Peter Casey, director of Lloyd's and London market supervision, HM Treasury; Andrew Davey, manager risk financing, Royal & Sun Alliance; Patrick Devine, Allen & Overy; Philip Hooley, Clyde & Co; Andrew Pincott, Elborne Mitchell; Ambereen Salamat, Davies Arnold Cooper and John Young, chairman of the Lloyd's regulatory board, joined Global Reinsurance co-editor, Lee Coppack, for lunch at the Baltic Exchange in the City of London.

Lee Coppack: It always strikes me how little was said about the London market company insolvencies compared to the problems at Lloyd's. Are there things which have come out of those liquidations and schemes of arrangement that tell us things for the future in terms of avoiding that sort of insolvency recurring?

Patrick Devine: The main emphasis is on trying to secure solvent schemes of arrangement, largely because you literally save them before they go down. We have seen one or two come down the pipeline which is very good news. Anything that stops any insurer, whether it is at Lloyd's or in the company market, going down has got to be good news for certainty and it has got to be good news for name and reputation.

Unfortunately, we do not actually have a kind of full Monty kit of legislative environment around it at the present time, more of a cut and paste job. If you are insolvent, you can go to the Policyholders Protection Board and say: "Please stump up for the third party liability losses", which they have a statutory duty to do. If you try to save yourself from going down, you cannot get a contribution from them, which could be quite critical as to whether you are not going to have a solvent scheme of arrangement. They are confined as to what they can do in the context of solvent arrangements, which is a pity. But by the same token, they cannot give out money willy-nilly. If the solvent scheme of arrangement does not work, they would be on call in any event. That is the main emphasis that I see as a very positive thing to emerge from a very unfortunate set of circumstances.

Ambereen Salamat: There is the advantage that the parent company has in a solvent scheme in the sense that one of the main attractions is the commutation of liability in that early realisation. That brings the cut-off date a lot earlier than letting a company be in run-off for years on end, and it is a major advantage to a lot of parent companies assessing their liabilities at that early stage.

Peter Casey: Because the regulatory reform legislation will go on a repeal and replace basis, we shall in any event have to replace the insolvency provisions that appear in the Insurance Companies Act, so I think there is an opportunity to think again about what we would like. I have to say that has probably taken second, third or fourth place in thinking to some of the main regulatory provisions, simply because those have been so difficult to get right across the whole spectrum of financial services. But people are certainly turning their brains to that.

John Young: If you look back to the history of the last 15 or 20 years and look worldwide, how many cases are there of serious insolvencies among insurance companies which have led to pain and grief at the end of the process? I can remember the odd building society, the odd bank, the odd investment bank, the odd securities house.

Philip Hooley: I think there have been a number and if anything over the period, there is probably an increasing trend. As Patrick said, obviously what one would want is prevention rather than cure. If there is a way of diverting the train which is going to go off the track, one should seize that opportunity. With the risks which are becoming ever greater, as one gets globalisation, it is important to have the tools available to achieve that.

Andrew Pincott: But you cannot have a capitalist system without bankruptcy, can you? It is certainly not a realisable objective; it may not even be theoretically a proper objective to avoid all bankruptcy.

Philip Hooley: It is important that one does not have a complete failsafe because there has to be that element of risk for the participants that they cannot feel that there is inevitably a safety net whatever they do, otherwise someone might go out and take enormous risks and lose vast sums of money knowing someone will always try and pick them up.

Peter Casey: This is essentially what happened in the US savings and loan.

John Young: Savings and loan is probably the biggest international example of the last 20 years. But the concept of intensive care, the regulator really having to watch a particular insurance company for months and maybe years, to nurse it out of the horror scenario, that kind of work presumably goes on a bit.

Andrew Pincott: It went on rather more than it does now. The DTI was really rather famous for putting pressure on overseas owners to recapitalise insurance companies - I can think of one South African where there was a certain amount of pressure put on the shareholders in SA. I am not sure it worked - but there was a certain amount of lobbying.

Peter Casey: It is still known.

Patrick Devine: For the proper conduct of a market, there has got to be a distinction between free fall and controlled fall in terms of where regulators, accountants and lawyers can properly step in and try and minimise a loss. When insurers go down, they go down in quite a heap, very quickly. They have contractual relationships with reinsurers, etc. In a market, even a global one, you have a domino effect, which is why I look with favour on early intervention and sensible intervention to try and minimise the losses. Ceasing underwriting or ceasing underwriting in part, restrictions on investments etc, so as to assist in trading up is actually only what a prudent management would do, but sometimes prudent managements get themselves into difficulty. So I do not wholly favour a weak goes to the wall type principle if you are trying to regulate a very, very complex market.

