Transcript

Transcript of an IMF Economic Forum: Managing Financial Risks--The Insurance Industry

June 30, 2004

MR. TRAN: Good afternoon, ladies and gentlemen. Welcome to the IMF Economic Forum on managing financial risks, the case of the insurance industry. My name is Hung Tran, and I'm going to moderate this Economic Forum. And let me quickly explain to you how the agenda is going to be organized.

First, I will give you a some background information, including a short explanation of why the Fund, the IMF, is interested in this kind of topic and why we are holding this Economic Forum on this subject at this point in time. And, secondly, to give you short summaries of the key themes and issues that will lay the stage for the discussion from our panelists and then the discussion from you, the audience, later on. And then I will ask our distinguished panelists to give their presentation, about 10 minutes each. Afterwards, I will open it up for questions and comments from the audience and also from among the panelists themselves. And let's hope that we have an interesting and lively discussion.

First, why is the Fund interested in this subject, the insurance industry? At the Fund, and particularly in our department, which is the International Capital Markets Department, one of the things that we do is to monitor and analyze global financial markets developments with a view to identifying potential systemic weaknesses. And we try to alert the international financial community, the authorities, but also market participants, with these potential problems in the hope that remedial actions could be taken in time so as to prevent a bigger problem later on. In the jargon at the Fund, we call it "multilateral surveillance of international financial markets" aiming at crisis prevention.

In monitoring global financial market developments, we look at both the short-term, conjunctural developments--for example, what would be the impact of the Fed action on financial markets--and more structural and secular changes in the international financial system. And we look at these issues both in the context of what's going on in the mature market economies and, of course, the emerging market economies, and increasingly so, the spillover and the interlinkages between one to the other.

One of the most important changes in recent years in the international financial system, in our view, has been the blurring of the demarcation line between different sectors of the financial services industry and also the steady transfer of risk from the banking sector to various nonbanking sectors, including ultimately to the household sector.

This phenomenon first caught the attention of many observers when credit derivatives, particularly credit default swaps, were highlighted a few years back as something that banks in general are trying to use to reduce or to manage their credit risk exposure and other nonbank institutions, including insurance companies, were found to be net suppliers of these default protections or increasing their net exposure to these credit risks. We will hear more about this later on from one of our panelists, the scale and the distribution of this issue. But also, more importantly, in our view, when we begin to do work on this area, a steady, relative reallocation of credit risk or credit instruments from the banking sector to various nonbanking sectors, including first and foremost the insurance industry.

In fact, in the U.S. the insurance industry has for a few years held more non-farm corporate credit risk than the banking sector all together. This kind of development raised several questions: Where are the risks being distributed if it is being distributed away from the banking sector, which is, by and large, more transparent, better regulated, and so on and so forth? Are the recipients of this risk able and equally equipped to manage and to deal with the risks that they take on? In the process of change, are there possibilities of new gaps or vulnerabilities being opened up that an international financial institution like the Fund should be aware of? And at the end, what is the overall impact of these changes on global financial stability?

To try to find some answers to these questions, we have started to do work on the phenomenon of transferring risk and risk-taking and overall financial management by various nonbanking sectors. We started last time, about a few months ago, looking at the insurance industry. We published our findings in our report, in the Global Financial Stability Report, a copy of which you will find outside of the auditorium. You can follow it up by looking at our public website. You can subscribe to get the nice colorful hard copies by writing to our External Relations Department. But the last time we published our findings on the insurance industry, and the findings plus the discussion that we had with our panelists will form the gist of the discussion we have this afternoon here for you.

Next time, in September, when the next issue of the GFSR is coming out, we will look at the pension industry, leading up to early next year, in March 2005, with more detailed looks into the financial activities and financial management of mutual funds/hedge funds, and, finally, how the risk is being devolved into the household sector.

For each of the sectors we look at, basically we want to understand better and to identify the factors, market structures, market characteristics, regulations, accounting standards, and so on that shape the financial market activities, the risk-taking as well as the risk management culture and risk management capabilities of the institutions in that sector, and to see if the changes that I described briefly a while ago give rise to any new risks or any impact, potential impact on financial stability.

For the insurance industry specifically, some of our key findings which I will very briefly summarize here for you and set the stage for the presentations of our panelists and which you can find in more detail reading through Chapter III of the GFSR publication you have in your hand.

One, there are differences and sometimes significance differences in the composition of the asset portfolios of different insurance companies or national insurance systems across countries and regions, driven mainly by two key factors: one is market characteristics of the region or of the countries; and, secondly, by regulations.

The second point, we found that overall for insurance industry, because of the relatively longer duration for their liabilities, credit instruments appear to be an appropriate instrument for them to hold because they can more easily match that with their liabilities.

The third point, we found that in countries or in systems where a well-developed credit market with risk-based capital and accounting regimes exist for some time, these things have encouraged or provided very strong incentives for insurers to develop their own credit management capabilities and led them to allocate a higher share of their asset portfolio to credit instruments.

And that led to the next key finding that we identified, that national insurance systems having those features tended to be more stable compared to systems where the asset side of the insurer's portfolio is more exposed to equities, among other things. And particularly during the last few years, from early 2000 until late 2003, the decline in equity markets in many major countries, particularly in Europe, led many insurance companies, again, in Europe, having had higher exposure on their asset side to equity instruments, to suffer losses and to some extent to the degree that they ran into a solvency crisis, requiring their regulators and supervisors to intervene to help them.

Therefore, we feel also that driven by this experience, many insurance companies, again, mainly in Europe, have taken measures to increase their capital, to strengthen their risk management capabilities, and to allocate more in terms of their asset composition to credit products. And by credit products, I mean corporate bonds, other credit instruments, including credit derivatives. And in doing so, the regulators in those countries are also taking steps, particularly in the case of the U.K., with the Consultative Paper 195, within the European Union in the context of Solvency II project, are also taking steps to move toward a risk-based capital regime.

All together, we feel that these developments have definitely reduced the risk of the balance sheet pressures that these insurance companies underwent in the previous years. And if you put that next to the view or the assessment that for the banking sector proper, they have weathered the turbulence in financial markets in recent years very well, with much less problem and insolvencies than any previous downturn in the credit cycle or market cycle. The conclusion we have is that, by and large, the transfer of risk from banking to, in this case, the insurance sector seems to have been a positive development from the point of view of global financial stability.

So those are the basic themes that I just want to sketch out for you to set the stage for our discussion.

We are very happy and grateful to have our four distinguished panelists to be here with us today, taking time out from their very busy schedule. Mr. Charles Lucas, Director of Market Risk Management at the insurance company AIG, will be the first panelist to present his view, basically to say to us or describe to us how insurance companies, the practitioner, look and actually do and manage risk in their day-to-day business.

Second, Grace Osborne, Director of Advanced Analytical Group, Insurance Practice, Standard & Poor's, will give a view of how rating agencies go about looking at insurance companies, particularly in their risk-taking and risk management practice to form a view of their creditworthiness.

And then David Strachan, Director of Life Insurance Division of the U.K. Financial Services Authority, will give a view of how regulators/supervisors look at these issues from a regulatory perspective.

And last, but not least, Keith Buckley, Managing Director of the Insurance Practice Group of Fitch Ratings, will give us some updated information and assessment on the very useful and comprehensive global survey of credit derivatives that Fitch released a year ago.

