Cedents are becoming increasingly concerned about the security of their reinsurers, particularly in light of the global financial catastrophe. Diversification, a time-honored approach to managing risk, may offer part of the solution. Cedents may benefit from diversifying reinsurance placements among many reinsurers. Thus, the approach applied to asset management can be applied to reinsurance placements, as well. The syndication process carried out within the broker market results in an important reduction of the “no recovery” potential that could arise from reinsurer defaults. Diversification may reduce the probability of no recovery, even if the likelihood of default among reinsurers is correlated.
Diversification is a widely accepted principle in investing and asset management. We don’t need to know much about diversification to understand that if all of our invested assets are held in the common stock of a single company, a failure or severe set back for that company could be catastrophic. It is also intuitively clear that if we add stocks of high quality companies to a portfolio, we will reduce the risk of calamitous loss. Expanding the concept further, the capital asset pricing model (CAPM) is used to demonstrate more formally that diversification can achieve a benefit in the risk/reward profile of a portfolio by effectively reducing the volatility of a portfolio’s expected returns, obtaining both the benefit of averaging the expected returns of different stocks and reducing the volatility of returns. An event or development that has an adverse impact on one company may be offset by a favorable (or at least less detrimental) development at another company.
This thinking can be applied to reinsurance placements-not just investment assets held in a portfolio. A diversified reinsurance program can lead to a dramatic decrease in the probability of a “no recovery” event and the deleterious impact on a ceding company that can result from it. The principle of diversification suggests that, in spreading its reinsurance program among a variety of reinsurers, a ceding company will achieve the very important effect of diminishing the chances of a grievous financial loss resulting from reinsurance company failures.
There are several common considerations that would lead a ceding company to diversify its reinsurance placement. Often, reinsurance placements are so large that they simply require the participation of several reinsurers to provide sufficient capacity. Also, relationship management can come into play, particularly as cedents seek to build loyalty across market cycles and assure the availability of adequate reinsurance capacity when market conditions favor reinsurers. Further, a ceding company may wish to allow competition to help assure that it is receiving adequate attention, services, and creativity from its reinsurers, and thus will diversify by entertaining bids from new market entrants.
The consolidation of primary insurance companies has led to greater concentration of risk-both individual and aggregate. Years ago, it may have been unlikely that default on a single risk or treaty could harm profits severely or shake the financial foundations of an individual insurer. Conditions have changed, however, and the collapse of a single risk or treaty could impair a line of business or even an entire company. A prudent ceding company recognizes that-even if a single reinsurer has the financial capability and risk appetite to absorb all or most of a risk or treaty-counterparty exposure considerations may call for multiple participants.
Finally, as will be discussed in the following week, diversification may reduce the default risk of reinsurers.
Case Study Assumptions
Over the next week, we will publish several case studies to illustrate the implications of reinsurer diversification. The following assumptions will be used to quantify diversification impacts:
First, we will assume that a ceding company wishes to reinsure for USD200 million in casualty cover. We further assume the business is long-tailed and therefore that the ceding company will be exposed to the credit risk of the reinsurer at least 10 years. In fact, we will assume that the reinsurance transaction results in USD200 million of recoverables that will be outstanding in 10 years.
Next, for convenience, we will concern ourselves only with default or no recovery risk, whereby a default means that the ceding company will not recover from the reinsurer. In reality, of course, recovery isn’t all or nothing. Partial recovery or even delayed recovery represents a real loss to the ceding company and is only a variation on the no recovery case.
We need to establish default probabilities for high quality reinsurers. We are unaware of any specific data source that provides default information for reinsurers, so we will use data from a Moody’s report on default experience for corporate bonds as a proxy. Specifically, we will draw from the cumulative 10-year default rates for bonds with the three highest whole letter grades: Aaa, Aa, and A. (Moody’s attaches numerical modifiers to each generic whole letter rating classification from Aa downward, so that an issuer could be Aa1, Aa2, or Aa3. The default data we borrow from ignores the modifiers.)
The default probabilities that we are trying to capture are cumulative defaults over the first 10 years following a rating. These rates are approximately 1 percent for Aaa, 2.8 percent for Aa, and 3.4 percent for A (Exhibit 43 of cited Moody’s report). Consequently, we assume the Aaa rated reinsurers will have a 1 percent probability of default over a 10-year period, with Aa rated reinsurers at 2.8 percent, and A rated reinsurers at 3.4 percent.
These default rates are not purported to represent reinsurer experience-and that even cumulative corporate bond default rates will be different depending on the period from which they are derived. For example, the rates cited above are from a table capturing experience from 1920 to 2002, and the defaults rates are considerably different if derived from data from the 1970 to 2002 period (perhaps because of improvements in rating methodology or some environmental factor). However, the rates are useful in that they may represent both direction and magnitude of default probability over the three grades of credit worthiness.
Finally, we assume that the default of any one reinsurer is independent of, or not correlated with, the default of another. This is a convenient assumption for illustrating the points to come. We relax this assumption later.
- Default & Recovery Rates of Corporate Bond Issuers, Report # 77471, Moody’s Investors Service, February 2003
- Sean Mooney, Chief Economist