John Young: How important do you think it is that globally there should be a sort of mutual whistle blowing obligation on regulators to warn each other of early warning signs, problems? A lot of regulation is about stopping things, a lot about stopping things is about early warning and a lot about early warning is about whistle blowing. That means regulators, that means professional watchers. It means guys in businesses almost. Suppose there was, hypothetical example, an emerging problem, very early stage, with an Asian reinsurer taking on obligations they could or did not actually really intend to honour at prices which were commercially silly. Who would blow the whistle? How do you follow things up?

Philip Hooley: One of the things you do need in that system is co-operation between the regulators. Many of these institutions have their offices in other jurisdictions. So, first of all, the regulators need to be able to co-ordinate. Then in terms of whistle blowing, there may be a number of professionals who have relevant information. One of their problems is confidentiality, has been confidentiality anyway. To some extent, the accountants have been helped there in that they now have more flexibility to be able to go to a regulator and say: "We are extremely worried about what is going on here", than they had in the past. Maybe that is something the regulators need to look at. If they want co-operation, they will probably need to facilitate it.

Patrick Devine: Traditionally every individual country has regulated what has gone on within its borders, like a Victorian nationalistic model which no longer works. In the EU, we are now under co-ordinated legislation, co-ordination on a formal basis under a protocol between the supervisors in the various EU member states. To what extent this protocol is used for whistle blowing and for information purposes, etc is another matter but it is there as a kind of regional block.

In the United States the National Association of Insurance Commissioners (NAIC) tries to co-operate to some extent. What is missing in the context that you are raising is a kind of - it would be quite a burden on a clearing house - global, international organisation. Everybody stays in their own country but just has formal or informal lines of communication with each other.

Peter Casey: I think the International Association of Insurance Supervisors (IAIS) is beginning to move in that direction. It is a rather young organisation at the moment but I think it has that potential.

Patrick Devine: At the moment, the regulators very much rely upon the professionals. Look at the life sector as an example. You have got the appointed actuary and you may wish to try as an aggressive finance director to do x, y and z, whether it is cutting premiums for market share or whatever. At the end of the day, you have got to get it past the appointed actuary whose great god is going to be the Treasury in terms of justifying it to the Government Actuary's Department as to whether or not it is a prudent thing to do. Internally the controls are there, but whether or not they are set at international level, I do not think that is the case.

John Young: I was thinking that internationally early warning signals about global business have to be derived from a global early warning system if they are going to work.

Andrew Pincott: Take your hypothetical example of an Asian reinsurer, writing a huge volume of business. It is not entirely clear to me on what basis that would be the subject of a whistle blowing exercise, really, other than to keep it under monitoring. It would not justify an intervention, would it?

John Young: There is a distinction to be made between early warnings of potentially unhealthy, worrying international trends in behaviour on the one hand, and what I came in on, early warnings about insolvencies of particular companies. They are not quite the same subject.

Andrew Davey: It will give no comfort for the intervening period, but I have a very strong suspicion that the capital structure of the insurance industry will eventually become rather more similar to the capital structure of the mortgage market than the traditional share capital system that we use at the moment. The more you look at it, the more you can draw parallels between the development of a secondary market for mortgages and opportunities for the insurance industry to go in that direction.

If that is the case and if the securities that are put up to support that sort of environment are 100% collateralised cash, as they tend to be at the moment, then the whole default question tends to diminish into the background. That is not to say it goes away entirely, but it diminishes very much. You are no longer arguing about the relative merits of one insurer who might be A- or might be A+ or AA. We are arguing about an instrument that is indisputably AA or AA+. The whole order of magnitude changes.

Patrick Devine: It may be a partial solution but even in a securitised situation, first you are securitising part of your book of business. You will not necessarily have the quality of business right across the board that you would be able to securitise. You would pick your best book and your best cash flow stream. Even following a securitisation, there are still risks there: business risk, investment risk, credit risk, traditional risk, etc, any of which can turn nasty and bite even the best of insurers and reinsurers.

Lee Coppack: It seemed to me one of the problems with some of the cat bonds is that in order to get investors to be prepared to take the risk, they have had to pay away quite a bit of the money to secure the principal. The advantage the reinsurers still feel they have is leverage.