So that will round up the presentations from the panelists. Each will speak for about 10 minutes. Afterwards, I will be happy to open up for questions and comments from the floor.

Chuck, the floor is yours.

MR. LUCAS: Good afternoon to you all. It's a great pleasure to be here with you. Let me first say to my friend Mr. Groome of the IMF that I spent the first 25 years of my career as a central banker and actually have written a number of the kinds of surveys that Todd has done in Chapter III of this recent survey. And it's an excellent job. It poses a set of policy issues that could take quite a while to discuss fully, and, you know, Todd said, "Don't worry. You've got 10 minutes."

[Laughter.]

MR. LUCAS: So I'm going to have to focus here a little bit. I have actually condensed this talk out of two others that were scheduled for a total of three hours. So what I'm going to try to do is respond to the question that Todd asked me, so I'm going to demonstrate to you I actually heard the question.

The question he asked me is basically how does an entity like AIG make use of capital markets in the process of managing its financial risk. Phrased that way, it's a very broad question. And I'm going to keep the discussion at a very broad level and skip over things. And if anybody wants to push into any level of detail on some aspect of it, we can do that in the question-and-answer period.

This is the basic message that I want to leave with you: There is a huge change undergoing in the insurance industry in the way in which risk is managed, and talking about asset and liability risk in the sense in which a finance theorist would talk about it. There is absolutely no doubt about that. But the underlying characteristics of the way in which insurance companies are being managed I do not believe are changing very much.

One of the comments that is made in Todd's paper right at the beginning bears on the question of whether in this turmoil the transfer of risk is resulting in a net reduction or increase of risk in the system. And the couple of bullet points there are axioms in my mind that I think, you know, should be kept in mind right straight through this process.

Any capital markets transaction of any type is only a transfer of risk. It neither increases nor reduces the risk in the system in and of itself. It is simply a transfer from one entity to another.

The systemic issue, which has to do with the aggregation of all of those risks, in my mind has in the first instance and of dominating importance the question of whether or not in the process of that transfer risk concentrations occur or are diversified.

I'm not speaking for AIG here. This is just my own personal view. Obviously, I work for AIG, so they can't be terribly inconsistent with the views of the company. But as a formal matter, I am expressing personal views.

Just to check--I don't know the audience at all well--how many have ever heard of AIG? Okay. That's actually an impressive number. AIG is--no, it's an interesting phenomenon, because it's a very, very large company. It's the largest insurance company in the world. But the name recognition is not high. But it is a company with $700 billion in assets and operates in 150 countries.

Here's kind of the basic three points. I will comment a little bit more on credit derivatives, but the important fundamental point for an entity like AIG, and I think of the other large insurance companies this would be true, is that the use of credit derivatives as a part of credit risk management or asset management in the insurance part of the business is negligible. It's not zero, but in terms of, say, a notional amount compared to asset holdings, is in the low, single-digit percents. It is the underlying deep capital markets that are the core of the problem from the point of view of an entity like AIG. And I'm going to talk a little bit about some of the changes over the last 30 years in the capital markets which have affected the way in which we manage risk.

But really the part that I want to emphasize more than any other, I think, is that what capital markets bring to an entity like AIG is technology in the sense of finance theory that can be applied to a whole variety of risk management problems within the entity. And, in fact, my department, which is basically a risk management shop, a small army full of Ph.D.s in mathematical physics and applied mathematics, building risk management models, and about 60 or 65 percent of those resources are allocated to insurance liabilities, which are embedded option products where finance theory applies reasonably directly. So it is that aspect of what the capital markets have produced for the world that is, in fact, most important to the company.

There's a little bit of controversy, I guess, in the second bullet there, but it does, by the way, go to the heart of some of Mr. Tran's discussion earlier about Europe. It's tough to make money in the insurance business at the risk-free rate. It's tough to issue insurance liabilities and invest at the risk-free rate and earn a spread. It's for that reason that insurance companies that I know anything about--and AIG is a U.S.-based company even though it operates very broadly in the world--is fundamentally a credit risk intermediary. Of the $700 billion in our asset portfolio, virtually none of it is risk-free. It is almost all, in one form or another, a credit risk instrument.

That has been the case for the existence, the 85-year history of the company, and the appearance in the marketplace of credit derivatives as an incremental asset class, if you will, is a trivial incident in relationship to that overall picture.

Even worse, from our point of view, the existence of credit derivatives does not in the credit risk sense expand market opportunities. It may increase the liquidity of a name a little bit, but that's about all that change. Far and away the dominant share of credit derivatives that are traded in the marketplace are against names that have underlying bonds trading in liquid markets already. And the reason for that--anybody who knows capital markets understands this going in--is the derivatives trader has to have an underline in order to price the derivative. So the derivatives get created on the basis of credits that are already liquid. So it doesn't really add very much from the point of view of an insurance company. It probably does for other asset holders who are much more trading oriented. Insurance companies tend to be buy and hold.

This gets to very deep questions in terms of pricing of insurance liabilities, by the way, because actuarial assumptions typically are not consistent with risk-neutral pricing assumptions.

The really important part of what you could call credit derivatives or credit modeling kinds of things more broadly is things like not single-name default swaps but things like CDOs, collateralized debt obligations, which really are for a company like AIG simply an extension of a pre-existing securitization technology into the credit markets. And I'm sure there's at least some of you who are familiar with the underlying economics of the CMO market in mortgages, where it is actually possible to take a bundle of mortgage-based securities, which are already a securitization, and create a special purpose vehicle to hold those assets, issue liabilities, which are structured differently than the underlying securities, and sell those tranches and make money in the process. It's an important thing. The dealers that do this make money in the process, which is to say the value, the aggregate value of the restructured assets exceeds by some amount--small, but a positive amount--the value of the underlying mortgage-backed securities that collateralize the risk.

That type of entity has no mortgage exposure at all. As a matter of fact, it ends up that's why they are AAA subsidiaries with basically no net exposure. But in the process of creating a structure, it adds value.

This same concept is applied to something like a CDO where underlying assets trading in a marketplace can be repackaged and restructured and sold. And it does create a slightly different twist on the asset structure, the risk structure of an asset, which can be of interest to an investing entity. So a company like AIG will use those asset opportunities to fine-tune the risk composition of its portfolio.

The interesting side of the question, I think, really is, I think, the liability side of particularly life businesses, life and annuity businesses in insurance. It's in that area where in my view capital markets technologies are invading traditional businesses and causing companies to completely rethink the way in which they construct and risk-manage those classes of products.

Effectively what those products are, they really aren't products, they're features of products. They're guarantees, they're financial guarantees of various types. There are minimum benefits or withdrawal benefits or accrual benefits in an annuity product or a life product that are effectively quasi-financial options. And it's a class of exposure which is appropriately modeled by finance theory-based models which are inherently stochastic models. Actuarial approaches tend to be deterministic approaches. So effectively what you're doing is taking standard finance theory as it exists in the capital markets and modifying it as appropriate to deal with the special features of a product that appear in the context of an insurance contract and, in fact, estimating fair value and risk associated with that liability. That creates a foundation for hedging of risk using capital markets instruments directly, and it is in that form in which capital markets are really penetrating the insurance industry most immediately.

