Economic capital blues

Rating agencies and life insurers have not always got on, with squabbles over how to account for diversification across business lines a prime reason. But could the rise of economic capital management put an end to the arguing?

June 27, 2005, and one of the world's largest insurers, Munich Re, is holding a risk management investors day meeting in London. The company's stakeholders listen intently as Munich Re's chairman of the board, Nikolaus von Bomhard, board member Jorg Schneider and Charlie Shamieh, head of central division integrated risk management, give a presentation outlining their position on economic capital. About a third of the way through comes the gripe about rating agencies. "Limited quantitative recognition of diversification represents a significant barrier to aligning capital positions to the pure economic views," they state. It seems Munich Re is annoyed that rating agencies are not taking account of diversification across business lines when examining economic capital models. Indeed, the presentation mentions a capital adequacy target rating of AA when rating agency Standard & Poor's (S&P) official rating on the company is just A+/stable.

Munich Re declined to be interviewed on the subject, but the criticism of rating agencies comes at a time of great change in the industry in terms of modelling risk. Companies are creating, and increasingly relying on, sophisticated stochastic models to quantify the risks they face. This data is then used to come up with internal capital adequacy numbers. Last month, one of the UK's largest insurers, Prudential, took the unusual step of restating its 2004 results under European Embedded Value and International Financial Reporting Standards. It also attempted to provide a picture of its economic capital position as of the end of last year (see box).

The drive towards sophistication, accountability and transparency is creating immense challenges, as much for the rating agencies, which face shareholder and policyholder criticism for remaining one step behind the industry, as for the insurers themselves. "The rating agencies are probably in a bit of a difficult position, because it's only really been in the last couple of years that economic capital has started to be used on a consistent framework," says York-based Rob Kerry, head of capital management planning and strategy at Norwich Union Life, part of the Aviva Group.

Views are mixed, however, on how effective the rating agencies are. "I think they do a pretty thorough job of talking to management and getting a good grip on what's going on," says Colin Ledlie, head of the actuarial function at Standard Life in Edinburgh, adding that he will not be drawn to comment on past criticisms of the agencies.

So what do the rating agencies think? Have they got a grip on economic capital and are they taking account of diversification properly? Certainly, all three main rating agencies are making moves to change the way they use economic capital as part of the ratings process. At the end of May, S&P issued a note stating that as insurers are driving towards the creation of powerful enterprise-wide risk management tools, which are dynamic rather than static, it too is reviewing its approach to assessing enterprise risk management. However it also warned that the current models suffer from complexity, a lack of market standards and from being subjective, especially regarding the benefits of diversification. S&P is officially maintaining a sceptical stance when it comes to insurers' claims that risk diversification reduces their economic capital requirements by significant amounts.

"Some companies are saying they can carry 20%, 30%, 40% less capital because of diversification. We remain pretty sceptical about diversification effects at those levels," says Rob Jones, S&P's London-based insurance analyst. He does, however, offer some hope to those on the other side. "Over time, it may mean we recognise diversification more than we have done in the past," he concedes.

Evolution

The rush by rating agencies to change the way they assess the financial strength of insurance companies is in line with the way the insurance industry itself is rapidly evolving. Solvency II is forcing European insurers to adopt risk-based capital regimes whether they like it or not. And of course in the UK, the Financial Services Authority introduced risk-based regulation at the beginning of this year.

Meanwhile, the industry has come together and published the European Embedded Value Principles document, a voluntary set of principles for reporting supplementary embedded value, which the chief financial officers of 19 major European insurers have agreed to adhere to. "The move towards European embedded values is making companies assess how much capital they need to hold against different risks," says Kerry.

S&P says its new approach to measuring risk will seek to be dynamic and take account of both qualitative and quantitative aspects of a business, but as yet it will not be drawn on the details of how its process will work. "During the course of this year we will be talking to companies with a view to implementing some new criteria, but we will not prejudge what the outcome will be," says Jones.

Moody's has not made any formal statement, but confirms its insurance division is also re-examining its ratings process. Simon Harris, managing director for insurance at Moody's in London, says that historically, unlike S&P, the agency has not had a capital model, but has instead focused on qualitative factors, such as the strength of the franchise and the quality of the management, in coming up with a financial strength rating. But it is also re-examining its methods. "Around the middle of last year, we started to have some pretty intense discussions with a variety of European and UK groups about their internal capital models, and we are now going down the path of model testing and so on," he says.

Fitch says it is constantly refining the processes and procedures it uses. "We would agree that the modelling we do internally is never going to be as sophisticated as what the major companies do internally," says Harish Gohil, director within Fitch's insurance section in London. "Clearly, they have much more resources and much better access to detailed information, such as individual policy details. We're trying to come up with something more consistent and simpler than the company's own system, but still sophisticated enough that we can compare companies across the market in a consistent way."

But what does that actually mean in practice? The position the rating agencies seem to be taking is to state that they have a 'holistic' approach to measuring risk without giving a formal breakdown of what this entails. "We discuss the models and the key assumptions they use, and we use our knowledge of what different companies do to try and compare and see whether there's any particular assumptions or parameters that need a bit more probing," says Gohil.