Andrew Davey: I think that is true but if we are addressing the narrow question of the risk of default, then it seems to me that a securitised insurance industry is going to present less problems to regulators for people who find themselves unable to collect claims as a result of the default of the reinsurer than would a traditional insurance industry. Not to mention enormously simpler administration of the process in the real world as well. I have argued elsewhere that the insurance industry has too much capital to be efficient and too little to be safe. They are mutually exclusive objectives, it seems to me. If you hit some middle ground, which is both half way efficient and half way safe, that probably says volumes about the skill that has been in the insurance industry.

John Young: I think one of the most fascinating questions is how far into the traditional insurance market, whether the commoditised retail end or the bespoke wholesale end, will alternative risk transfer techniques or products go. It is really big business and there are some really clever people doing it. I just wonder if you took a five year view how far they would actually get into replacing ordinary insurance cover in the sense we understand it.

Andrew Davey: I have argued, I do not know how convincingly but certainly lengthily, that capital market approaches are going to be very much tools for professional insurers and reinsurers to use and the use of those techniques is going to generate benefits that those insurers and reinsurers can in turn pass on to the retail customer. I would not say that they are not going to be used by the ultimate retail buyer but they are going to be more a creature, in my opinion, of the reinsurance and insurance business. That is not to say that a major corporate, a BP, a major pharmaceutical company, whatever, is not going to find itself big enough that it has the critical mass to do something in that area itself.

John Young: The clever City value add, if you can call it City, is that you have got people who are jolly good at broking schemes, whether to insurance companies or to Smith Kline Beecham as it were, taking part of the complicated risk profile on their own book and then hedging themselves. Arguably this is the way of the future.

Philip Hooley: I think that is right. The entree is going to be through the major insurers and reinsurers but the companies are looking at it. I think Alcatel is one. Perhaps in particular areas where they find it difficult to place a particular risk like political risk, they will say: "Well, we deal in so many different countries, there just isn't the capacity in some of the areas to do it. It is an extremely labour intensive exercise for us to do. We would much rather find a different solution."

The irony here is that to some extent this is a product being promoted by, if you like, Wall Street, but in order to really sell it and to make it work effectively, they actually need the skill base that is found in reinsurance, because they need someone to be able to correctly value these risks and price them.

Andrew Pincott: Doesn't it depend on which particular technique you are talking about?

Philip Hooley: It varies, I agree.

Andrew Pincott: In broad terms I was thinking of catastrophe linked debt, derivative options on the CBOT, probably the newest of which is the catastrophe equity puts. I am sure there are a few other tricks but it is probably a mistake to lump them altogether, because each of those techniques has different advantages and disadvantages.

Philip Hooley: And requires different disciplines in varying degrees. Some of them appeal to very different markets. Some of them are looking at, as you say, catastrophe risks and some of them, if they develop, will look at much more mundane risks, a book of business which is motor or life or whatever, to see what can be done with respect to that in terms of disseminating the risk.

Patrick Devine: A lot of the work of the past few years has been experimental in many respects. We still do not have a mainstream, continual flow of these products. They are still being developed. These products you are looking at will actually be replacements in a way for traditional reinsurance. I feel more confident with the idea that they will work alongside them. Everything at the end of the day in this market comes down to price. We are in a soft reinsurance market at the moment, and to create a bespoke product, a capital markets product, does cost a lot of money.

Andrew Davey: The R&D cost does tend to get amortised.

Patrick Devine: Over a period of time you will get round to a kind of plain vanilla motor book. Take one of these down off the shelf.

John Young: Reinsurance at the moment is quite cheap, isn't it? So maybe that introduces a bit of a slow down. The real derivatives experts are not short of business, so it may take some time to evolve but I am sure it eventually will.

Patrick Devine: I agree with you. I think it is going to have really quite a considerable impact, in terms of the wooden wheel to the pneumatic tyre, but I do not think that they are going to be replacements, necessarily in the short to medium term. They are going to live alongside each other on a complementary basis. It is just another option that you can exercise.

