Helen Yates: Looking back at last year's hurricane season, this has had a very big impact on various different aspects of the industry. We have obviously got these alternatives to reinsurance, which are growing in importance, but we have also seen adjustments to catastrophe models and from the rating agency point of view changes in capital requirements, how will that impact upon strategies this season?

Marcus Rivaldi: Well I think the models have done a great job in indicating to us all that we may be living in a riskier world at this point in time, particularly in the short term rise and fall in US marine exposure, and a number of modelling agencies have got different expectations for what in the short term they know are long term averages with regards to these types of storms. The way we look at it is that our capital modelling is meant to be a venture market tool for us, to get a feel for how much capital is in the business, but we also really look to see that businesses are grown up and mature enough to have a real view of how much capital they have got and how much they need. So all this work we are doing on enterprise risk management is to really try and establish that people have a good understanding of the risks they are running and, over time, we can envisage a situation where more and more information will come from companies' own models and they will rely less and less on our own model patterns. Companies already have a better understanding of the risks they are running. It is all about intelligent usage of modelling.

Thomas Blunck: It clearly shows where the direction of insurance and reinsurance is going. It is very clear that one of the core competences for both, and for brokers as well, is becoming data availability, granularity, modelling competence and then solutions to mitigate the exposure.

Matthew Grant: The models were changed partly to reflect the increased frequency of events and secondly to learn from Katrina. So as different modelling agencies have taken their view on the way they should reflect the frequency, as Marcus says, there are different views out there about what is the right way to think about frequency. So whereas in the past you saw a convergence of the view of the hurricane risk in the US, this year there is going to be more variety out there and I think that is going to be a challenge, particularly for the brokers placing the risk, to really understand when the reinsurer talks about their view of risk whether they coming at it from an RMS perspective, AIR perspective or the EQECAT perspective, because they are all slightly different. The second thing comes back to the issue of data. The models, and you learn from past experience, are now making much more differentiation between information that is associated with different types of risk, and so there is going to be much more variation now in the portfolio that has been analysed this year to the one we were analysing last year. Just one statistic on that kind of brings this home. If you look at a commercial building in Florida for the same location and you assume it has got the best defined characteristics to resist hurricanes versus the worst defined, the previous models had a difference of three times the loss estimate and it goes up to 18 times now, so you can see the massive changes in portfolio analysis that have occurred.

Peter Grove: The modelling is essential to understanding your portfolio of course but you must have a grip on what cover you are giving, what sums insured you have given, and you start with that information and then you look to the modelling. But you must start with the fundamentals, what sums insured have I given? That to me is the key point.

Jeremy Brazil: And that is exactly what we are doing as well, not forgetting the models but starting from basics, what have I actually got out there? What is the probability of losing that, distance from the coast, construction type etc? Then we put it through a model.

Thomas Blunck: You asked before about whether a company should have a centralised or decentralised buying approach and I think the discussion clearly shows the need for a central unity for setting the standards for the data, for the modelling, for the exposure measurement and then for the big exposures as well as the buying side.

Helen Yates: Do you think anything has changed since last year? Was there perhaps an over reliance on modelling?

David Priebe: Well, define over relying. I think what's changed is a better understanding of the tools available and I think there is an even more intense utilisation of those tools than there was before. I think Peter may have said it earlier that the mistake in the past was, "Oh this is the number so therefore I manage to the number". What is happening today is, "Let's really understand my exposure, let me understand what the tools are saying, what are the changes that have been going through those tools and then develop a strategy around that". So there has probably been a broader use of tools, "Let me look at RMS, let me look at EQE, let me look at AIR, let me look at my own assessment and let me make very sound judgements around what that means and then how I manage that risk". Thomas is correct from a global perspective; you really need a central unity to look at the entire landscape and to set clear guidelines in terms of data quality and in terms of how we are going to make sure we get consistency from that so that you can use the tools more effectively. So I have seen a step up.

Thomas Blunck: One more point, models are always models, they are not reality and they will never be, so we will always have a certain amount of uncertainty. That clearly gives us the message that we have to go beyond modelling and that means in the end we have to do stress test scenarios as well, and I think that has perhaps not been the focus in the past but going forward it will be. Use modelling and stress test scenarios.

Marcus Rivaldi: Would it be fair to say that the updates to the models effectively came too late this year to influence any renewals?

David Priebe: No, no, no. Our clients and certainly the reinsurers sitting round this table would say they have already put them into effect, if not running 6.0 in their models, they have made their own assumptions as to what 6.0 means, so no, it is not fully baked but it is baked into the analysis clearly.

