Anticipating firming insurance prices, experts say more risk managers are considering or adopting alternative risk strategies that include forming a captive insurer.

Putting a company's workers compensation risks into a captive typically is among the top reasons for moving to a captive arrangement—aside from the tax advantages and other benefits a company-owned insurer may provide. However, forming a captive may not be the right strategy for all employers, and the costs can outweigh the benefits.

Workers comp, general liability and auto liability typically are the top three risks covered by a captive. Workers comp is favored because of the line's premium volume, the long-tail nature of its claims and the relative ease in evaluating the exposure statistically, consultants say.

“Workers comp for a lot of organizations is really the big nut in terms of premium volume,” said Jim Swanke, director of risk consulting for Towers Watson & Co. in Minneapolis. “If you are going to be doing anything in your captive, you are going to be focusing on the largest-size premiums and the premiums that have the longer tails in terms of payouts. Workers compensation by definition falls into that category.”

Sean B. Rider, managing director of sales and consulting for Willis Group Holdings P.L.C.'s global captive practice in New York, agreed.

“For U.S.-based firms, workers compensation is one of first places people go for captive utilization,” Mr. Rider said. “Because workers comp in particular is long-tail business, it has a lot of transaction activity; and with enough size and scale, it can be very credible statistically so you can have a lot of rationality around the use of a captive for workers compensation.”

While the workers comp insurance market remains competitive, it appears to be transitioning from recent years' soft pricing and is driving discussions on stabilizing costs, said Robert Hessel, senior managing director for large risk casualty in the Atlanta office of Beecher Carlson Holdings Inc.

“The consensus is we are in a market that is trying to change. Comp is getting a bit more difficult and so people are perhaps more sensitive to alternatives today than they were a year or two ago,” Mr. Hessel said. “Because the market has been so soft, the tendency has been to lower retentions and purchase more coverage; but to the extent that changes, people will adjust.”

Captives are just one option for employers seeking to retain more risk, said Ellyn Casazza, senior vp in Atlanta for Marsh Inc.'s captives solutions group.

But current market conditions are leading employers to weigh increasing their retention and, in doing so, some are evaluating whether a captive may be their best option for assuming additional risk, she said.

Some employers are more concerned about the firming of insurance pricing for risks other than workers comp, Mr. Rider said. Cost increases for product and professional liability insurance may be a greater concern.

But even employers considering putting their professional and product liability risks in a captive likely will want to put their workers comp risks in the facility because the line's statistical reliability provides a “nice stable base” for a captive, Mr. Rider said.

The decision of whether to launch a captive typically should be based on the size of a company's projected losses rather than its revenue, consultants said.

“We use (projected losses within a year) as a benchmark to determine if a captive is potentially going to add significant tax advantages to outweigh the costs to run the captive,” Ms. Casazza said.

That usually means a company must expect about $3 million in retained loss costs per underwriting year, she said.

The decision of how much risk to retain, however, must be decided separately, before determining whether the parent company will pay the retention or whether it will be financed through a captive, Ms. Casazza said.

Loss history, financial capacity to bear risks and commercial insurance market conditions should help determine the amount of risk to retain.

While $3 million in retained losses is an average benchmark for considering a single-parent captive, losses of as little as $250,000 may be appropriate for group or association captives or 831(b) captives, Mr. Ridder said.

An 831(b) captive, established under Internal Revenue Code Section 831(b), allows insurance companies with less than $1.2 million in annual premiums to elect to be taxed by the federal government only on their investment income.

One benefit of operating a captive is that upper management is likely to support safety and risk management department efforts to reduce losses and enhance the financial benefits of a captive insurer, the consultants said.

Companies also form captives, in part, to help them take greater control over their risks and risk costs, Towers Watson's Mr. Swanke added. “Control is right up there as one of the (top) reasons—and to be less reliant on the commercial insurance marketplace and (insurance) cycles.”

But the real economic benefit of a captive derives from potential tax advantages, Marsh's Ms. Casazza said.

Under a typical insurance program, income tax deductions for claims are derived over time as claims expenses are paid, the consultants said. Because of the long-tail nature of workers comp claims, deductions might be taken over years while reserves sit in the employer's balance sheet.

But a captive allows its owner to accelerate tax deductions. The entire amount of projected claims costs can be taken immediately as a tax deduction when funding is put into the captive for a claim. Meanwhile, the money placed in the captive to cover future claims expenses can be invested.

Yet companies weighing alternatives of taking on more risks need to be careful because captives can increase costs rather than reduce them, Mr. Hessel said.

“You have to consider administrative costs of the captive and, depending on how you structure the captive, your frictional costs can be higher than under a qualified self-insurance program or a deductible program,” he said. “You need to give careful consideration to how you structure the program so you don't increase costs unnecessarily.”

Frictional costs to consider can include state premium taxes and assessments, which can be less under a fully insured program.

“If you have a lot of exposure in states that have a difficult tax or assessment regime, it can make a huge difference,” Mr. Hessel said.

Apart from tax considerations, operational expenses average about $80,000 to $100,000 per year for captive management fees, regulatory fees, audit services and legal advice, Ms. Casazza said.