Andrew Pincott: One of the drags on the development of some of these techniques is, of course, the regulatory side of these things. I am sure they would actually take off at a greater rate of knots if, in fact, reinsurance did not benefit, as it were, from the regulator's favour. I am not saying the regulator should not favour reinsurance; there are jolly good reasons why he should. But, for example, if instead of protecting your book of business by conventional excess of loss reinsurance, you wanted to protect it by a programme of CBOT CAT options, you would suffer a significant disadvantage, because when considering your solvency margin, you are entitled to reduce the solvency margin by the reinsurance reduction factor, which you could not do if you made essentially the same recovery by hedging your exposures on the CBOT.

Andrew Davey: And with arguably better security in the process.

Andrew Pincott: There are considerable advantages. First of all, you get paid more speedily. It may be cheaper, certainly when catastrophe reinsurance rates go up and the transaction costs are probably cheaper. There is no scope for the sorts of complicated reinsurance disputes about the meanings of wordings and what the Court of Appeal recently described as the "kebab principle" to creating reinsurance wordings. The disadvantages at the moment are two. One is regulatory. It is not recognised for solvency purposes. And the other is inherent in the system that you have actually got to stand the basis risk, which is that your book of business isn't quite the same as the book of business which forms the index.

Andrew Davey: If it was only basis risk we were talking about, I believe that insurers will find that that's their role. I think insurers will become traders of basis risk, quite deliberately. From a regulatory point of view, outside this country at least, there are some jurisdictions where people are starting to embrace this idea. In a way regulators are saying to themselves: "Clearly, this is going to happen. It is to the benefit of the policyholders whom I as a regulator exist to protect. I should be in a position to have some hand in controlling what is going on. Therefore, I will permit it to happen, it being, let us say, the use of derivatives in lieu of reinsurance. And I will create a regulatory environment where this becomes possible, and I will regulate it rather than maintaining a position where it is not possible in the first place."

John Young: We do not start from a complete empty space. In the securities market, stock in trade is stock. People have for some long while now been developing ways of hedging their risk in that stock in trade. Insurers' stock in trade is insurance, so the same process might begin to develop. There is a lot of nous in both the banking and securities field about how you tailor capital requirements to effective hedging models. It has become an art and a science to keep up with it. Should people be allowed to devise their own private, in-house models how best to hedge risk, keep it to themselves and do jolly well, in which case does the regulator tick the model? Or does the regulator prescribe certain sorts of haircuts that you ought to have if you have got certain kinds of exposures, regardless of how you are capable of hedging them or not? Now this kind of work may well come into the insurance field. The Treasury is as aware as anybody of how this whole business has developed in the securities and banking field. There is a lot of expertise there.

Peter Casey: We are actually very conscious that when we get into the Financial Services Authority (FSA) we will pick up some of this expertise and some of this thinking. One thing that slightly depressed me was the talk of all these things as though it was all a matter of whether the minimum solvency margin was triggered, because already I would be pretty twitchy about most companies that were close to the minimum solvency margin. Already one is trying to get an appreciation of the actual risks that a company is running, the risks that this particular reinsurance portfolio poses and arguably we have been much too soft on reinsurance portfolios in the past. We have still got some way to go, not least because we have vastly less resource than the banking supervisors. But the trend is going to be much more towards careful analysis of the risks and, indeed, towards risk based supervision, which means something slightly different, rather than a mechanical approach, which does something only when you get very close to the solvency margin. My colleagues are talking an awful lot more about sound and prudent and risk to policyholders, ie the general trigger powers for intervention, than about solvency margins these days.

Philip Hooley: The other issue that needs to be worked out is in the future if these products are developed is that as disputes arise what laws are going to govern these products. The laws that the banking industry are used to using whereby they have a close analysis of the complex document they have drafted before they entered into the agreement on this product. That is quite different from the complex laws that the insurers use when they do a deal on the risk. In the latter case the documentation is often quite flimsy and primarily the obligations are placed on one of the parties who does the investigation and presents the risk, and the other party then decides how he is going to value that.

John Young: Sometimes they do not understand the terms they are using.

Philip Hooley: Absolutely. The players in each of those fields are used to dealing with their own laws, rules and regulations. They are not necessarily very used to dealing with the other's. There may have to be a blend.

Andrew Pincott: But Philip, looking at CBOT catastrophe options one of the great advantages is that they are commoditised. There is very little scope for dispute. When it comes to settlement date, there is a payment. It has all been margined, because it is on a recognised exchange. You get your payment the following day, unlike the potential for disputes with reinsurances about what is one event.

Patrick Devine: If people do not get paid, it is in event of default. It is a far simpler and more brutal mechanism.

