, that the rent-a-captives typically are used by group-owned captives, rather than those in which the agent or broker also is an owner.

Arch expects a captive to have at least $3 million in premium the first year, and a good chance of reaching $10 million within three years. “We also are looking for the individual accounts to be fairly substantial–ideally, at least $300,000 to $400,000 in premium,” he said. Typically, the captive insures the three main liability lines for its members: workers comp, auto liability and general liability.

Kozal said he has yet to see the softening market significantly affect the alternative marketplace. While conditions certainly are softening “in the more commodity driven lines … the primary casualty market is still fairly tight,” he said, especially for difficult classes like nursing homes, contractors and certain manufacturers.

While agents and brokers primarily come to Arch to form captives for their top-tier accounts, their motivation sometimes is to find a way to cover difficult-to-place risks, Kozal said. He said it's fairly easy to tell when agents and brokers are simply looking for a place to dump a bunch of poor accounts. “Those (deals) aren't getting done,” he said.

Kozal said Arch will take on difficult risks–as long as the captive is selective. As an example, he cited a captive for California house-framing contractors. The captive is run by a wholesaler, and multiple retail agents access it through the wholesaler. The group has a tight underwriting committee, which, Kozal said, is its key to success. “In fact, that group's underwriting is tighter than ours,” he said. Selectivity is essential, Kozal said. “When underwriting difficult classes, you look to write the best” in those classes.

Rich Turner
Liberty Mutual Insurance Co.

Liberty Mutual prefers to work with homogenous groups of clients who are large enough to take on risk, according to Rich Turner, Liberty's managing director of sales for alternative markets.

Liberty Mutual usually does not work with program administrators, Turner said. Rather, all captive opportunities are brought to the carrier by agents and brokers. Some coordinate with the carrier in the creation of a group captive or a cell in a rent-a-captive but usually do not join their clients in taking risk. They may, however, provide the captive with loss control, claims or management services on a fee basis. They are not involved in underwriting. “We don't give out the pen,” he said.

Large agents and brokers may bring the carrier existing captives for which they already provide required services. Small ones, on the other hand, may have one or more clients interested in forming captives, Turner said, but not the resources or expertise to create them. In that case, the insurer prepares feasibility studies and works with the agents and brokers to set up and manage captives in appropriate domiciles, he said.

The carrier does not work with captives that are owned 100% by agents and brokers, and in which they retain any underwriting profits. “I think agency-owned captives in general have taken a big turn backward,” he said, citing the Spitzer investigation. “You also have a number of brokers that by their charters are no longer allowed to participate” in such captives, he added. He said the relatively few insurers still working with agency-owned captives require full disclosure to captive participants.

Turner said he saw a sharp increase in interest in captives during the hard market. Even as the market softens, he said captives remain attractive for difficult-to-place risks. For instance, Liberty Mutual has created captives to insure homebuilders' completed operations exposures. “The client is really looking to take risk, because the cost of buying the coverage is almost as large as the policy limits,” he said.

Turner said valid reasons for a business to get into a captive include a desire to control all elements of a risk management program and a belief that the traditional marketplace is way overpriced for the participant's liability exposure. A bad reason, on the other hand, is a desire to reduce premiums when one already has a competitive quote from the traditional marketplace. “What you really are looking for is a client who says, 'I have done something that obviously has changed the risk factors of my operation, which is not being recognized by the traditional marketplace,'” he said.

Property exposures aren't as well-suited to captives as liability exposures, Turner said, since losses tend to be paid immediately, meaning there is less time to earn investment income from the loss fund. Captives also are ill-suited for catastrophic property exposures, he added. “Captives tend to be most attractive to predictable risks with high frequency and low severity.”

From an agent's or broker's point of view, one of the attractions of a captive is that a client who goes into one is no longer interested in shopping its account every year, Turner said. The ability to offer clients captive services also differentiates agents and brokers from competitors that can offer only traditional transfer-of-risk products, he said. Captives are “hands-on” products for insureds, Turner noted, which means the agent or broker will be more involved with a client's senior management, fostering a deeper relationship. The agent also can offer a number of fee-based services to lower losses–and the clients will be interested, he said, “because it's their captive.”

As the market softens and traditional insurers sometimes cut rates below expected loss costs, it becomes harder to set up captives for the most desirable risks, Turner said. Nor will captive participants necessarily benefit from from smaller “gaps” in such a market, he said. In past soft markets, insurers reinsurers narrowed the gaps by dropping their attachment points for stop-loss coverage, Turner said–and sometimes got burned in the process. “I'm not sure what's going to happen in this marketplace,” Turner said. “My guess is that there is going to be more discipline.”