The next to the last bullet there is to me the largest issue associated with this trend. It is the degree of culture change that this challenge poses for an insurance industry. It is not unlike what happened in the banking industry 25 years ago, 20 to 25 years ago, when financial options appeared in the banking industry, and bankers had to learn to think in very, very different ways and develop different analytic tools to manage the risk successfully. Or you could take the question back yet another generation and discuss it in terms of the thrift institution crisis in the United States in the 1970s, where what fundamentally drove that problem was the existence in a mortgage contract of an unrecognized financial option in the form of prepayment rights or the rights to sell, to transfer a mortgage to another buyer of a building. So it's an extension risk, prepayment risk, two-way risk that is embedded option. But we all remember how much turmoil and difficulty there was in the industry, in that industry when it had to recognize the existence of those financial obligations and figure out how to value and hedge them. I am not predicting anything.

Actually, Todd said 12 minutes. You said 10 minutes. I'll finish up with this.

The option risk--since this is one of the fundamental differences between this class of exposure for an insurance company and a traditional insurance company, a traditional insurance exposure. A traditional insurance exposure is inherently diversifiable. You know, if you do auto insurance, you have a probability distribution of how many accidents will occur, you know, per million passenger miles or however you want to measure it. And you can write a policy which can be expected to perform in the context of a portfolio on the basis of the portfolio characteristics. It's a diversifiable risk. Option risk isn't. These are market risks if the equity market goes down, the fact that you've written a lot of guarantees all at the same level means that all of the liabilities gain value to the policy holder at the same time. So you have the opposite of the kind of diversification that occurs in a conventional insurance product, and that is why it is so difficult to change the way in which these risk management approaches are executed within an insurance company because you have to think about the problem in a completely different way and manage the risk in a completely different way.

I'll skip over that. You can read those. I don't have to read them. Those are the two particular points that I wanted to leave with you, and I guess I am completely out of time.

[Applause.]

MS. OSBORNE: Good afternoon. This afternoon in my presentation, I plan on giving you a very quick overview of the rating process at Standard & Poor's, a recap, if you will, of the health of the life insurance industry from our perspective for both North America and Europe, and then really delve more into the tools that we use to assess the risk at the life insurance entities that we rate, and then speak, if we have time, of course, to some of the challenges ahead for the life insurance industry.

Very succinctly, on this particular slide I've identified four or five constituents of our ratings, but at the end of the day, the ultimate goal behind what a rating agency needs to do is to develop a timely, objective, well-informed assessment of the credit risk associated with the entity we're rating--or it could be the issue that we're rating--and provide that to the market so that they can make informed decisions from whatever perspective they're looking at it.

Standard & Poor's, all of our models and meetings are on our website, and any of our formulas and conclusions that we derive from our modeling as it relates to an insurance entity would be provided to that company at the end of each analytical process.

Very quickly, since there will be a slide a little later just talking about our ratings, just to put everybody on the same level, we have four ratings that would indicate a financial strength rating is in the secure category and four that would be suggesting that it's vulnerable. You will see with Standard & Poor's, you'll have pluses and minuses that will be associated from the spectrum of very strong to very weak, and that's just trying to give the user of the rating more information regarding the degree to which we think that the financial strength, relative, perhaps, to some other competitor, insurance company or issue that they're looking at, so they can put it on some sort of a spectrum.

Okay. What do we actually do? And how do we do it? We have at Standard & Poor's a systematic process that is adhered to globally. So no matter whether your rating agency analyst is out of the New York office or Tokyo or London, we are using the same fundamental approach. We will assess on six different factors--competitive position, management corporate strategy, operating performance, investment analysis, capital and reserve adequacy and liquidity, and financial flexibility. Clearly here you will see that it is a composition of both quantitative and qualitative factors. There's always a focus on the models and the output and how that's trying to either, you know, pinhole, if you will, the particular rating that a company has, but it really is--the rating will be bringing all of these particular factors in and weighing them as deemed appropriate.

From a spectrum here--and these would be financial strength ratings on specific individual live insurance operations as of last Friday--you can see that the North American sector for life insurance is rated higher than our counterparts in Europe. I think there's probably two key things that are driving that rating distribution. First, Hung had mentioned it earlier in his opening remarks of their findings was certainly the Europeans were heavily impacted by losses related to their investment strategy where they emphasized more equity holdings, and so their capital positions were absolutely devastated in the early 2000s.

Also, I would say partially but maybe not dramatically causing this is the fact that a lot of the European life insurance companies are parts of global groups, and so you would also have non-life operations that would be part of their group. And inevitably, when you're part of a group, you may have to be at times subsidizing other parts of the organization, and certainly the non-life insurance sector had some extreme difficulties, both between '97 and 2001 in particular. So I think there were a couple reasons why we would see--but certainly I think the key take-away here is looking at the life insurance sector, it's clearly a secure sector from our perspective.

This is just an overview of the strengths that we see in the life insurance industry: strong, recovering capitalization; credit trends are good; certainly improved asset quality that we're certainly noticing in the past year or so. We are expecting that the rise in interest rates will be gradual; if this is the case, then it's a positive. You'll see my next slide. If it's not, I have to hedge it.

Also, we're seeing improved operating efficiency. There has been significant merger and acquisition and consolidation activity in the life insurance sector, and we are seeing that companies are far more focused. They're able to provide their product on a more efficient basis. And I do think all these key executives in the various operations have really learned and seen others go through a very difficult time, and we suspect there will be far more financial discipline with the management teams prospectively.

Challenges, and there are plenty. Reducing credit spreads. They have been tightening. I have a slide a little later that will kind of graphically show that. Uncertain interest rate outlook here. If there is a sharp increase, I think this would have a significant adverse impact for the life insurance sector. We're not expecting equity markets to provide too much oomph in their investment portfolios. So they're really going to need to make sure that their products are priced appropriately.

We do see evidence of increasing product risk and reinsurers backing away. So the life insurers will be, in our view, holding on to this risk.

We are certainly seeing increasing regulatory burden. I have another slide that kind of outlines that in just a couple moments. And then just so I have a little bit of a balance here with the European insurers, they are going through a traumatic time right now where the various regional accounting standards are trying to converge and address and follow a more common practice. So we should see--it's likely to see that there will be restatements of financial statements. There will be a different viewpoint of what's the financial strength of an entity.

What we do at a rating agency is, in a sense, we get our first financial statements whenever the company has published, and then we really start to go in and try to kind of conform them to be more like what we might see in the U.S. So there are a lot of adjustments that we do in order to come up with what we think is the security of the company.

I only added here or included for this presentation our midyear outlook, which we just published last week or the week before, for North America, only because my European colleagues really do it by individual regions, so they'll talk about the French market, the German market, and they'll mix it in between life and non-life business.So just for ease of presentation, I'm only including the North American. And we have moved that to stable, which is suggesting that if there are downgrades, there will be an equal number of upgrades. And so that will be more of a company-specific issue that will drive a change in rating rather than a macro kind of economic situation.

Really quick here, I know we'll be pressed for time, so this is just a historical perspective showing you the pooling investment yields that these life insurance companies had to face. The good news, I guess, is that with rising rates, we will start to see that somewhat improving. We will also expect that because of this pressure, if you will, on the investment portfolio, life insurance companies really started to make different selections on the asset allocations that they were pursuing. And we also saw that for some companies--and this is definitely on an individual basis--there might have been a little more loosening of the tolerance, of the mismatch between their assets and liabilities, just in the hopes of trying to get a better yield. So we'll have to see how that plays out as the market starts to change now.