Philip Hooley: But it is going to be looked at by players who are not used to those types of documents, and they are also dealing with a product which is, perhaps, lying part way between the traditional capital market product and what they see as their own insurance product. I agree once you have got it on to the markets and you are actually dealing with it - just as soybeans or anything else - then you have removed that because you do not have the documentation, but, of course, the examples we have up to date have partly been special risks that have been written and ironically, so far have been primarily written by reinsurers.

Patrick Devine: It is a wholesale market you are dealing with, remember. If the question is: "Whether disputes on capital market products are going to lead to litigation or arbitration among insurance and reinsurance litigators," the answer is no, because they are capital markets products. The traditional insurance and reinsurance litigation legal market will only get involved if it is an insurance product. If it is an insurance product in the first place, it is not a capital markets product.

We are talking too generally in a way, because as we have already said there is a whole range of these products. The big trick - trick is the wrong word - the big feature of putting these products together is to make damn sure that when you draft them they do not have the characteristics, the legal characteristics, of an insurance contract.

Philip Hooley: Yes.

Patrick Devine: Because it then all just falls apart. The issuing bank or the issuer etc, etc., if it were an insurance contract, would have to be authorised in every country in which the thing was marketed, and you run into all kinds of massive regulatory problems. Where there may be disputes is as to whether or not somebody has actually got the documentation right.

Philip Hooley: Exactly.

Patrick Devine: It is very difficult to know as a matter of English law what is or is not an insurance contract. There are very solid parameters, but this is an area of law that will be pushed at the margins by these products. Even in the regulatory side, we have got banking exemptions from the Insurance Companies Act, because in some respects it is still not wholly clear whether this is a traditional bank product or an insurance product.

Peter Casey: What I find slightly depressing about this discussion is that it is in terms of what the formulation is, therefore, which regime you can manage to bring it under. My interest is, what are the risks? And eventually, what are the risks that the policyholder does not get his or her money? Now, what I have not heard is - save in the context of CBOT options - a discussion of whether the risks of ultimate non-payment are greater or less in any of these particular techniques than they are with traditional reinsurance, which is the paradigm that we are all used to dealing with.

Andrew Pincott: Peter, you do not know unless it is tested by the same sorts of catastrophes and things which have tested reinsurance for the past few centuries. These are modern products. There was, as I understand it, a pilot scheme by the insurance commissioners of California, Illinois and New York where for three years they monitored a programme of CBOT catastrophe options to replicate excess of loss reinsurance. On the basis of that programme, they have given the green light for people using CBOT catastrophe options in place of excess of loss reinsurance.

Peter Casey: At least if you are inviting someone to go into this field and allow it, you are inviting them to take some view - and a view which I agree may be proved wrong in either direction by events - on the risks posed by that technique. That is where I would want the discussion to focus, and I would want to be saying: "Okay, if such and such an insurance company has got a Cat-e-Put, what kind of allowance ought I to make for that in looking at whether that company is soundly and prudently managed with a proper regard to the interests of its policyholders?"

Patrick Devine: Are you focusing then on the rating of the issuer, because we have unregulated reinsurance companies, outside of the UK, with which UK direct insurers are transacting reinsurance business. They are passing on, with the full knowledge and approval of the Treasury, risk to these reinsurers who are not regulated, certainly do not have a Standard and Poor's rating, etc. Trade that off against a transaction between a direct insurer and an AAA rated internationally and globally regulated merchant bank.

Peter Casey: You are persuading me that there can be substantial counter-party risk in the reinsurance transactions. I was actually already persuaded of this.

Patrick Devine: I'm sure.

Peter Casey: Perhaps I ought to take a tougher line on those and that is a reputable argument because there are, in certain places, significant counter-party risks. One of the things that I would expect us to be doing in a future regime is looking very carefully at the reinsurance programmes of companies. So, yes, I know that those risks are there. There are one or two countries whose names coupled with the word reinsurance terrify me. Now, ought I to be similarly terrified about any of the other instruments and if so, which ought I to be looking at in evaluating, as I keep saying, whether the insurance company is soundly and prudently managed?

John Young: Okay, I am a company and I know that I am about to receive $1 billion over the next five years in lumps and I want to hedge against the value of that flow adding up in sterling. You can do a deal.