I think a lot of the clients have been doing their exposure assessments and clearly they are becoming far more conservative. They have done what Jeremy said, they have looked at their concentrations of exposure and focusing on that, plus their overall portfolio, and then making decisions on how to mitigate that risk, that is beyond the tolerance of their risk strategy. But to the best of my knowledge they are pretty much ingrained, people have already brought that to bear and we saw that in the seven reinsurance programmes that were executed in April and July. The impact of the new awareness came through and we saw it in pricing and we also saw it in terms of the quantum that companies were looking to protect against.

Matthew Grant: We started a communication effort back in November as soon as we had a sense as to what direction the numbers were going to move in, and between November and the models coming out in May we gradually became more precise about where the changes were going to be. And the feedback we got is, as David said, that people felt the numbers, when they came out, were reasonably consistent with what they would expect and in some cases I think people had slightly over-estimated, but by and large they were consistent. We had a conference out in the States and the comments were all along the lines of, "That is extraordinarily different to what I thought it was going to be". There seemed to be a lot of recognition.

Peter Grove: Last year the models gave a false comfort to a lot of management. That will not be the case going forward. I think the insurance industry understands a lot more about models this year than sadly what it did last year.

Jeremy Brazil: Picking up on that point, I think people were complacent because, to be fair, there had not really been anything, talking about wind. Every day is a quake day and what would have happened if there had been a huge earthquake in San Andreas? People were complacent because they were getting the number they wanted it to be rather than sitting down and challenging it and saying, "Is it the right number? And if it is the right number, shouldn't we factor in some margin and say we should work within that number". And that is what has changed, people have now recognised that capital and people's jobs are on the line, we have got to get it right going forward. The way to get it right is you either don't write cat business or, as Peter says, you just keep focusing on what the exposure is.

Helen Yates: Do you think the general risk appetite of insurers on all catastrophes including earthquakes has reduced in this last year, given that a few large insurers have decided not to rewrite certain policies?

Jeremy Brazil: We have seen it on the insurance side and I think our open market underwriter was saying this time last year, pre-Katrina, there might have been $1bn or $1.5bn of Fortune 500 risk, and now that is probably down to $200m or $300m in capacity, so the insurers are taking it on board. The markets that I felt moved the most after Katrina was the energy market and the few markets where the insurers and reinsurers move together very quickly, and comparatively speaking I think they have got a slightly more robust model than perhaps we have got on the non-marine side.

Thomas Blunck: I think there is one point we haven't addressed here but it goes beyond this season. Most of the companies have looked at the exposure using the new models and know more or less the aggregate, but the next step will be to drill down that exposure. Through the risk capital allocated to each and every line of business and region, to the different products, and then they will find out that they have very interesting products in one area but the other ones don't really earn the money they should. So perhaps there will be a total reshuffling of the strategy in the front end in the next year or two and it will then be defined by the risk appetite and the available reinsurance or capital market capacity.

Helen Yates: Do you generally see this as a mismatch, like Jeremy was talking about, where insurers are taking on more of the risk because reinsurers have got less of a risk appetite this year?

Thomas Blunck: It might be a mismatch yes and it will have implications on the product and the regional strategy of the insurer, that's true.

Jeremy Brazil: I wasn't saying it was a mismatch in that respect, it is a mismatch I think between understanding the model's risk appetite and capital requirements between the marine and non-marine market. I don't think there is a mismatch between what the direct insurers are doing because, and you mentioned Allstate, they made a decision to keep where they are at and they are going to have to buy in a lot more cover to do that.

Peter Grove: The reality is there is a limited retrocession market this year, therefore the pass the parcel theory doesn't work in the same way and that is why you have less capacity and frankly, I believe it is that simple.

Marcus Rivaldi: And it's also just human nature. The number of executives you speak to who say, "If we get it wrong this year, next year you will be speaking to a different management team".

Jeremy Brazil: Our successors will be deciding where we went wrong.

Marcus Rivaldi: Exactly. It is one of those issues about how general bullishness leaked out of the market after Katrina last year.

Jeremy Brazil: I expect David and Peter may have seen this but I don't think there is a mismatch but rather a difference of opinion of how some of the people in the market look at the models. How good is their aggregate, how good is their data, do they really understand the capital requirements, do they understand the rating agencies and their models? I think that is why, if we have another bad year, there will be some people who will have a modest loss within their expectations and other people will have surprises again.

Helen Yates: Do you think there is enough understanding of the models?

Peter Grove: I think there is a lot more today, post-Hurricane Katrina than what there was pre-Katrina. I am sad to say that I think there was an ignorance in many quarters before and when people spoke of one-in-100, two-in-250 and one-in-1,000, the belief was that period was how long it would be before the next big one came about. I don't think it was particularly understood that obviously a global exposure is a global exposure and major catastrophes can occur in many parts of the world almost at the same time. So just because in one particular location they can judge the one-in-1,000 risk, it doesn't mean to say that all locations you have on your books are judged that way. That wasn't understood by everyone before. I believe it is understood by the majority of people now.