This is suggesting here--I made a comment earlier about bond spreads tightening, and they certainly have. But you can see, if you just look at it over a couple years, that it had tightened far more in the mid to late 1990s, and so there is a question whether we might see some more of that.

Rising interest rates is a fear. This one, if it's too rapid, we at S&P are not expecting that to occur. We really do think the Fed has been very vocal in trying to make sure that both financial institutions, including banks, let's say, and insurance companies, were clearly aware of the direction and were taking corrective action.

This is just very quickly some of the implications for the industry. You know, near term it's positive, so there will be a reduction on the pressure on spread income. Certainly rising annuity surrenders, how will companies react? Will they change from their pricing discipline? That will be something we will be looking for, and certainly looking at companies' active asset/liability management has been key as we've analyzed life insurance entities, and we don't suspect that is going to change prospectively.

So now really getting to the heart of what do we actually do when we go into a life insurance company and try to figure out what capital needs to be allocated for the risks that they're assuming. What we do is--and this is an approach that is adopted by all of our analysts. The factors that we would elect to use will vary from region to region, only because it's just recognizing that the assets, let's say, a French insurer may be investing in would look different than perhaps what a U.S. insurance company will. And so the factors that would be applied to those investments will be reflective of our assessment of the asset default risk that would be associated or the volatility where there is equity securities. So some markets are--you know, there's a lot less liquidity. Some markets there is just no ability for companies to really quickly change their investment strategy and, hence, a sudden movement could increase their volatility.

So what we'll do is we'll look at their statutory surplus or their gap equity. We will make adjustments based on their double leverage. This is something that regulators, at least in the U.S., don't focus on, but we certainly do at S&P, where traditionally an insurance company who wants to--oh, gosh. Okay. Well, real quick, one more second. Then we have additional tools that look at certain segments of a life insurance portfolio so that we can do a tailored look and look at both the assets and the liabilities and really how the net risk exposure is at the company.

I'll tell you what. You have all my handouts there, and I guess I thought I was going to talk quicker than I did. So at the question-and-answer, if you have any questions as you peruse that as people are walking up, not that you're going to look at when other people are--you know, feel free to ask me.

[Applause.]

MR. STRACHAN: Good afternoon. I'm David Strachan from the Financial Services Authority. What I'm going to do very briefly is talk a bit about the experience in the United Kingdom of how our life insurance industry -- [tape ends].

-- doing so, I hope I can illustrate some of the issues that other speakers have been referring to in their remarks.

But at the outset, I'd just like to take two minutes and talk to you about the FSA because we may not be familiar to you. We're a single regulator for all forms of financial services businesses. So we cover banks, insurance companies, investment banks, you name it, prepaid funeral plans, we just about regulate it.

We look at financial soundness, at capital adequacy, but also consumer protection in terms of the sales practices which firms use in relation to their customers.

The legislation gives us four statutory objectives which govern our activities, and they are protecting consumers, maintaining confidence in the market, promoting public awareness of the financial system, and reducing the scope for that system to be used to facilitate financial crime. And those activities govern everything that we do. I mention them because they're relevant to a couple of things I'll say later.

Traditionally, as Grace mentioned in her presentation, the U.K. life industry has held a high proportion of its assets in equities. It wouldn't be unusual, looking back a few years, to see equity backing ratios of 70 percent, 80 percent, or indeed one or two isolated cases almost 90 percent of asset portfolios. Partly that's a reflection of U.K. market structure, but I think also partly a legacy of periods in the United Kingdom of extremely high inflation in the '70s and '80s and the early part of the 1990s.

Now, that high equity market exposure had for a number of years stood the life insurance and indeed their policy holders in extremely good stead. I mean, they had benefited particularly from the bull market in the 1990s, feeding through into returns on policies. But it did mean that the bear market that set in in 2000 and lasted into the spring of 2003 created significant difficulties for our industry.

The position, I have to say, wasn't really helped by the fact that the capital adequacy regime for life insurance companies, which was in place at the time, was not terribly responsive. I'm not going to go into the detail today. I don't think that's really relevant. But I will make a couple of points which I think have wider application.

First, the operation of something that we call the resilience test--think of it as a fairly crude market risk stress test--meant that as equity markets fail, insurance companies came under pressure to move into less risky assets, to which they responded, by and large, by selling equities, even when that went against the best judgment of the management of the companies concern.

That was giving rise to two things. First of all, it gave other market participants a fairly easy target to aim at, and, secondly, what we saw was the risk of fueling a downward spiral in equity markets.

Secondly, although assets responded fully and immediately to changes in prices, because we operate on a marked-to-market basis, liabilities didn't. Even though following a sustained fall in equity markets, the benefits paid to insurance policy holders typically also fell. So what we had was a mismatch. The assets were falling sharply; liabilities were not, even though in reality those liabilities would have been lower because of decreased payouts in the future.

Well, what did we do about it? Well, basically two things. We modified and at times suspended the operation of the stress test, the resilience test, and we also made it possible for insurance companies to apply to us on an individual basis, not market-wide, for disapplication of some of these rules, but only if the company could demonstrate that it remained strong on a realistic basis. And what a realistic basis is I'll explain in a couple seconds.

Did our actions have a positive impact? Well, one of my colleagues was in Denver when we announced that we stood ready to make these changes, and by his account, a CNBC reporter said, well, I'm not sure what the FSA is or what the initials stand for--which proves that we don't have the brand recognition of AIG--but the letter that they've just sent out to market participants announcing these changes was surely one of the key movers of the market. So I suppose if CNBC says it's true, then it must be true, but I do think that that did have a significantly beneficial effect on the market and indeed was consistent with the need to pursue our market confidence objective.

So that's the past. What about the future? Well, we're moving to a new capital adequacy framework which will take effect at the end of this year, and indeed we are publishing the final rules on Friday of this week.

For those of you with memories of a particular David Lynch program, we're going to have a so-called "Twin Peaks" calculation. That will consist of a regulatory peak based on the EU directives which govern everything that we do, and indeed which will provide an absolute minimum capital adequacy requirement. But then the second peak of the twin peaks is going to be the so-called realistic peak, and that represents an assessment of expected liabilities arising both from contractual guarantees and a fair provision for benefits that the companies expect to pay, such as future annual internal bonuses.

Now, those of you with a keen interest in this can have a look on our website at these rules on Friday, but, in essence, the main changes embodied in this realistic calculation are three-fold:

First, as I mentioned, the need to include fair provision for expected discretionary payouts. The previous approach didn't require firms to provide for these future discretionary payments, even though both the firm itself, its management, and its policy holders expected such payments to be made. And the basis on which these payments will be made are set out now quite publicly in documents that the insurance companies make available to their policy holders.

The second--and this, I think, very much goes back to a point that Chuck Lucas was making--we're requiring insurance companies to move to more modern market-consistent techniques for valuing options and guarantees. As I'm sure you know, guaranteed values and benefits are a major issue for many European life insurers. We are absolutely no exception to that and could indeed form a fascinating seminar in its own right. But leaving that aside, until recently even some of the largest firms used pretty unsophisticated techniques for valuing options and guarantees, though many of them use fairly simple deterministic methods, and it's only recently that they're beginning to adopt stochastic modeling techniques. And that is changing very rapidly, and it's true to say that it's making some managements and boards look very carefully at their balance sheets in the light of the results of these calculations.