If someone is running an insurance company and gets offered a selection of - I will call them products - devices, to be able to hedge what he thinks his risk is. Two questions. One, what actually is the risk on that insurance contract? Two, how well is the hedging going to fit to that risk? It is easy for the insurance company to say: "Make it for £100 million." So, you have only got the second question: When he is guessing what his real exposure to risk is in the first place, that is his job, and he is then guessing whether the hedge product is going to fit to produce the right amount of money.

There are two difficult running-a-business questions for him, aren't there? He is used to the first one, but the second one is, I think where you are getting to, Peter, isn't it? He could completely under-guess the efficiency of the bottom line of his hedging programme. Does that mean that the regulator has to replicate his expertise in assessing whether a product is likely to produce the goods or is there any easier way of regulating?

Patrick Devine: It happens before it gets to the regulator, because this is done on a daily basis. The lawyer says: "I read it this way and it should produce x. It is for the actuary to come and say: "Yes, indeed. It does produce x because I have crunched numbers." This is actually happening on a daily basis much before it reaches the likes of the Treasury. This is a great growth industry for the actuarial profession in terms of coming in and advising a company. Indeed, it is quite interesting to see how the major merchant banks and investment houses who are really pushing into this area have recruited a considerable number of actuaries to do this exact type of risk assessment that you're talking about and they use them for the production of these products.

Andrew Davey: It is certainly true that when you go down the derivative route, you swap a certain amount of credit risk for a certain amount of basis risk, to pick up the point that was just made, but equally you can model the basis risk. There is no absolute certainty in the model but it is probably fair to say that you can model basis risk more precisely than you can model credit risk. At the end of the day, it is a trade worth making. On top of that, nobody is necessarily assuming that one would get a 100% solvency credit for a derivative anyway. As we all know, we do not get 100% for a reinsurance.

John Young: Especially as the derivatives that might be invented cannot necessarily rely on open and understood indices, markets or whatever.

Peter Casey: To a certain extent it comes down to this question: if you were writing the FSA rule book for the section dealing with insurance, what should you be wanting to say about both conventional and less conventional forms?

Andrew Pincott: At this juncture, you do not have to say anything, do you? Because you favour reinsurance. In so far as the hedging or derivative instruments are not just simply protecting investment risk, you do not have to worry about them because you do not really contemplate them. They are not in your view. If they are used at all, they have to be accounted for as a liability.

Peter Casey: Are you saying that if you were Howard Davies, you would be urging me to draft a rule book that said nothing about these instruments?

Andrew Pincott: From the point of view of insurance company solvency, I do not think you have to make any change. The problem is that by not making a change, you are inhibiting developments that are being seen in the market.

Peter Casey: Are you saying that I should make a change or that I should not?

Andrew Davey: I would argue that you should.

Peter Casey: Fine, and what should that change be?

John Young: Here is a small set of simple principles to guide insurance company management in a sound and prudent approach to the protection of their business and their balance sheet by bespoke-design derivative products.

Patrick Devine: We already have that.

Andrew Pincott: Yes, there is prudential advice from the DTI about investment protection derivatives.

John Young: You could add that, furthermore, we will be talking closely to any insurance company who devises weird and wonderful paraphenalias of schemes and departments and boffins to elaborate its chance of maximising its returns from the derivative products themselves.

Lee Coppack: The thing that strikes me is that, at least so far, the people who are getting involved in buying and, in the case of the CBOT writing, these products are reinsurers.

Patrick Devine: I think that is a slightly different thing in the sense that they understand what they are buying. They understand the risk and they have a portfolio which they know how to manage. It is like the USAA bond in the US which was largely certainly bought by a good chunk of the reinsurance market who understood exactly what it was they were taking on board.

Peter Casey: You are quite right in saying that one would need to look at correlated risk between what has traditionally been in the asset and what has traditionally been the liability side of the balance sheet in any insurance company that starts to take these on.

Patrick Devine: Can I give my answer to your question about the hypothetical question put to you by Howard Davies?

Peter Casey: That rule book actually has to be written.

Patrick Devine: Quite. I think it is critical for the City of London, not only for the insurance industry but also because the banking and investment community are really very focused on these products for the insurance sector. We are talking about UK plc here and I would not like to see it slipping away specifically to Geneva or New York or whatever.

Andrew Davey: As it will. As it is.