Matthew Grant: It is still a very imprecise way of measuring things. Probabilities are a much more effective way of looking at it but as Peter says, there is a very big difference between the return period of a Katrina-like event happening again in New Orleans versus a return period of a $70bn loss and I think more people get it but in a way, it would be good if we could find new terminology to describe risk in a way that was unambiguous so everybody understood it.

Helen Yates: Someone mentioned starting the whole process earlier. Is there any feeling that because we don't know what is going to happen with this year's hurricane season, people might want to wait and see what's going to happen?

David Priebe: I would say the process has started. We are already fully engaged in developing the programmes for the upcoming 2007 period. In fact, I think even those clients who have just completed their mid year renewals are now thinking about what they could do to supplement that coverage and prepare for 2007, maybe look at earlier strategies around that, so clearly the upcoming wind season will have a bearing on those strategies. Most importantly, what is going to be the impact on availability, which will have an impact on strategy? But right now people are moving forward and working hard to make sure they are getting the data in a format that is going to enable those risk providers to give maximum capacity at the best price and try to create greater transparency and clarity to the process.

Rupert Boswell: David, I have seen it written that some of the potential additional capacity providers in the capital markets are holding back or have a tactic of holding back until later in the day in the renewal season on the footing that they could charge more and come to the rescue at the last minute.

David Priebe: Well it is back to the market dynamics. I think there are a lot of people who said, "Darn, I wish I hadn't sold so much capacity in 01/01/06 as I did because I could have really used that in April, May, June and July", and everyone is always playing with that sort of theory and that is a concern.

Rupert Boswell: Hedge funds are sort of nimbler in a way aren't they? You don't have to produce these business plans with people pouring over them in quite the same way.

David Priebe: The hedge funds are clearly there to maximise return and they want to make sure they optimise that capital that they are going to deploy and so they are serious at looking at strategy. And when it comes to strategy, whether they're going to layer out tranches of their capacity or capital over the course of the year, how much they're going to allocate at 01/01, how much they bring back and add to the market at 01/04 and what is going to be there for the Florida season, in May, and what is going to be there for the post-Florida season. And that is going to be an important dynamic to be really on top of, because that flow of capital will have a bearing on availability and price and so I think as a reinsurer trying to access that capital, or a primary insurance company putting together their core reinsurance strategy, they are also going to have to think about that in terms of the dynamics of their programme. It is going to be like the old days. The days of, "I put my programme in place on the day I like to put it in place and it runs for a year" are gone. It is really going to be a continuous process.

Thomas Blunck: Here we go again adding an old-fashioned view. We try to offer our clients a stable capacity and don't wait until the last moment where we can have a beauty contest to say who is paying more, that is part of a value proposition but of course it is linked to the fact that we want to have risk at fair prices and if we don't get them the capacity goes in a different direction.

Rupert Boswell: Is it hedge funds in London that are more interested in the industry than US-based hedge funds, or can't you tell?

David Priebe: If you look at the level of hedge fund involvement in the reinsurance business it is mainly dominated by the US funds but it has been growing elsewhere. One of the key things we have been doing a lot of, which is under the radar, is education and really trying to broaden the range of investor interest in insurance and in reinsurance, to draw them in to supplement and to support the reinsurance capacity and capital. So up until about six months ago it was predominantly North American-based investors who were really into investing in insurance risk but it is broadening. We are seeing a growing interest in terms of European-based funds.

Thomas Blunck: I think the part of the capital markets that is taking on insurance or reinsurance risk is a very juvenile market, with just a couple of years of experience. The next task for us in the industry is to educate the market and to further develop products so that it fits the risk appetites of the capital markets, and that is a process that will be developed over the next few years, and only then will we be able to increase the capacity of the capital markets for the reinsurance industry.

David Priebe: Right now there is a range of products from far more insurance securitisations to collateralised reinsurance programmes on a parametric or index basis, to capital supporting professional reinsurers in the form of sidecars. But one of the key issues is transparency, understanding that this is a properly structured and priced deal because the investors want to make sure they are not being taken advantage of. There needs to be clarity in understanding who is going to validate the creditability of that proposition and how they know it is a fair one at a good price. Back to the start of the conversation - in terms of will people hold back to see if the price is going to move up or down - it is a free market, so you are going to see a lot of interesting dynamics.

Marcus Rivaldi: The markets are actually broadening their appetite into new areas where we are seeing new deals.