Greg Lang
Munich-American Risk Partners

While the soft market may be slowing the growth of alternative-market programs, there continues to be strong interest in them, according to Greg Lang, Munich-American Risk Partner's senior vice president for business development and marketing. “We continue to see many new agents coming in and taking risk for the first time,” he said.

Munich-American works primarily with program administrators or other wholesalers on the formation of captives or rent-a-captive cells. Thanks to program administrators' underwriting expertise in their niches, these books of business typically have better-than-average loss ratios, Lang said. “Our theory is, Why not share in some of the fruits of your labor (by partaking in favorable underwriting results), rather than just earn commission?” Munich-American normally looks for homogenous books with $10 million or more in premium volume, although Lang said there are exceptions.

Munich-American can issue policies for a captive, work with a program administator's existing carrier or ar-range for the services of a fronting carrier. It can provide aggregate excess insurance for a program administrator's captive as well as offer the services of two Bermuda-based rent-a- captives. The program may cover a single line of coverage, but Munich-American also can arrange alternative-market programs for multiple lines, Lang said. On any given program, the company may or may not arrange for claims service, captive management services, or even a fronting carrier, he said. “We're very comfortable with the unbundling of services.”

In evaluating captive opportunities, Munich-American looks closely at program administrators' financials to gauge their ability to take on risk. It also looks at their programs' track records to ensure they have been maintaining tight underwriting standards. “If the program has grown more than 200% during three years of a softening market, that could be a red flag,” he said.

Munich-American expects program administrators to accept a significant amount
of risk in their captives and rent-a-captives. For a $10 million program, the “gap” for a program administrator might work out to $1 million above expected losses. In the softening market, however, Lang said, the size of gaps may be coming down a bit as reinsurers become “a little more aggressive with their aggregate attachment points.”

Like Tangram's John Shea, Lang noted that program administrators have to collateralize the gap and sometimes don't initially understand they have to do so every year, as new business is written, a phenomenon Lang referred to as the “stacking” of collateral. He said they also do not always fully grasp that in casualty lines, where losses can take several years to fully develop, the unused collateral for a given program year might not be returned to program administrators as quickly as they might expect. As premiums shot up in the hard market, so did collateral requirements. Now that premiums are leveling off or even falling, one consolation is that collateral requirements are dropping too, Lang said.

As did some other people contacted for this report, Lang said the Spitzer investigation into broker practices has discouraged the formation of group captives owned 100% by retail agents. Without adequate disclosure, such arrangements can have an inherent conflict of interest, in that agents could put their interest in earning underwriting profits ahead of their responsibility to get their clients the best deal. Captives owned 100% by program administrators or other wholesalers present no such conflict, he said, since unaffiliated retail agents and brokers bring them their business.

Roger Greiner
Markel/SMART

Specialized Markel Alterative Risk Transfer was founded in July 2003 as the platform for the Markel Insurance Group to produce alternative risk business, according to Roger Greiner, president. It targets large individual commercial risks, public entities and homogeneous casualty insurance programs, he said.

Greiner said most of the unit's clients have been program administrators, although it also works with retail agents and brokers, reinsurance intermediaries, third-party administrators and insurance consultants. For program administrators, SMART has helped create group captives and rent-a-captives for habitational risks, restaurants and taverns and assisted living facilities, among other classes, he said.
Under the usual scenario, an admitted or nonadmitted Markel company issues policies for the program, Greiner said. Actuaries then determine a program's “working layer,” where most of the losses occur, and cede part of it to the captive on a quota-share basis, he said.

Suppose, for instance, most losses amounted to $250,000 or less for a captive or rent-a-captive providing $1 million of general liability coverage to its members. The issuing company might cede 25% of the $250,000 working layer to the captive, he said, while retaining the rest of the exposure. If the total expected losses within the working layer were $2 million, then the captive would be responsible for 25% of them, or $500,000. That would be the captive's “gap,” Greiner said, which the program administrator would be expected to collateralize. Then a Markel company might retain 100% of the losses within the remaining $750,000 layer of the $1 million policies, he said.
Besides sharing in the underwriting profits (or losses) of their captives, program administrators are paid commissions based on whatever functions they may perform for the insurer, Greiner said, which could include rating, quoting, and binding business, and billing and collecting premiums.

SMART also helps program administrators who operate purchasing groups, which are entities authorized by federal legislation that enable multiple, homogenous risks to buy certain kinds of liability insurance from traditional insurers on a group basis. SMART provides insurance for the purchasing groups from one of the Markel companies, Greiner said, which issues a single policy to the group. The program administrator or wholesaler then issues certificates to the group's members, affirming their status as insureds.