The third area is that as well as a realistic valuation of the assets and liabilities, you also need an explicit capital requirement, a capital buffer on top of that. There we are requiring a risk capital margin, arrived at following the application of a number of stress tests, and those tests cover equity risk, interest rate risk, real estate risk, credit or counterparty risk, and, indeed, changes in persistency rates, rates at which policies are surrendered.

We're satisfied that those five areas do cover the material risks in a U.K. life insurer, and we've confirmed that by some independent work we had done by Watson Wyatt, which we will publish.

Now, there's clearly some market and policy holder consequences from these changes. Moving from a relatively unsophisticated, unresponsive capital framework to a risk-based on is bound to have such impacts. What are they likely to be?

Well, first--indeed, as we've already seen--all firms have to a greater or lesser extent rebalanced portfolios. I was talking about equity backing ratios of perhaps around 70 percent a few years ago. As of the end of 2003, that's down to around 35, 38 percent, though, again, there are still some above that, but that is the average figure. So that's quite a significant change.

But I really don't think that's just about regulation. I think it's also a fraction of changes in the inflation outlook, and indeed, the fact that the guarantees in the products are now biting much more now than in the past and, indeed, are being covered increasingly by bond investments by the companies themselves.

Second, there's a huge amount of portfolio management, portfolio restructuring and hedging of market and other risks going on, which is extremely good business for the investment banks.

Third, policy holder benefits have been reduced, and in some cases, companies are now explicitly charging for these guarantees embedded in the products.

Are the policy holders bearing more of the risk than previously? Well, I think it's more complicated than it first appears. Clearly, to the extent that firms are now less able to smooth the volatility of returns, then, yes, policy holders are more directly exposed to market movements.

Similarly, some are now explicitly charging for guarantees, as I mentioned, but much of the debate certainly in the U.K. has been affected by the changes that we've introduced to make the products much more transparent, to make explicit what was previously there but probably less obvious, less explicit to the policy holders concerned.

And that I think takes me to my concluding point, coming back to one of our statutory objectives, and that's really the importance of consumer understanding of the risks and rewards of financial products. As a general proposition, the consumer is being expected to bear more of the risk that hitherto would have been borne by government, by employers, or indeed, by financial institutions. I mean, that seems to me to be a very clear trend and one that's not going to be reversed.

As I mentioned, we have this objective of promoting awareness and understanding of the financial system, and that's an area that we're spending an increasing amount of time and money on in terms of consumer education, working in partnership with other relevant bodies in this area. Clearly, there are absolutely no quick wins. It takes a long time to imbue a greater understanding of the risks and rewards, but we certainly see it as one of our major priorities.

And before somebody waves a piece of paper at me, I'm going to stop there. Thank you very much.

[Applause.]

MR. BUCKLEY: Hi, everybody. Again, I'm Keith Buckley. I'm Managing Director and I head the Insurance Group at Fitch. And what I'd like to talk to you about today is not necessarily insurance specifically, but more to give you a sense of what we view as far as the global trends with credit derivatives. I know it's a narrow topic but one that's certainly relevant when you're thinking about newer or evolving types of risks that certainly plays back and has significant implications ultimately for different sectors and different companies in the insurance industry.

So what I'd like to do is to break it down into six broader areas: first, give you a real quick overview of the market in a survey that we've conducted; then talk a little bit about market flows, who's buying and who's selling the risk and what types; very briefly talk about what we saw as some of the key reference entities, credit events, and counterparties involved in this market; and then finally close on a couple of issues that we're thinking about related to credit derivatives.

So, first, as far as the market itself, I think there's--you know, everyone here would recognize that it's a very rapidly growing but still pretty immature market. If you look back in 2002, there were probably under $2 trillion of notional exposure. If you look into 2003, we estimate about $3.5 trillion. So, again, very rapid growth, about 75 percent.

Given this, we think there's a potential for both risk concentrations as well as risk dispersion, which I think ties in very nicely with the comments we heard earlier from Charles that this is really a capital market transaction and risk will be raised or lowered depending on where it goes and how it's built up.

We recognize, too, that as capital and, you know, risk is flowing between entities, there's a potential for information asymmetries. Basically, the companies that are originating the credit risk theoretically have an information advantage versus those that could be acquiring it through a sale. That's not necessarily the case, but it's one of the key things that we're keeping our eyes open for.

Because of that, too, we recognize that there's low transparency and disclosure at this point related to credit derivative risk. If you go through the annual reports and footnote disclosures of companies, you can't really track where this is moving. You get certain disclosures on notionals and market value changes, but not much more than that from a lot of companies. And I'll be talking about how we address that issue.

Because of that, it is hard to, as I mentioned, track on flows by region and sector. I think there's been a lot of anecdotal discussion, and certainly our analyses have indicated that banks are the main beneficiaries, that they're the ones who originate a lot of credit risk and ultimately want to relieve themselves of some of it. And certainly the risk sellers or those taking the risk are not always apparent. Those we'll talk about in a minute. The insurance industry certainly is on the end as far as the sellers. But when you hear a lot of the anecdotal discussions about a credit event occurring, banks are often to raise their hand that they've shed the risk, but you don't see a lot of those raising their hand saying we're the ones that took it on.

Because of some of the disclosure issues, we launched a pretty comprehensive credit derivatives survey that we published in two stages last year. We targeted about 200 financial institutions and identified about $1.8 trillion of notional. So probably not a bad starting point.

Our goal was to better identify what some of the market flows and who the market participants were, and I think we got a pretty good start on that--not perfect but pretty good.

The emphasis in our analysis was on protection of sellers or those taking on the risk, because that's where we want to recognize in our business of rating companies if someone is taking on a disproportionate exposure.

Now, the information that I'll be presenting to you today is based on the survey published in 2003 as of 9/30/02, but we're in the process of updating the survey based on year-end '03 numbers; and where I can, I'll try to give you some anecdotal evidence on where things are shaking out in the new numbers. But I think the big-picture take-away is that the general themes and trends that we're talking about based on the '02 data are carrying through into the '03 data as well. So I think as far as getting into the nitty-gritty, until we're able to publish and dissect all the information, there probably won't be great revelations at this point.

As far as the global credit risk transfer, what we observed was that, as I mentioned, global banks and broker-dealers are basically transferring credit risk to three key areas: the financial guarantors, the mono-line bond insurers; the insurance and reinsurance industry; and to a lesser extent, but interestingly enough, regional banks, especially those in Europe. And I give some numbers here on the amount of protection and coverage that's being sold and bought.

As far as the gross positions, this is a real interesting slide, in my view. If you look at, say, the single-name credit default swap exposure, again, looking at the gross exposures, what you're really talking about are a lot of the overall trading activities. And certainly when you look at the first row over on the left-hand side, the banks and broker-dealers are very heavy into the trading and market activities, and that's why their gross amounts are so significant. They're certainly the ones adding liquidity to this market and making sure that it happens; whereas, most of the other players are more modest.

