The next step: U.K. insurance regulators are adding modeling requirements to monitor solvency and to evaluate risk. Could U.S. insurers benefit from a similar approach?

A revolution is taking place in the way regulators assess the financial stability of insurance and reinsurance companies in much of the Western world. It is being driven mainly by the Europeans, who are exploiting modern methodologies and software developments to move regulation onto a new level.

The United States, however, is remaining aloof from the process. Perhaps this is because the industry has avoided the scale of insurance company disasters that has afflicted some other major markets in the past or maybe because the method of regulating solvency among 50 states would make the reform process too unwieldy and fragmented to contemplate.

Either way, with recent downfalls of companies such as Reliance, Kemper and Superior National, is it time the United States sits up and takes notice? This question goes beyond asking how best to protect buyers, important as that is. Despite their initial skepticism, insurers who have experienced the new type of regulation find that it has all kinds of benefits. It encourages disciplines and processes that optimize both profitability and the use of capital. It also could improve transparency in compliance with the requirements of Sarbanes-Oxley.

Australia and Canada

The whole world watched as Australia suffered a series of insurance and reinsurance disasters at the end of the past decade, but hardly anyone seemed to notice when in mid-2002 it installed a rigorous new regulatory regime, effectively relicensing all players. Regulation swung in a relatively short space of time from a one-size-fits-all system based mainly on premium income and reserve levels to a much more sophisticated risk-based method of appraisal.

At the heart of the new system is the principle that insurers should assign a risk charge to a variety of sources of risk, with the precise weighting dependent on the type of risk. They must be able to demonstrate to the Australian Prudential Regulatory Authority that they have enough reserves set aside for a 75% probability of sufficiency. They must, in addition, hold capital charges in excess to bring the total probability up to at least 99.5% over a one-year time horizon.

The Canadians, since the mid 1990s, have required a dynamic capital adequacy testing analysis of each company. This approach is more of a stress test of the balance sheet based upon several plausible adverse scenarios and does not address the range of probabilistic events or ask for an opinion on a prescribed threshold being met.

Europe and the United Kingdom

The European Economic Area (consisting of the European Union and such non-EU countries as Switzerland) has, of course, a far bigger insurance and reinsurance industry. It now is going through its own process using many of the same underlying principles as in Australia. Coming under the Solvency II banner, it aims to introduce a risk-based capital approach to insurance regulation in 2009.

The United Kingdom, meanwhile, has decided to take its own initiative. It has, according to some estimates, the world's second-largest property/casualty insurance sector. Lloyd's and the London market are, of course, leading suppliers of reinsurance and surplus lines insurance to the United States.

The first major decision of the U.K.'s new regulator, the Financial Services Authority, when it assumed its powers in 2001, was to introduce a new approach to prudential regulation for the insurance and reinsurance industry. This is coming into force during 2005.

Like the United States, the United Kingdom has decided to use a risk-based capital approach. In addition, however, the FSA requires each insurer and reinsurer to make an internal capital assessment. It is a critical difference.

An internal capital assessment is a form of self-certification, confirming that an insurer or reinsurer has enough capital to support the business it is writing and the other risks inherent in its business. The assessment requires the company to build a model that identifies different potential outcomes based upon the specific risks of the company. (See "How Internal Capital Assessment Could Work," page 62.) This is not too different in principle from company-specific catastrophe models, but in a much broader scale. The list of risks includes the following categories, though their relative importance and weighting will vary from company to company:

* Reserving risk on prior business

* Underwriting risk on planned business

* Small claims below a designated threshold or severity level

* Large claims above the thresholds

* Catastrophe

* Reinsurance programs

* Credit risk on reinsurance past and future

* Asset risk

* Market/Credit risk

* Liquidity risk

* Operational risk

* Group risk (A company may have limited risk on its own business but it may have relationships with other companies within its group that may cause risk to be transferred to it and thus require additional capital to be held.)

While this list may seem similar to the risks the National Association of Insurance Commissioners' risk-based capital formula tries to address, the methodology is fundamentally different. The FSA has not set out a formula by which companies can measure themselves against specified criteria. It expects, instead, that all but the small insurers will create stochastic models for their businesses. These are financial simulation models that enable companies to test the resilience of their operation in any number of possible real-life scenarios, such as high loss experience, stock market or reinsurer failure or adverse claims development.

One of the models' strengths is the ability to take full account of variability and also of the dependency that exists between different types of loss, which proved so devastating to balance sheets after Sept. 11. This enables companies to see where they are most vulnerable and to devise capital-efficient strategies for anticipating extreme events.

The FSA is under no illusions about its ability to rule out the possibility of insurance or reinsurance failure. Trying to achieve such a utopian state of affairs, as well as being a quest for the impossible, would be utterly capital-exhaustive. Instead, the FSA requires companies to demonstrate through their models that they could withstand a one-in-200-year combination of adverse circumstances. This is an approach somewhat similar to how rating agencies have addressed catastrophe management in the United States, but now looking at all risks impacting the balance sheet.

