While capital adequacy proved fairly robust following another active hurricane season in 2005, greater consistency across capital models is needed, argues Vasilis Katsipis

The great majority of companies affected by the hurricanes and other natural catastrophes in 2005 seem to have been able to withstand the financial impact of these events, although there have been a small number of downgrades.

When it is considered that each of the seven major hurricanes of the last two years ranked among the 20 costliest US catastrophes ever faced by insurers, one has to admit that the industry has been very robust in absorbing the impact of these catastrophes. There are several reasons contributing to this but two factors emerged as the main drivers.

First, the timing of these events was particularly clement for the industry, arriving at a time when rates were still relatively high and investment yields had recovered from the lows of the early 2000s. Therefore, the combined effect for many companies was that the financial impact of Katrina, Rita and Wilma, significant though it was, still remained an earnings event. That is, it did not really encroach into insurers' and reinsurers' capital and surplus.

Second, the capitalisation of the industry as a whole had been recovering over the last three to four years, not least because of improved technical results. This meant that even those companies for which these catastrophes resulted in a capital event were able to attract additional capital in a very short time. The level and speed with which this additional capital was injected into the industry has thus been critical in the case of these companies.

Therefore, the great majority of insurers and reinsurers have either not had their capitalisation impacted or indeed were in a position to very speedily raise additional capital to cover the economic impact of the catastrophes.

MODEL WEAKNESSES

While the 2005 catastrophes did not reduce the capital available in the industry, they did expose shortcomings in the abilities of insurers to measure their exposures. To start with, there is the higher frequency of high-claims hurricanes that has been witnessed during the last two years. As a result of this kind of activity, especially the 2005 season, experts are starting to reassess previously held beliefs as it becomes evident that the intensity of storms around the world is increasing. Clearly, this is a major headache for the industry when one considers that exposures and probable maximum losses (PMLs) are estimated based on models that utilise historical hurricane activity most commonly built up over the last 100 years. In many cases, the way the models were used exacerbated the problem, such as in cases where the models were used with the storm surge and, more importantly, the demand surge options switched off.

These were the main reasons why the industry as a whole has had to re-examine both its exposures assumed and re-evaluate what constitutes the correct technical price. This in turn has directly impacted the estimation of the economic capital available to each company as it underestimated the net PML (which is typically excluded from a company's capital and surplus) and at the same time resulted in an underestimation of the exposures assumed by certain insurers.

Therefore, while catastrophe models are valuable tools in monitoring the normal distribution of catastrophe losses, there is a need for careful monitoring of zonal aggregates in order to understand the maximum potential loss of any given company.

REVISING CAPITAL MEASURES

As a result of potential shortcomings in the catastrophe models, while insurers are trying to refine their estimations of their exposures, rating agencies are trying to evaluate the further potential impact of similar catastrophes on those insurers who are rated by them. The two main lessons of this catastrophe season are that the industry as a whole needs to ensure greater consistency of the models used and that it needs to exhibit a greater discipline in estimating its exposures.

AM Best now specifically requests that options for demand surge, storm surge, fire following earthquakes and secondary uncertainty are included in the loss estimates provided by companies. Additionally, material sources of catastrophe risk, such as property structure and contents, additional living expenses, business interruption and so on are also required to be included in the insurer's loss estimate.

As part of its analysis, AM Best has been requiring a company to ensure that its capitalisation is able to withstand the financial impact of a severe event (typically defined as a 1-in-100 year windstorm or 1-in-250 year earthquake). With the frequency and severity of catastrophic losses escalating, it has now become imperative to analyse a company's ability to withstand a further such event within the same financial year. This is clearly going to impact a company's available economic capital, reinsurance recoverables and counterparty risk, as well as its loss reserves. When estimating the impact of the additional catastrophe, its net after-tax effect (that is, net of reinsurance) is subtracted from the available capital of the company. At the same time, reinsurance recoverables are increased by the difference between gross and pre-tax net PML while the reinsurance credit factors are increased to reflect a potential downgrade of one notch impacting the whole of the reinsurance industry. Finally, loss reserves are increased for pre-tax net PML and are added to the existing reserves.

The stress-tested capital of the company may be up to a maximum of one financial strength rating level below its stand-alone rating. This applies to all companies worldwide and, as mentioned earlier, is a reflection of the increased severity and frequency of catastrophic losses. The only difference that applies among companies is the fact that the impact of the second loss is "tempered" depending on the event that leads to the catastrophe. In cases where the main exposure is earthquake then the second event is a 1-in-100 year event, while in the case of losses arising from terrorism the test is applied either to the modelled loss output or to 75% of the largest aggregate exposure. In addition to stress testing, the capitalisation of a company is also analysed in terms of the potential impact of additional debt issuance on its leverage and coverage ratios as if capital was to be replaced by the issuing of debt.

GETTING UP TO SPEED

AM Best's stress test is being rolled out to rated companies throughout 2006. However, the first companies to be impacted by the requirements for stress-tested capital were those insurers with significant exposure and losses arising from the 2005 hurricanes. These were followed by the Class of 2005, where the absolute minimum requirement for a company to achieve an "A-" rating was a score of 175 in Best's Capital Adequacy Ratio (BCAR). Assuming that a realised need for additional capital is identified, a realistic approach is taken that safeguards policyholder benefits.

In that respect, this is similar to 2005 for both insurers and reinsurers that needed to raise additional capital. In most cases, where a need for additional capital is identified, there is a specific timeframe attached and communicated to the rated insurer. Generally, this is common among both insurers and reinsurers alike. The only difference following the experience of 2005 has been the need to ensure that reinsurers went into the renewal season with an updated rating to best protect the interests of policyholders. The rating community as a whole managed this well considering the limited number of companies that remained under review during the renewal period.

The revised capital requirements will mainly affect property lines, and more specifically insurers who focus on natural catastrophe covers. It is clear that the stress test is likely to increase the level of capital required for this and consequently to reduce both the return on capital and profit margins (assuming that there are no other changes). However, the impact of this methodology in the marketplace is going to be minimal.

Insurers are cognisant of the fact that rates have needed increasing, irrespective of the decisions of rating agencies.

Equally, companies that were in a position to properly assess their exposures and risks assumed were coming to a very similar estimation of the capital needed independent from the views of rating agencies. This was the implicit reasoning behind the strong influx of new capital into the market, which was independent of the methodologies of ratings agencies and more reflective of the prevailing market conditions (that is, the expectation that the industry is now going to go through another phase of stable rates and a relatively benign claims environment). In other words, capital requirements are dictated by reality and claims experience of the insurers and reinsurers affected, rather than the methodology of any rating agency.

- Vasilis Katsipis is assistant general manager at AM Best Europe.

www.globalreinsurance.com
- Storm loss rises unlikely to hit ratings

- AM Best: cat methodologies require rethink