Finite reinsurance being replaced by structured reinsurance following scandalsPosted On: Sep. 4, 2011 12:00 AM CST
Finite reinsurance deals, tarred by scandal and regulatory suspicion, have been replaced by “structured” transactions that put to rest any doubt whether there is a genuine transfer of risk while providing a multiyear alternative to traditional reinsurance.
While finite reinsurance managed loss volatility after it took off in the late 1980s, concerns grew that it was more of a loan, with critics alleging it was being used to hide the true condition of insurers.
A Financial Accounting Standards Board ruling in 1993 that there must be a reasonable chance of a significant loss for a transaction to be accounted for as insurance led the market to develop the “10/10” rule, whereby a transaction was considered insurance if there was at least a 10% chance of a 10% loss.
The image of finite reinsurance was stained in 2008 when investigations by then-New York Attorney General Eliot Spitzer and others into finite deals led to the convictions of several former executives at Stamford, Conn.-based General Reinsurance Corp. and New York-based American International Group Inc.
“Finite came under a lot of pressure in the Spitzer years,” said Mike Schnur, Chicago-based partner with TigerRisk Partners L.L.C.
Last month, however, a three-judge panel of the 2nd U.S. Circuit Court of Appeals in New York overturned their convictions in the 2008 case and ordered a new trial, ruling in part that the jury was instructed improperly.
Even if the defendants' names are cleared eventually, the damage to the reputation of finite reinsurance is done, with various forms of structured reinsurance now taking its place.
Even so, there is no firm definition as to what constitutes a structured reinsurance transaction, sources noted.
“My definition is: products where there is a degree of risk-sharing between the parties,” said Ed Hochberg, Philadelphia-based executive vp and global product leader of risk transfer products and analytics at Towers Watson & Co. “That is a good way of looking at it, but there are many variations on the theme from there.”
“What is meant by structured is structured over time,” said Donald A. Paterson, CEO of Paterson2, a Westlake Village, Calif., firm that develops alternative risk transfer programs. The transactions typically are for longer than a year and involve some sharing of risk among the parties, he said.
Depending on loss activity, return premiums may be paid by reinsurers to insurers under structured reinsurance deals or, if losses are heavy, the insurer could be forced to make additional payments to the reinsurer, Mr. Paterson said.
“I would call it alternative reinsurance,” he said of the deals being put together today. “It's reinsurance that's not traditional.”
Whatever the name, the rationale for today's deals—and the clear disclosure of risk transfer—differs from that used in developing finite reinsurance products, Mr. Hochberg said.
“What we used to see 10 or 15 years ago,” he said, were deals “with an awfully significant element of discounting loss reserves. A lot of that has gone away. That element of the deal is much less significant than it was.” The aim of some finite transactions was to “achieve an accounting result, whether it was the movement of a result from one year to another, or whatever the objective was.”
Today's deals by comparison are more “capital-oriented,” Mr. Hochberg said. “They are designed to truly transfer risk in a way that is comfortable for all parties...When we do a structured deal these days, there's not a question of, "Where's the risk transfer?' The risk transfer is very obvious.”
Finite reinsurance truly began to vanish when a soft insurance market began hardening in 2001, said Dan Malloy, executive managing director at Aon Corp.'s Aon Benfield reinsurance intermediary unit in New York.
“People were underreserved” after the soft market years, Mr. Malloy said, “and deals were done to create better results. People opted to buy less expensive finite deals, and some of those ended up going very badly for reinsurers. As the market hardened in 2002 and 2003, buying shifted to deals with more risk transfer with higher margins because the market could support it.”
Structured reinsurance arrangements are useful for workers compensation, property/casualty and some property catastrophe exposures “where you aren't transferring the 100-year risk,” Mr. Malloy said. “It might be at the bottom of your property catastrophe program.”
“Typically, you want good actuarial or modeled data” on risks that have a fairly comfortable range of potential outcomes, he said.
“The motivation of the client is typically something other than the transfer of risk for a price” in a structured reinsurance arrangement, said Mr. Malloy. “They want to retain the economics of the business in some fashion” and the risk- and reward-sharing features allow them to do that, he said.
The multiyear aspect of structured reinsurance products also can make them useful for risks that are not as easily modeled, Mr. Paterson said.
The Y2K risk that computers might crash at the turn of the century is an example of an exposure with consequences that could develop over time that would be suitable for a multiyear risk-sharing arrangement in a structured reinsurance program, Mr. Paterson said.
Sources said structured reinsurance is well-suited as an aggregate stop-loss product. “You take all the risk of a company and put it in one basket,” Mr. Paterson said.
Not all markets will write such whole-account coverage, though, because it tends to cannibalize their traditional business, Mr. Paterson said. “There are not many markets doing that.”
Because opinions vary on what constitutes a structured reinsurance deal, experts say it is hard to know how much of the coverage is written.
“One man's structured reinsurance deal might be another man's traditional deal with some bells and whistles,” said Mr. Malloy.
Before it fell out of favor, finite reinsurance was around 5% to 6% of the reinsurance market, experts said. Structured reinsurance is “certainly no more than that,” said Mr. Schnur, and likely far less, other sources said.