Greiner said SMART works with retail agents and brokers primarily on group captives or single-parent captives arranged for their clients. The captives generally are created to coverage the general liability, auto liability, and occasionally, specialty or professional liability exposures of middle- market accounts. Workers comp exposures are not accepted. Single-parent captives are used for accounts that pay at least $3 million to $5 million in premiums, Greiner said. Retail agents and brokers generally do not share risks in these captives, he added, but are paid commissions for business placed in these captives, just as they would be for business placed with any insurer.

Excess public entity business is another target market. “As long as there is significant risk-taking on the part of the account itself, SMART will provide the excess capacity beyond their SIR,” Greiner said. Recently it began working with public entities that share their risks in legislatively created pools. “We write a reinsurance treaty behind the pool to support an excess layer that the pool might need to provide capacity to its members,” he said.

Despite a softening market, Greiner expects a steady gain in business. “We're looking for controlled, profitable growth, where we can build relationships over a long period of time,” he said.

Daniel T. Keough
Innovative Captive Strategies Inc.

Innovative Captive Strategies (ICS) was created in 1999 as an affiliate of Holmes Murphy & Associates, a large privately held regional insurance broker in Des Moines, Iowa. Dan Keough, the organization's president, said ICS initially was formed to give Holmes Murphy the ability to offer captive services to its own clients. It has since expanded its scope to provide such services to other agents and brokers. ICS typically works with retail agents and brokers having $10 million to $100 million in revenue, Keough said, but to date has not worked with MGAs or program administrators.

While ICS has at times worked with 100% agency-owned captives it more frequently works with member-owned group captives, Keough said. They tend to be created for homogenous middle-market businesses that use them mainly to insure common liability exposures, including auto liability, workers compensation and general liability. ICS also helps form single-parent captives for large businesses, he said, those paying $2 million or more for insurance. Such a large business also might use a rent-a-captive to separately insure unusual or difficult-to-place risks, he added.

ICS functions as a consultant and facilitator of alternative-market programs, generally taking no risk itself, Keough said. An agency might approach ICS, for example, for help in creating a member-owned group captive for contractors, Keough said. ICS would prepare the necessary feasibility studies, present proposals to potential fronting companies and excess insurers, and arrange claims management and other services if necessary.

Most agencies come to ICS by referral, Keough said. These agencies have close working relationships with clients they place with traditional insurers, he said, and they also want to be involved in their clients' alternative-market programs. An agency working with ICS typically sits on a captive's board and on its committees, and sometimes shares its risk, he said.

Keough said agents' interest in captives and similar arrangements increased during the hard market, as insureds sought alternatives to the higher premiums and reduced coverage offered by traditional insurers. “Customers wanted more control,” he said. That is still a motivating factor, he said; another is Eliot Spitzer. The New York attorney general's attack on contingency commissions has led some agents and brokers to wonder whether they eventually might disappear under pressure from regulators or middle-market clients. As a hedge against that possibility, some agents and brokers are looking for revenue options that create value for clients and in which all charges are transparent. “Captives met both those criteria,” he said.

Christopher Payne
Aon Alternative Risk Underwriting

According to Christopher Payne, now is a good time for agents and brokers to explore alternative-market options with their clients. Payne is president of Aon Alternative Risk Under- writing, which is part of Aon Underwriting Managers. ARU facilitates the creation and operation of agency-sponsored, insured-owned group captives. A typical group captive might provide workers compensation, general liability and business auto insurance for 10 to 50 midsize clients, paying an average of $700,000 in annual premiums. Aon Alternative Risk usually arranges these captives for sponsoring midsize and larger brokers, and occasionally for smaller agents.

Payne said the number of available fronting markets has expanded a bit, as insurers that had pulled back during the hard market now entertain new opportunities. Reinsurers also are more interested in group captives, he said. Some formerly might not have taken the time to grasp the difference between them and program business, “in which an MGU might try to write a few thousand doctor's offices,” he said, but that's no longer the case.

Pricing for reinsurance and other components of alternative-market programs also are softening a bit, along with the market in general, he said. Because of Hurricane Katrina, group captives insuring property exposures aren't going to see a reduction in reinsurance costs any time soon, Payne said, but on the other hand, “I think there's a lot of capital out there, so I don't see this turning the market.”

Payne said that when markets soften, some new players will become “reckless” with pricing, “but I think the more intelligent buyers will see past that and say, 'Look, this is still a lot more attractive than what I've paid historically.'” Also, he said, such accounts will continue to be drawn to the concept of controlling and investing in their own insurance programs, rather than “writing a check to somebody.”

All in all, Payne said, today's market offers agents and brokers a low-pressure environment in which to talk about group captives to clients who are interested in opting out of a cyclical insurance industry and large enough to finance their expected losses.