Now, if you flip to the next slide, which is the net positions, what was sort of left over after all the trading nets out, you see a very different view in that the banks and broker-dealers were actually, again, on the net purchase side where they were relieving themselves of the risk. So they're number one as far as the amount of notional flowing through their organizations, but at the end of the day, especially the bank side, they're getting rid of it. And who's taking it up but on the insurance and financial guarantor side. So, again, consistent with the broader perspectives I gave earlier, but I think those--looking at the difference between the two slides I find very interesting.

Now, if we isolate just those entities in the surveys that we identified as net sellers--so this is not everybody. It's just those entities. These are summations of how much net sales we're seeing by type of entity. And, again, we see the smallest amount, $94 billion, being the banks, which, again, are mainly the European banks, that are actually taking on some of this risk on a net basis via credit derivatives. The next area would be the traditional insurance and reinsurance companies, and the largest area, probably not surprisingly, are the financial guarantors.

Now, this next slide I'm going to spend a few minutes on because it really talks about the risk appetite for the protection sellers, what risk are they taking on. So when you see the distributions by rating category here, those are really distributions of the underlying credit risk that's being taken on via selling credit derivative protections. And what those represent is that in the case of single-name swaps, what have you, that's the actual rating on the entity, the underlying entity within the swap. Where there's a portfolio product such as the CDO, the rating represents the actual rating on the tranche that's being picked up.

So I think there's a couple things that you see here. One is that on the insurance and reinsurance side, which includes financial guarantors, most of the exposure is AAA. And that may not make sense. Why are people taking AAA risk and passing it along among, you know, other entities? And a lot of it has to do with the fact that the financial guarantors have been somewhat willing to take marked-to-market risk onto their income statements and relieve it for others. So, in other words, via guaranteeing CDOs or synthetic CDOs, which are pools of credit default swaps, they're willing to take some of that earnings volatility on, and they're getting paid for it, even though the risk, the true credit risk is very insignificant.

The other areas are more traditional along the lines of just more traditional risk transfers. And certainly the banks have a broader area because I think they're not providing that ultra-AAA marked-to-market exposure. They're actually looking at taking on different types of risk.

Now, to understand this Exhibit 2, I think you want to understand what types of risk and what some of the strategies behind the different types of entities, and I want to talk about that for a minute.

When you talk about the financial guarantors, they see credit derivatives as a very natural extension of their traditional business, which is to provide wraps or guarantees on debt. And whether they do it in the financial guarantee contract where it's an insurance, or whether they do it through a credit default swap, they see the credit risk as being basically the same. It's credit risk. That's what they do. They guarantee that. And this is the differences in the accounting treatment where the credit default swaps have this, as I mentioned, the marked-to-market treatment.

So it's really not an expansion of credit risk. It's just a different type with different types of accounting.

The insurance companies have a different perspective. In that case, a lot of the risk transfer that we're seeing is through cash CDO investing. And if you're a life insurance company, a lot of what you do is that you take monies from policy holders and you invest them and you try to earn a spread on that. And they're somewhat neutral as to whether they're buying actual bonds in, say, IBM or AIG or other companies or whether they buy those in a pool through a credit CDO. So to them, again, it's an alternate form of cash investing.

You're also seeing for some of the property casualty companies, though, and reinsurers where they're actually wanting to take on credit risk as an alternate to providing property insurance or liability insurance, et cetera, because they see it as adding diversification to their portfolio of insurance.

Now, in many cases they're not doing it necessarily within the insurance or reinsurance company. They're doing it through an affiliate financial products company, et cetera, not unlike what you might see with some of the banks. But, again, they're trying to accept credit risk as a diversification effort and, quite honestly, during the soft market for a lot of traditional products as a growth effort. And that's one area where we're probably having our radar screen the highest, is when we see companies getting into new areas because they can't grow in their traditional areas. Do they really understand what they're getting into? And, honestly, I think a couple of the companies didn't quite understand what they were getting into and have subsequently exited.

This next slide pretty much merges the two that you saw before and just gives a broad overview as to the type of risk being transferred via credit derivatives. And I think the interesting take-away there is that it's relatively good risk. It's generally an average rating of BBB-plus, and where you see the high-yield component, which is about 8 percent, that tends to be at this point more fallen angels, et cetera. But we do expect that over time, as the market matures, we'll probably see more originally originated below investment grade risk, less for financial guarantors who don't like to write below investment grade risk, but maybe more the traditional insurance and reinsurance companies. But, again, we'll have to see how that develops.

As far as credit events, these were the ones that we identified back in last year's survey, and I think the names are probably not surprising. Certainly WorldCom, Enron, Marconi, as the top three, would probably be whatever everyone would come up with if I asked you blindly what do you think.

As far as the '03 survey, a couple of names that are popping up--and, again, not surprising--are companies like Parmalat and British Energy.

The next slide looks at top counterparties, and, again, I don't think this is surprising at all. But basically when we looked at sort of the gross exposures and saw that a lot of the banks and broker-dealers were the ones providing liquidity, well, sure enough, if you look at the top counterparties, it's those entities. And I think the only question that we raise is that about 70 percent of the notional from a counterparty basis seems to flow through these top ten companies, and is that too much? Is it too concentrated? It may, it may not be. Time will have to tell.

If we want to look at the financial strength of the issuers, again, I think that the providers of credit protection are those that are generally pretty highly rated. So you tend to see whether it's both the insurance companies or the European banks, but it's pretty highly rated entity with a big amount in the AAA and AA range, and the rest in the A and some BBB. And, again, I think when you get down into the BBB range, it tends to be some companies that were downgraded.

The one thing that we question, though, is: What is the implications of rating migration? And I think you have seen that some of the original ratings the companies expected to have from counterparties aren't the ratings that they have currently. About a year and a half ago, you probably had four big major AAA-rated global reinsurers. Now you have one left after downgrades. And you've seen other reinsurers, et cetera, see pretty significant downgrades, such as the French reinsurer Score (ph) that not too long ago had AA level ratings and is now down in BBB territory. So that will, I think, play out over time as far as the migration of ratings is to the strength of the counterparties. But, again, so far, strong entities.

So what are the key issues that we think about in sort of quick conclusion? And I think there are three that probably make a lot of sense to people: just sort of management information and risk management, financial disclosure, and hedge funds. And, very briefly, when it comes to risk management, I think our key conclusion is from what we've seen, most companies are doing a reasonably good job. We have taken no negative rating action because of credit derivatives for any of the protection sellers. And I think that says a lot. Certainly we see mixed levels of sophistication, but where we tend to see less sophistication, it seems to be those companies that have relatively modest amounts of exposure. So what we're doing is monitoring to make sure there isn't a lot of growth before the positions are in place to do good monitoring. But at the end of the day, it's a key area that we look at from a ratings perspective.

As far as disclosure, I'd have to say there we're probably not thrilled with what we see in the public areas. That's why we did this survey. We just think if we're an outsider looking in, relying strictly on public information disclosures, you cannot understand what companies are doing as far as shaping and reshaping their risks and what type of risk they really are taking on their balance sheet. So certainly we've been pushing hard for those that have authority to do so to sort of push and require better disclosure in this area so that outsiders, investors, rating agencies, et cetera, can better understand the flows.