Patrick Devine: I would like to see neutrality. As a direct insurer in the UK, there's no difference to your mind buying reinsurance and a capital markets product. That it becomes a commercial decision on the part of the finance director and the board advised by their actuaries or whomsoever as to which of these two products best fits the requirements of the direct insurer and their policyholders. At the moment, we have a constraint against reliance and use of these products by direct insurance and, indeed, UK reinsurers. It will ultimately come down to this, if the capital markets products are worked on and perform in certain areas better than traditional reinsurance and traditional retrocession, then all we are doing is handicapping the UK insurance industry in not allowing them to participate in these products.

It is interesting that it is only since we had the third life and non-life directives that insurance companies have been able to purchase derivatives to hedge against exchange rate risk. There are very good reasons for that, but there are no rules in the UK that stopped the local sweet shop owner from buying derivative products. There may well be a set of circumstances where they say, like Winterthur: "Let us issue a bond on hail damage to motor vehicles, 4,000 vehicles and the bond will trigger." That was an experiment that was done by Winterthur; they did not actually transfer much in the way of risk. They did that to see how these things worked. What I would like to know and the question I would put back to Howard Davies is: "Why is it you would wish to discourage the use of these products by this leading industry that we have here in the City of London and the rest of the UK?"

Andrew Pincott: Doesn't that come back to what his brief is - which is prudential supervision of insurance companies? At the moment, you cannot plunge in feet first about these because they are very much experimental. People do not have a history, as they have with reinsurance, about how these things will perform in catastrophic situations.

Andrew Davey: I do not believe that is true. The insurance industry does not have that history but the history and the law surrounding derivatives is every bit as deep as the law surrounding insurance. It is merely a question of crossing the line.

Andrew Pincott: I was not concerned with that. I was concerned with the economic effects of hedging your exposures on the capital markets with, for example, a CBOT catastrophe option versus a tailor-made excess of loss reinsurance. At the moment, we do not have very much evidence as to what the effect of basis risk is.

Philip Hooley: Does that suggest, as we were discussing earlier on, that really if there were to be a rule book in this area, you would want to try to keep it to principles rather than a detailed exposition? Speaking as a litigator, one of the interests I have in relation to the new FSA rule book would be whether it is to be construed literally, which is the traditional English approach to litigation, or whether it is to be given a purposive approach which is very much a European approach to litigation. The FSA could help themselves considerably there if they dictate which they would prefer. That is not to say that that would not necessarily bind the court, but it would certainly help.

Lee Coppack: Ambereen, what is your view?

Ambereen Salamat: Having heard what everybody is saying, I think there has to be the flexibility built into any rule book which will allow for hedging of risk the way we have discussed, but I feel it is an area which is so undeveloped that it does have to be done quite carefully, because from the regulator's point of view, the obligation is to protect the policyholder. I do think it has to be quite a cautious approach and slowly introduced. It is difficult to sit here and say: "These should be the guidelines."

John Young: Are you saying that regulators must restrain themselves or that businesses must be restrained from going too fast?

Ambereen Salamat: No, I think it is difficult to legislate but I do think the flexibility has to be there in order to allow the introduction of these sort of products.

Andrew Davey: At the end of the day, these products are going to come into existence. They are, as we speak, in existence. The question is whether a regulator in a particular jurisdiction has the ability to control what is happening or not.

Andrew Pincott: The point is, they are done as surrogates at the moment because most of them are fronted by reinsurances. Risk is being hedged on the capital markets and then there is quite a lively industry in people converting these into fronting reinsurances that will pay out the indemnity under the reinsurance contract that is actually the mirror image of what will be recovered on the hedging programme.

Patrick Devine: You front through reinsurance companies purely to avoid Treasury constraints/regulations. The issue is for an insurance regulator, you cannot just say: "My interest stops at Dover." You have got a situation where we are allowing UK direct insurers to do deals with unregulated reinsurers and reinsurers who do not have any rating. That is number one and that is the situation we have lived with for a long time. Two, where we have got products which are available the board of directors of various UK insurers and others believe are better for specific purposes for their own direct insurance company.

I think you cannot just say: "It is happening in Bermuda. It is beyond us." If that fronting reinsurance company goes down and cannot pay or if the bonds that are behind it do not pay, it is the man on the Clapham omnibus here in the UK who will be picking up the phone and talking to his MP. I am not suggesting that you do but you cannot say: "UK insurance regulation stops at Dover", because there is a whole series of these transactions and chains. That goes back to the initial discussion as to what extent should there be closer co-operation between the regulators.