If this were simply an exercise to keep the regulators at bay, it would add no lasting value beyond creating more paperwork. In fact, the FSA is insisting upon good practice, and U.K.-based insurers and reinsurers are coming to appreciate that it has widespread business benefits. Simulation modeling enables the company to understand the drivers behind the business and the likely financial consequences of any change of strategy.

While the NAIC approach is risk-based, it stops short of what the other regulatory entities are trying to do. It does not quantify the true underlying variability potential of a company and what are the actual dependencies between different risks for the company.

This shortcoming has caused companies to see the NAIC approach as a "plug and chug" approach (or just another accounting hoop) to get a final answer without asking, "How is my company different?" and trying to better understand the risk characteristics.

To address this concern, the NAIC has tried to install some soft variability qualifications into its actuarial opinion requirements by asking the actuary to comment on the reasonable range of outcomes or areas for potential adverse development, but this only addresses the area of reserving risk. In its present form, furthermore, the NAIC approach does not ask the actually to consider the full range of potential outcomes or, for that matter, to have the company hold capital to off-set the potential deviations noted in the opinion.

Although insurers and reinsurers do not normally welcome increased regulatory impositions, it may be time regulators, rating agencies and companies look toward a different model that can be beneficial in its own right. Indeed, several Bermuda companies already choose to carry out such modeling exercises. There is no regulatory imperative, but they can see the business advantages of doing so.

By the end of this decade, if everything goes to plan, the insurance and reinsurance industry throughout Western Europe will have introduced a similar regulatory regime to the one described in this article. The American system should take a close look.

Key Points

* The National Association of Insurance Commissioners requires insurers to use a risk-based capital approach to gauge financial stability.

* Other regulatory entities have been exploring ways to enhance she risk-based capital approach.

* In the United Kingdom, the Financial Services Authority will require insurers in 2005 to employ a risk-based capital approach and a form of self-certification that utilizes simulation modeling of all the different risks inherent to the operation.

How Internal Capital Assessment Could Work

Many U.S. insurers and reinsurers already are familiar with the principles of stochastic modeling, having seen a need to run catastrophe models to address rating agency queries. As many companies became more comfortable with this process, they dived into the modeling and the use of a risk measure, probable maximum loss, to really understand the catastrophe risks of their books of business. Companies embraced the technology to design better reinsurance programs, adjust pricing needs, focus marketing efforts, and in general manage their probable maximum losses.

In a similar way, meeting the Financial Services Authority's internal capital assessment requirements may lead to improvements in the following areas:

* Capital adequacy and allocation;

* Product and pricing strategy;

* Scenario testing;

* Reinsurance purchasing;

* Security ratings;

* Business expansion;

* Mergers and acquisitions; and

* Transparency.

In the example shown, the model has produced a simple response to a demanding question, does the company have its product mix right? It shows the likely outcome of a series of alternatives for an insurer that specializes in two lines of business, with strategy four representing the status quo or base. It investigates the capital required, expected returns and profits and the probability of meeting financial targets in a series of different scenarios.

How management reacts to this information will depend on a number of nonactuarial factors, including risk appetite. A reasonable response, however, would be to increase the proportion of business in line one by 10% and make a corresponding reduction in line two as per scenario five. This maximizes the likely return on capital and probability of meeting targets, while reducing the capital required to a minimum.

A similar type of chart or graph could be created for pricing, measuring the relationship between rating levels and profitability. The most revealing areas, however, often tend to be in reinsurance purchase and asset risk. Starting with reinsurance, companies almost invariably find that they are leaving themselves unnecessarily exposed in certain areas, while being over-reinsured in others. The resulting change in purchase strategy often can reduce reinsurance expenditure while also enhancing corporate stability. An internal reinsurance strategy model also can be used as an effective and healthy challenge to the status quo.

One aspect of a rigorous modeling exercise is that it both demands and creates transparency. Of particular interest in the current environment, it will answer the question as to whether a particular finite reinsurance or other alternative risk transfer transaction represents a genuine transfer of risk. Stochastic scenario testing will flush out any failure to provide balance sheet protection. Companies that meet the criteria, on the other hand, will be able to demonstrate compliance with this aspect of Sarbanes-Oxley.

            Line 1 as    Line 2 as
Strategy    % of Base    % of Base

1              70%         130%
2              80%         120%
3              90%         110%
4 (Base)      100%         100%
5             110%          90%
6             120%          80%
7             130%          70%
8             140%          60%
9             150%          50%
10            160%          40%
11            170%          30%
12            180%          20%
13            190%          10%
14            200%           0%

Contributor Tom Hettinger is managing director at EMB America, property/casualty, actuarial consultants belonging to the international EMB Consultancy family.