"In quite a few cases we've asked companies to stress-test their own models," adds Andrew Murray, another director in Fitch's insurance team. "We've taken their assumptions and gone back to them with our own set of assumptions and said 'Okay, if you didn't assume this level of volatility and instead assumed this other level, what would that do to the economic capital results?'"

As with S&P, Fitch says there is a lot of uncertainty, and as yet very little answers, when it comes to diversification. Gohil says: "Another technical difficulty is that, even if you accept that capital is mobile, how do you encapsulate the diversification in terms of understanding and quantifying the correlations between different assets in different countries, and so on?"

Last August, Fitch actually changed its ratings outlook on the UK life insurance sector from negative to stable, citing improvements in risk management mainly as a result of regulatory pressure. "One of the reasons we changed the outlook on the sector was that enterprise-wide risk management would lead to a better understanding and mitigation of the risks associated with the business," says Gohil.

Rating agencies can perhaps be forgiven for being in the dark when it comes to understanding economic capital management. Norwich Union's Kerry says to get a true and accurate assessment of a company's economic capital position a rating agency would have to "come in and crawl over a large amount of work, which would take them two or three months". He means getting down to a level of granularity where the agencies would be, for instance, collecting economic capital numbers on individual product lines. Clearly that is never going to happen. The hope is that the industry will move to a point where the rating agencies can take a top-level view that gives an accurate reflection of economic capital needs, with such views being consistent from company to company.

At the moment, however, all parties concerned are just starting to find their feet. "There hasn't been reliable information for rating agencies to base anything on," says Hitesh Patel, a partner and insurance industry consultant in KPMG's financial services division. "Most of the models are in varying stages of development and are inherently subjective," he adds.

"A lot of the data that's needed to run these models and truly understand the capital position has perhaps been in the company but has not been accessible, or in other cases not actually collected on a consistent basis," adds Nick Dexter, KPMG's London-based director of life actuarial services. Getting the data consistent across all the entities in a large international group is no mean feat, says Dexter. And it will probably take a few attempts just to get this resourcing right. Kerry says that ultimately the changes currently taking place in the industry will mean policyholders and investors will in future be able to place more faith in insurer financial strength ratings. Until then, a state of flux prevails.

S&P's definitions of insurer financial strength

Prudential's internal target solvency level is set as the historic equivalent default rate on an AA-rated bond. But in what way does this relate to the financial strength of the company? S&P says its long-term insurer financial strength rating levels do not represent specific probabilities that a company may or may not meet its policyholder obligations over a particular timeframe. Instead, it offers the following definition of an insurer financial strength rating: "A Standard & Poor's insurer financial strength rating is a current opinion of the financial security characteristics of an insurance organisation with respect to its ability to pay under its insurance policies and contracts in accordance with their terms." The definition applied to an AA-rated company is: "An insurer rated 'AA' has very strong financial security characteristics, differing only slightly from those rated higher." Above are examples of the current S&P financial strength ratings of some major insurers. Munich Re A+/Stable Standard Life Assuranc A+/Negative Norwich Union Life & Pensions AA/Stable Prudential AA+/Stable

The Prudential Route

At the beginning of June, Prudential restated its 2004 financial information under European Embedded Value and International Financial Reporting Standards, and stated that as a result the group's underlying capital strength, cashflow and dividend policy were not affected. The company said its internal capital assessment showed an economic capital requirement of $1.8 billion, taking into account diversification benefits, compared with available capital of $3.4 billion. It defines economic capital as the amount required to ensure Prudential can meet its existing contractual and discretionary policyholder obligations and remain solvent over a 25-year time horizon, within "a strict target solvency level". Prudential began its economic capital project three years ago. "I think the main driver has been from the business side, but the response has also come from the way regulations are moving," says James Matthews, the company's London-based head of investor relations. The framework comes down to projecting cashflows and capital requirements for each of its main business units in the UK, Asia and the US over a number of stochastically generated simulations. This process, says the company, uses a group solvency model and captures 80% of the business. The remaining 20% is modelled on a stand-alone basis and aggregated with the main results using a correlation matrix approach. For each simulation, and in each of the projected 25 years, business units calculate their capital surplus, transfers or requirements using their own asset-liability models. The projected capital transfers to and from business units are aggregated at group level along with group level cashflows, such as interest on debt and expenses. This gives a group level capital balance for each scenario in each of the 25 years. To take account of capital mobility restrictions, Prudential says capital transfers to and from business units are triggered at a solvency level "that reflects a suitable level of operating capital, based on local regulatory solvency targets, over and above basic liabilities". The group says it uses an iterative modelling process to calculate economic capital as the amount required, such that the cumulative number of projected defaults (a negative group capital balance) is less than a predetermined rate reflecting its internal target solvency level - set as equivalent to the historic default rate on a AA-rated bond (equivalent to a cumulative probability of default of 44 out of 1,000 simulations over 25 years). Prudential says its economic capital model covers all material risks in each business unit, including asset-liability matching, credit risk, underwriting, persistency and operational risk. Prudential splits its economic capital requirements into the following risk types: Risk type % of Group economic capital requirement Asset-liability matching 28% Credit 47% Underwriting (mortality, longevity and morbidity) 10% Persistency 2% Operational 13% Source: Prudential