And, finally, hedge funds are something that we want to talk about because I think notionally we've heard anecdotal evidence that maybe 20 to 25 percent of activities could be flowing through hedge funds. We're not sure if that's true, but it's what we're hearing talking to a number of companies. We actually asked 50 hedge funds to participate in our survey, and all very politely turned us down, which is honestly what we expected. Hedge funds tend to be very secretive and don't like to talk about what they do. But certainly it highlights the disclosure issue and just a big sort of black box part of this market that exists and likely will exist for some time.

So, in closing, I just want to mention that we do plan to keep updating the survey. If there's market interest, we should have the new '03 results sometime over the next couple of months. We're trying to really use this as part of best analytical practices. We not only publish this information but analysts use it in rating companies because we're getting better information than we think most market industry observers out there.

We like to keep an open dialogue on these issues, and certainly there's some additional information beyond what we publish that we can provide companies on request.

We launched a new website called fitchcdx.com that sort of houses now a lot of information. So if you're interested in credit derivatives without any--no fee involved, so I want to make sure that's clear, you can go and get a lot of good information. And we ultimately will have more and more credit derivative research forthcoming.

And I won't mention it, but I'll end with just a quote up there from Martin Mayer on derivatives that's really what we try to figure out from a ratings perspective: "We don't think this is necessarily happening, but it's what we keep our eyes open for."

So thank you.

[Applause.]

MR. TRAN: Thank you very much to all of our panelists for a very stimulating presentation.

We have about 15 minutes or so for questions and comments from the audience. Would you please activate your audio button on the left- or right-hand side of your seat? QUESTION: I have a question for the two rating agency representatives, although the other gentlemen are most welcome to comment. The question is about the role of the ratings groups in supervision of the insurance sector worldwide.

The background of that is that, as you probably know, there's evidence coming out from research done by Jerry Caprio and Wes Levin that active supervision has no effect on the rate of insurer failure, and that this is an increasingly active model where market supervision will, in fact, become a major part of the new model, and that's contained in pillar three of Basel II and probably in pillar three of Solvency II.

We're actually seeing this occurring in the market already with New Zealand basically getting rid of its supervisor and relying entirely on ratings. Turkey is now putting that into its new insurance law. The rating agencies are really only the mechanism available to look at the reinsurance sector, and the Mexican supervisor has built this formally into their system by specifying minimum required ratings. And that's the model that's being pushed in all developing and transition countries now for reinsurance supervision.

So I think there's going to be an increasing role for the rating agencies in the future supervisory model. The problem is that my own personal experience with the rating agencies is they've been good trailing indicators of insurance company failure rather than leading indicators, and I've got some very close personal experience with that. And the question is whether the rating agencies are changing, and to what extent they feel they are good leading indicators of insurance company failure, or at least distress?

MR. BUCKLEY: Again, I think from the Fitch perspective, maybe to put a few things in perspective, we do not see ourselves as regulators--it's on, okay.

Again, from a Fitch ratings perspective, I just want to clarify that even though we rate insurance companies and can have an influence over them, we don't consider ourselves regulators, do not want to be regulators, and don't want to be characterized as regulators. And it's something that we really can't do. We can rate companies. We can ask for voluntary information. But we have no power or authority over them. We can't subpoena information. We can't demand information. We can't cause them to take actions. They just react to what we do in order to try and achieve ratings.

So I just want to make it very clear that ratings agencies don't see themselves as regulators. We provide opinions, and in doing that we recognize we can have influence over companies. but it's a voluntary process where managements respond to what we look for rather than something we can dictate on them.

That said, I think there is--if you look at what our core mission is, then it's to rate companies prospectively. I think that we've probably had, in my mind, a pretty good track record doing that with a few notable exceptions that probably, you know, get played up more. But I think more and more as a company we have been stressing with analysts that they need to look forward and need to identify situations on a prospective rather than reactionary basis. And a lot of the modeling that we've put in place and a lot of the rating actions that we're taking are more and more looking forward rather than waiting for proof or evidence.

And I think that that's something that's pretty important to talk about, is that at the end of the day ratings are not driven by models, they're not driven by facts. They're ultimately opinions based on a lot of subjective elements. And the hardest part we have is trying to predict aspects of the future.

And it's also an area where we tend to be criticized more because some managements, if we downgrade them on a prediction of the future that hasn't occurred yet, they can say we're being speculative and damaging to what they do. So there's a tension there that we have to manage, but at the end of the day our goal is to make sure that the rating is as prospective as possible, and like I said, if I look at least our track record over the last couple of years, I feel good about it.

MR. TRAN: Grace?

MS. OSBORNE: I would concur really with Keith's observations there as far as the goal of a rating agency and really what our main job is.

I do think we do have--we do exert an enormous amount of influence, and we can't ignore that. In a sense, that as we develop, we find ways of looking at risk, we're going to apply them. We're going to apply them consistently across the entire sector. So that whether companies themselves will have already advanced such initiatives internally or not, just because we are going to apply it consistently so we can have a fair benchmark going across all credit, it does suggest that managements will change some of their behavior in order that they can give us the data so that we can make those evaluations.

That being said, I think that has a bit of a positive connotation than negative. Our big challenge--and I totally concur with your comment earlier, though. Our job is to get it out--our opinion out. It has to be objective. It has to be well informed. But it has to be timely. When we see a change in the credit quality, we have to be proactive in indicating that to the market.

And I do think looking back historically, you could say that we gave management teams more opportunity when they gave us valid-looking business plans that we vetted to see would work, they sold this part of the operation for X, you know, will that--will that improve their capital position? If they redeploy their efforts over here, will that--and I do think perhaps looking back we gave management teams maybe a little too much time to execute. And as issues started to surface, then that influences the ability of them being able to execute on a business plan.

The other thing, though, is the market does not really look for a rating agency to be too volatile in its ratings because it kind of creates uncertain volatility there, whether it's real or it's just a knee-jerk reaction to something that we think is happening, we're not sure, but we're going to make the call early on. And so I do think it is a balance. It's a challenge. But it has to be a well-informed kind of opinion, and we have to make it timely.

So it is a challenge, and I think all rating agencies, you know, acknowledge that our usefulness is to the users. And so if we don't become--if we're not giving them the information they need to make their decisions, then they'll look for other alternatives.

MR. TRAN: David, would you like to add something?

MR. STRACHAN: I'd be delighted to exchange my post bag of letters from policy holders and Members of Parliament with those of Keith or Grace. But just a few points.

I think if you look at the capital adequacy system, certainly in the United Kingdom, and probably also in a number of European countries, companies, in fact, hold much higher than would be required by the regulatory minimum to satisfy the needs to maintain a certain rating. So I've had it said to me by other audiences that actually in terms of capital, rating agencies have provided a stricter discipline than many regulators over the past few years. Now, that will change to some extent, certainly for us, as we bring in new capital adequacy frameworks. But it has been a feature of the past.

The second point is that clearly the rating agencies that do focus predominantly, almost I guess exclusively on financial statements and on capital adequacy and the management of risks, certainly a large part of our role is looking at how insurance companies deal with their policy holders and customers. And, indeed, historically, some of the greatest problems that we've experienced have been in the area of mis-selling and dealing with customers. So that's obviously an area which I don't think it comes within the agency's real ambit.