Peter Casey: What I am hearing is essentially nothing between at the moment: "Stop the ball. That is terrible. You must change that," and "At the moment let the board of directors do what they like." So take me the next stage along.

Andrew Pincott: Isn't the exemplar case to follow what the insurance commissioners in California, Illinois and New York have done? Namely to have a controlled programme to monitor the effect of, in that case CBOT catastrophe options, to see how they do differ from more conventional excess of loss reinsurance with which the DTI and now Treasury have been familiar. That it is actually an experiment. I am not sure that I recollect that sort of experimentation by the DTI.

John Young: I am surprised the conversation has been so far, unless I misunderstood it, restricted to how insurance companies should organise their businesses using these things and how the regulator should react. What about the separate question - I think it is separate - what if Goldman Sachs finds a way of going straight to Eurotunnel or BP or Marks & Spencer offering a really highly sophisticated way of insuring against commercial risk which either could not be covered by the traditional insurance market or has been attempted but nothing has been found that does the trick?

Patrick Devine: It is the next generation.

Andrew Pincott: I think it is happening. It is certainly happening in the US.

Andrew Davey: It is happening in this country. There is a lot of proprietary information here but there are certainly transactions, at least in contemplation if not actually executed, that do precisely that.

John Young: So Eurotunnel could go along to Goldman Sachs and say: "I want protection against the chances of fire, protection against the chance of this or that movement in the European economies and protection against the number of apples coming in from Italy?"

Andrew Pincott: Take a common example. If you want to transfer the political risks associated with export, you have essentially got three alternatives at the moment: state maintained export and credit guarantee departments, political risks insurance in the conventional insurance market and credit default options on the capital market.

John Young: And is the "you" in this case, are they being tackled already by Goldman, JP Morgan, SG Warburg?

Andrew Pincott: I am sure they are.

Patrick Devine: Let us say you wanted to build a Eurotunnel and whatever the cost was - call it $10 billion - you wanted to raise that. Traditionally, you would raise that money by syndication of banks and you would end up with 250 banks participating on that loan and they will just lend the money. Then you have to go out and buy separate insurance for the Eurotunnel and the risks associated with it. You could - and it has already been done on Docklands Light Railway, for example - set up Eurotunnel plc, issue a bond to the market for whatever amount of money you needed and you would pay, say for example 8% return. You issue the bond, you capitalise the project, and the income stream to the bondholders is actually the income from trains going through the tunnel.

You can take it one step forward and say have an A bond and a B bond, where the return for the B bond holders is going to be impacted by losses. If there is a fire or whatever, they will get 5%, but they are taking that risk so the coupon on the B bond is going to be 9%, as opposed to the 8% you pay everybody else. You can throw into that pot some of the more difficult areas of insurance, such as political risk, though that is not really a problem in western Europe. If you try to do one of these out in a sensitive country in Africa or wherever, the question then is: "What is the appetite of the capital market for that type of insurance risk." We have already seen that the people who actually buy these bonds tend to be from the insurance and reinsurance industry by way of spreading their portfolio of exposure.

Andrew Davey: But there is capital market appetite. You can argue there is considerably more appetite at the moment among investors who are non-insurers than there is supply. There is all sorts of theoretical reasons for that, the ability to diversify, the ability to move the so-called efficient frontier and so on. There is no shortage of non-insurance capital interested.

John Young: If you are sitting on one of these bonds and you wanted to exit to somebody else down the line there is not a ready market in these sophisticated instruments.

Andrew Davey: As we speak, no, but there is a secondary market growing up. There is, for example, a secondary market in the USAA bond. No one is going to pretend it is highly developed or that it is enormously liquid but it does exist. From a practitioner's point of view, the fact that there is a secondary market is an advantage. One of the shortcomings of the insurance industry is that it is highly illiquid. Having done a transaction, one is stuck with it.

Peter Casey: Going back to the example of a bond with variable interest, from my point of view as an insurance regulator, it actually raises very few questions in relation to the issuer. After all, to a certain extent, the issuer bears the risk, which in traditional terms would have been one of inadequate insurance, of the fire costing more than is gained by reducing the interest.

My interest as a regulator is this: who has bought those bonds and what is their value? If the buyer is an insurance company who will carry it on their balance sheet and the balance sheet will lose value if the coupon ceases, there is a potential problem. But from the issuer's point of view, if the bonds are fully collateralised - then problem, what problem? The question for me always has to be what is the risk that they eventually pose that policyholders will not get paid?