After years of accusing reinsurers of under-reserving, rating agencies appear to have given the industry the "all clear". Ronald Gift Mullins reports.

From the late 1990s into the early 2000s, reinsurers pumped billions of dollars into reserves to cover decades old asbestos, environmental and workers' compensation claims that kept ballooning in settlement costs. Like a broken record, the rating agencies continued to warn that reserves were deficient by billions, and hinted at downgrades if the condition persisted. Even with high interest rates, reinsurers barely eked out a net profit.

What a complete and merry change has arrived within the past year. Although the results are not all in, 2006 has the look of being the most profitable year on record for reinsurers. Combined ratios in the low 90s, even for the largest of reinsurers, are not exceptional - but one reinsurer, Montpelier Re, reported a combined ratio of below 36% for the fourth quarter. With such a cascade of earnings, the drumbeat for increased reserves seems to have been muffled, or at least subdued. Still, rating agencies remain cautious in their assessment of the reserving status for legacy case reserves, IBNRs and loss adjustment expenses.

Kevin Lee, an analyst at Moody's, observed that, "Based on 2005 data, we believe there is a modest reserve cushion in the 2004 and 2005 underwriting years. Reinsurers have already released a portion of reserves from recent underwriting years, particularly in property lines, to offset reserve hikes for older underwriting years." He said the reserving situation will be seen more clearly when the 2006 data rolls in.

The view of Laline Carvalho, director, Standard & Poor's, is that currently the property/casualty business is reaching a level of stabilisation of reserves. "As of 2006 there were few companies that posted reserve developments," she said, "and for those that did, the amounts were not as large as in the late 1990s, early 2000s." She believes reserves will no longer be a major item for the vast majority of insurance and reinsurance companies. "Most have hit favourable reserve development for the year."

One demonstration of the rollercoaster ride of reserving: Moody's estimated that beginning in 1996, carried-reserve adequacy for US non-proportional casualty reinsurance was about 1.5% deficient, but fell rapidly to bottom out at 63% in 2000. Then, year-by-year, the deficiency lessened until at year-end 2005, it had narrowed to less than 2%.

An executive decision

Today reinsurers, like primary insurers, use exotic software programmes to actuarially begin the process of developing reserves. But compared with primary insurers, reinsurers have the distinct disadvantage of being one step removed from the underlying details of the risks and policyholders, yet are still attempting to predict the future. And even with sound actuarial methods to develop ranges for reserves, there is still room for debate and judgment in reserving, "but the selected reserve amounts must be defensible to various constituents, including boards of directors, external auditors, investors, regulators and tax authorities," said Lee.

Certainly, the degree of reserving is open to opinion. Usually there is a range of reserves for a specific risk, or class: a low and high, and the middle. The choice of selecting which part of the range is a matter of informed opinion. The trick, however, is to be as accurate as possible.

A former CEO of a reinsurer, who asked not to be named, commented that reserving is determined by a combination of internal and outside actuaries as well as the CEO and CFO, who also have a say in how much to reserve. "In some cases, depending on a number of factors, such as will the reinsurer show a loss if reserving is too severe, a little shaving will be permitted."

"I am not sure that reinsurance executives are more prone to being aggressive with balance sheet accruals, such as reserves, than other executives" said David Bradford, senior insurance industry analyst for Advisen, "but since the real value of a specific long-tail claim is not exactly known, there is always a little wiggle room. The way it has been done in the past, in good years companies stock a little more away in reserves, in bad years, they're a little more conservative. It's done less now, but it would not be surprising if primary insurers and reinsurers were to take advantage of the big 2006 profits to pad out reserves of long-tail claims."

Today rating agencies, Sarbanes Oxley, the threat of class action shareholder suits, and the reluctance of outside accounting firms to certify a company's financial report if it's the least bit suspect, are encouraging executives of insurers and reinsurers to be more attentive to reserving appropriately. Propelling the drive for transparency was the uncovering of an accounting gimmick involving a finite reinsurance contract between AIG and General Re. This faux contract was intended to beef up AIG's reserves, apparently in an attempt to fool analysts, regulators and shareholders.

Greater transparency

It is a common misunderstanding that reserves are manipulated, but a raft of regulations ensure that doesn't happen, according to Joseph Sieverling, senior vice president, Reinsurance Association of America. "A company has to make an actuarial judgment of a claim, developed from using professional actuarial standards that deal with significant historical uncertainties in long-tail lines. Applying modern citations in recently settled cases can seriously change the amount to be reserved on old liability claims. Tort legislation has begun to reduce the size of awards and there has been attempts to rein in the legal profession. These have had a major positive effect on long-tail reserving, so the deficiency has lessened."

In addition to internal controls, Sieverling says that the increased transparency demanded by Sarbanes Oxley has made company executives more aware of needing sound reasons for their actions. "Following the collapse of Enron and the demise of Arthur Andersen, clearly the auditing profession is highly alert to auditing risk and the responsibility to shareholders. They are very aware of the penalties that could be brought against them."

Mark Jablonowski, senior research analyst at Conning Research & Consulting, admits that if a CEO wanted to improve reserving and the facts are not there to support the action, "he has to steam roll the audit committee, the CFO, internal and external actuaries and other powerful watchdogs. That would have to be some super powerful CEO in these highly restrictive times to manage that maneuver."

"I think that on an individual company basis, under-reserving deliberately did not happen," Sieverling observed. "But if you take the global reinsurance market, there may have been some misrepresentation of reserves. For example from 1997 to 1999, insurers and reinsurers were underwriting all kinds of business at low rates. There was a feeling within the industry that the business was less risky than it was. Under-reserving was going on, not to improve earnings, but because the companies felt the risks were not going to bring huge losses and thus didn't need hefty reserves." As time proved, this brought enormous additions to reserves in later years as the parents of several large US-based reinsurance entities supplied billions of dollars to increase reserves for those risks.

Getting it right

The former CEO of a reinsurer remarked that there is no "completely accurate 1+1 formula for reserving, especially in the long-tail liability lines such as asbestos and environmental risks". Reserving is an art and a science, says Lee of Moody's. "The science comes into play when analysing historical data. Unfortunately, history doesn't always speak to the future - there are data lags and structural changes to the market. That is why the art is needed."

Is reserving an art more than a science? Conning's Jablonowski asks. "There are a lot of judgmental actions that have to be considered," he continued. "Reinsurers have had to react to surprises in the past - A&E claims, an emerging soft market, medical malpractice claims, these surprises caused them to take very large charges for reserving. Now they are attempting to determine by some quasi-scientific measures if the long-term loss ratios are fair. And reserving is an art, in that reinsurers know there are surprises on business that is already written, but it is difficult to quantify it with actuarial maps. Extrapolate or strip out all the A&E losses but you can never get rid of the unknown factors. It is very much a challenge for companies now to try to set reserves."

Most companies have revamped their operations to bring actuaries and underwriters in closer communications earlier, when a risk is presented. This reduces some of the surprises later in reserving for that risk. "You can never get rid of volatility in reserves," S&P's Carvalho said. "But you can diminish volatility by taking a prudent approach. There is no one scientific method of reserving. At end of day it involves a lot of judgmental decisions, a judgment call; the final decision cannot be based on a scientific decision alone. Improved communications between all parties involved in accepting a risk can be a real advantage down the road."

Ronald Gift Mullins is an insurance journalist based in New York City.