But, lastly, and rather more positively, I think, we would certainly see pillar three, in the general sense market discipline, as being very important, and anything that we regulators can do to enhance the effectiveness of market discipline is something we should think about. And it's no secret we certainly use inputs--ratings as an input into the supervisory process. It's something we will monitor for our companies.

I wouldn't go quite as far as Grace's slide, which I think said that regulators rely on ratings.

[Laughter.]

MR. STRACHAN: If only that were possible. But we certainly sort of take them fully into account when carrying out supervision.

MR. TRAN: Chuck, as the object of one of these ratings and regulations, do you want to say something?

MR. LUCAS: Yes, a couple of comments.

I absolutely agree with David. Regulatory minimum capital requirements for an entity like AIG mean absolutely nothing. Those are minimum requirements, and presumably they are associated with establishing an investment grade level or something like that. And the level of capital which a company like AIG--AIG's business objective is to retain a AAA rating, which implies a level of capital nearly twice what would be required by some regulatory minimum. So the regulatory minimum has nothing to do with it.

The implication of my comments, I would suggest to you, in terms of the question that was asked is kind of interesting, I suspect. It doesn't really matter whether it is a regulatory capital model or a rating agency capital model. Those capital models are taking on very similar fundamental characteristics analytically. The key aspect of Basel II which is going to show up on the insurance side is pillar two. Let's throw the third pillar on the table just for discussion.

Now, I think the problem for the rating agencies is how to create the equivalent of a pillar two level of assessment and this goes right straight to the question of the complexity of the liabilities in insurance businesses and the nature of the analytic structures that have to be applied to them. That implies a commensurate beefing up and increasing complexity of the capital models that are used to ascertain capital adequacy. And the spirit of pillar two of Basel II is that an entity should use its own models in making those assessments and that the supervisor in a non-quantitative way assesses the appropriateness of those models to the business and then the risk management of the entity.

The part that is going to be most difficult for the rating agencies is to create the equivalent of that with respect to their capital models, where access to information is only a part of the problem. It's a question of the resources that are required to actually collect all that information and digest it properly where the range of the models that's in use is proliferating like crazy. It's tough enough for the regulators who actually have more resources to throw at this problem than the rating agencies do. It's tough enough for them to do this. I think it's going to be that much more difficult for the rating agencies.

Bottom line, the overall problem is getting more difficult fast for rating agencies and regulators, and the solution to the problem is not altogether obvious right now.

MR. TRAN: Any other questions?

QUESTION: I was specifically going to ask you, can you be a little more definite about who--what regulator is going to be at AIG evaluating your risk management models under what U.S.--well, what regulator, I guess, anywhere is actually going to be in there and who it is who's qualified to come in, I assume at a consolidated level, to look at your risk management system?

MR. LUCAS: I don't have a clue.

[Laughter.]

MR. LUCAS: I'm not trying to be facetious. You asked a bunch of different questions. Who is competent to do it, and I think obviously the implication of what I'm saying is that it would require a pretty sophisticated regulator to deal with an entity like AIG, given its complexity. There's no doubt about that.

But it's not clear what the consolidated--who the consolidated supervisor is going to be. I mean, that's something that is going to get worked out over the course of the next couple years. But there's no decision that's been made. There's discussions with the regulators about the subject, and it's not decided. It's not clear. It's not clear who is appropriate. I mean, there isn't any regulator that exists now that sort of has a portfolio of expertise that fits very well. That's evident. it's probably a banking supervisor who sort of in the abstract would come closest to that.

But at the moment, as far as I know, that's not under discussion. I don't know what the answer is going to be. I think it's a very interesting problem, and it's going to be very, very difficult for--I mean, just in the conversations I've been involved in with various regulators who might be candidates for that role, you're really, you know, starting from very basic fundamentals for what the business is and how it's run. And it's really a much earlier stage of discussion than it is with the rating agencies. We're just exploring--beginning to explore the issue.

QUESTION: I would like to raise the issue of globalization and the industry we're discussing now, which is the insurance industry. International financial markets have started much, much earlier than the current trend towards globalization of economics. The globalization of economics has had stumbling blocks dealing with culture differences. What is the experience or what do you project to be the experience when the insurance industry tries to penetrate globalized markets more than it has? Will there be problems with the culture differences that exist?

MR. TRAN: Chuck, why don't you take the--

MR. LUCAS: Well, you know, I guess AIG is the world leading insurer that fits into the category that you're talking about. The company was founded in China in 1919. Its largest retail penetration is in Asian countries, but we operate in something close to 150 countries.

Globalization of financial markets presents some very interesting sorts of issues in the context of the regulatory framework that exists in the insurance industry. Most of those countries of the 150 countries have domestic insurance regulation that severely limits the capacity of the insurance company domiciled in that country. It doesn't matter whether it's a subsidiary or a branch of an international company or not. It's the operation within that country that sells insurance products in that country is limited in the currency denomination of its assets. So it basically is required to buy assets, to invest within the country that issues its liabilities.

It doesn't have access--we don't have access to the international capital markets with respect to the assets of a great many of those insurance companies domiciled in those countries. And that's a tremendous handicap to the insurance company in terms of its capacity to diversify its asset risk.

I mean, obviously, if you diversify your asset risk, it involves currency risk, or it requires the existence of, you know, developed derivatives markets to hedge the currency risk back into local currency and achieve asset diversification.

The existence of globalized financial markets creates market opportunities to do that kind of asset diversification, but the regulatory environment largely prohibits it.

MR. TRAN: Thank you.

We have time for one last question. If not--okay. One, please.

QUESTION: The U.S. Export-Import Bank provides financial support to U.S. companies to help facilitate trade of U.S. goods overseas, and our mandate is to support the sale of U.S. goods into countries that oftentimes the commercial market will not function in.

One example is the area that I work on is in project finance where Ex-Im Bank provides financing to infrastructure projects such as a power plant. And an issue that we deal with is when a project company has to get insurance to insure property damage, terrorism, and all of that, oftentimes we're finding that these insurance companies are reinsuring the risk with a local entity.

So in the case where a multinational insurance company is reinsuring the risk with a local entity, how do you see that in terms of is this a complete pass-through of risk, or do things change when the risk is transferred to a local entity? And how would it impact the financial stability of the international markets?

MS. OSBORNE: From a capital perspective, the way we address it is we go back to look to see if we have made a credit assessment of the reinsurance entity. To the extent that we have, then we will apply a credit factor, in a sense a reinsurance recoverable charge, that is reflective of that risk.

So to the extent that perhaps in the case that you would be talking about, really may be an unrated entity, it's very large, it's a very prohibitive charge. So the insurer does not really get the full benefit in our capital model of actually passing the risk on to the reinsurer.

MR. LUCAS: I think that the problem is basically a currency hedging problem. If you reinsure offshore, in the first place it's going to be difficult to find a reinsurer that knows the country, typically. But if you then reinsure offshore, you're either going to have a currency exposure or you're going to have very expensive reinsurance because the reinsurer faces the currency exposure inherently. One or the other. So it's going to tend to be cheaper to buy the reinsurance locally, and it creates the problem that Grace is talking about. It's a credit risk problem. So you're really transferring, you know, a market risk problem into a credit risk problem.

MR. TRAN: Well, we've run out of time, so we have to stop here. Thank you very much for joining us in this forum, and would you please join me in a round of applause for our panelists.

[Applause.]

MR. TRAN: Thank you very much.

- - -