David Priebe: You have to, because the market will not expand if all we are looking at is Florida, south east wind, north east wind and California quake. The market will stop and it won't grow much further than it is. So to really stimulate the market you need to bring in a broader diversification of uncorrelated risk, be it in different zones or be it in motor and other securitisation areas.

Thomas Blunck: By using capital market solutions with most of the products you have basis risks. It is very well known you have to address that basis risk and not only within your company. For us it is very clear that you have to address it with the rating agencies together with your regulatory authority - they will look at that as well with your internal model. If you don't get a grip on the basis risk the products will continue to be problematic in terms of having an effective risk mitigation strategy.

Helen Yates: What is the rating agencies view on this?

Marcus Rivaldi: Our view on sidecars varies from business to business. In some cases the sidecars are effectively propping up a competitive position, allowing a business to maintain a top line. In such circumstances one could question whether the need for reliance on the sidecar is a sign of weakness. Other examples however can be construed more positively - the sidecar is being used to effectively and proactively manage risk.

We've talked about this earlier, about the longevity of appetite in this space amongst some investors. What if there is a bad year in terms of hurricane losses? Could we see existing or new investors avoiding the asset class and a resultant reduction in cover being provided by this source?

Peter Grove: It is great for short-term planning but it is not so good for long-term continuity and I think that is the worry with the capital markets at the moment. One can't look to the long term and guarantee that you have got a capital market product out there that can reinsure you in quite the way that a conventional reinsurer can.

Rupert Boswell: What would make them want to be long term? Given that they are happy to make money, what could change to make them want to commit?

Peter Grove: I think it is more of an understanding, as Thomas said earlier. It is a very young market in many ways and it is on a learning curve at the moment, like everybody is when they look at something new. And the more that product is understood the more long term the possibilities are.

Helen Yates: You say it is a very young market and mentioned educating the market on insurance so that they understand what they are buying into, but is there a risk they could be scared off, by things such as Olympus Re and another active storm season?

David Priebe: Certainly, our business is extremely volatile. It is a high risk business and therefore education is about really making sure those investors understand what they are getting into, what the risk/reward trade offs are and how that fits into their overall asset allocation. Insurance risk is pretty much an uncorrelated risk so for someone who is managing a large pool of assets, insurance risk - if written properly - should increase yield and reduce volatility. That is the theory and that is what you need to get people to buy into, and the more that they can understand and assess what they are investing in, the better. So it is critical that we really help investors understand the nature of the risk they are taking on.

I think Olympus Re is more of a concern for people like S&P, not the investors, because the investors capitalise their liability. It is about understanding how that falls through the chain of your capital and making sure that the exposure out the back end is recognised in terms of the exposure of your capital base and then how that is assessed on your balance sheet. So from the sidecar's standpoint what is key is getting the right balance between how much of that liability is collateralised or not. In an ideal world, 100% is collateralised but it may negate the return for the investor because it is now not leveraged to the point that it makes it attractive.

Marcus Rivaldi: Olympus is not the only example, there are others out there and when we talk about contract certainty, I have seen examples of very clear contract language yet there is some other area of dispute that arises between the capital market and the traditional insurer and things can get very messy. So picking the right partner, someone who really understands the insurance business is vital because if things do get messy they may resort to litigation.

Thomas Blunck: If you look at the products the insurance and reinsurance market is offering to the capital markets, they are very simple compared to the credit market for example. They have a far more structured approach, offering the investors very different opportunities in a structured way and I believe, going forward, we will also have to structure our products differently so that they offer a more economical, feasible and understandable solution for the capital markets.

Helen Yates: Can we move on to talk about the relationships between the reinsurers, the insurers and the brokers and basically what do you all want from the various parties you interact with? Would you like to comment on that Peter?

Peter Grove: I think the relationship is the most important ingredient in many ways. You mentioned Monte Carlo earlier. The communication end of it - client meeting reinsurer, client often being introduced to reinsurer by broker - it is vital. I bought contact lenses because it is important that you do look your future partner in the eye and that you get to understand each other, because during that communication you get a true feeling of the relationship and it is still vital, even in this day and age. The relationship helps the continuity. Sadly so much of what we do in our business is not black and white although we have heard Rupert arguing for the need for the black and white clarity, but at the same time life isn't that simple. The more people get together and the more there is that sort of communication, the better it is because that is when a client can really believe when the reinsurer says, "I will give you continuity", and that is when a reinsurer can really believe when a client says, "I will give you payback". But it is not just a case of going away and meeting people at conventions, it is very important too that the client travels to the reinsurer and the reinsurer travels to the client, as that is when you truly start to build up a relationship, and obviously the broker is in the middle of all that and that is a good thing.

Topics