Eric Paire: That is a point that leads back to the earlier discussion. Effectively when you start to look at your exposure from an ERM perspective or a Solvency II perspective, when you have some rating agencies talking about cat exposures based on 250 year events in the aggregate, then effectively you end up being more concerned by what happens over the aggregated period of time.

Following on from the comments made by Jonathan about basis risk, we are now helping many of our clients to look more in terms of sensitivity than basis risk. In other words, what the impact is of you being wrong on the assumption in the model or what the impact is of that trigger not acting exactly as it should. It is not really a question of having a given figure for basis risk; it is more a question of knowing how much at risk you have if your assumption on one given part of the modelling work is wrong.

Jonathan Barnes: I was involved in an indemnity bond that was arranged at the end of last year but it had a comfort warranty in it based on PCS losses. We were able to provide comfort to the sponsor of the transaction by modelling the basis risk around that industry loss trigger. The models allow you to inspect the losses of the sponsor versus the industry loss by each of the events in the stochastic event set. If you see that 95% of the time you are satisfying the warranty requirement prior to penetrating the indemnity layer, then the sponsor should gain some comfort from that. Maybe we will see the development of basis risk covers as explicit protections to augment the bond purchase.

Eric Paire: With basis risk, and going back to Alan’s initial comment on data, one other good thing is that the issue of basis risk is also pushing everybody to look very hard at the data quality of the line they have been using. The sort of study that you do for market loss you also have to do it on your own portfolio. You need to ask yourself whether the assumptions in the model are realistic. It goes back to the quality of data you are using and how effectively you use it.

Helen Yates: Can I just ask for a quick perspective from the rating agency on that?

Peter Grant: Do you mean basis risk?

Helen Yates: Yes, and in terms of how the rating agencies view vehicles like cat bonds. I know there is one or maybe two with “A” ratings but that’s about it right?

Peter Grant: S&P adopts a slightly more conservative stance than some of our competitors do. Recently, I profiled the State Farm cat bond issue. I think it was the largest ever at $2bn. Some of our competitors, Fitch in particular, were able to rate the uppermost tranche of that as “AAA”. We take a slightly more conservative stance in the way we apply our ratings. I think for a cat bond that could potentially be triggered by a single event, we cap our ratings at the “BB+” level, but I would need to confirm that.

Eric Paire: In practice you have trouble going up to “BBB” or above anyway.

Peter Grant: Yes. I guess to some extent that is a reflection of what we consider to be the inherent uncertainty posed by model risk and other factors. It is something we find very difficult to come to grips with at the higher rating levels. In terms of basis risk, again that is something that perhaps we have been less prescriptive on compared to some of our competitors. We certainly evaluate it but we would look at it on a case by case basis and factor that in.

Helen Yates: As non-traditional reinsurance becomes more mainstream, do you think these concerns about basis risk will improve as knowledge grows and investors become more comfortable with the product?

Jonathan Barnes: We have seen greater sophistication in the use of different triggers to try to squeeze out basis risk. Therefore, looking at the standard PCS US wind trigger you can take that data by state and then by lines of business and start to weight those to reflect the concentrations in the underlying portfolio. I do not think we are there yet, but the impression is that we are moving in the right direction.

Nigel Ralph: I do not think it is an either/or argument. Large companies like us will sell both products. It just so happens, as Alan pointed out, the amount of parametric type business we sell is very small, and certainly less than 10%. Companies like ourselves in our peer group can move from one area to the other so we would not really get too bothered if one market develops more than another.

Going back to Alan’s point about the aggregate and event coverage, which certainly will be valid now in the UK for flood losses, as reinsurers we can sell both products. We can sell event coverage and we can sell aggregate coverage. It just so happens that aggregate is far more expensive than event and generally the capacity that is available for aggregate coverage is much less than on an event basis.

Jonathan Barnes: It is a question of pricing value though, is it not Nigel? I think this year you have been providing great value.

Nigel Ralph: Yes, aggregate covers would have been a great buy.

Alan Fowler: To put cat bonds in perspective, we are buying around £1.2bn of coverage in the UK. When we are looking at cat bonds they are probably struggling to provide more than about 10% of that capacity. You have to put it in the context that even if we bought one it will not be a major component, it would slot in somewhere in that programme. It is not that it is a substitute at the moment, but just a part of your armoury.

Jonathan Barnes: It is a complement.

Alan Fowler: It is a part of the armoury and I do not see that changing any time soon.

Helen Yates: So perhaps this is a good moment to go back to traditional reinsurance buying and give Alex an opportunity to comment. What role does technology play in buying strategies and how has this changed over the last two to three years?

Alex Letts: Sitting here and listening to people talk about $100m cat bonds, it does rather put it in perspective when you are talking about streamlining the process to save a few tens or hundreds of millions of pounds. I think that is one of the reasons why modernisation in this industry has been so painful. Ultimately, when the chips are down and money has to be counted, it is about minimising the amount of loss rather than saving a bit of money in the back office. In the last few years, particularly post-9/11, as Simon said, the quality of the wording has become terribly important. It is about making sure everybody is clear about what was said and meant by whom and when, and what the process was to get there.

I think the importance of technology is more around creating as much certainty as possible around the process and what has been agreed, rather than saving a few bob in the back office. Saving millions of pounds for brokers is a very important thing at the moment because their margins are being squeezed in the softer markets. That is very important to them, as is distribution. In terms of the wider industry perspective, the role of technology in electronic trading is much more around keeping track of records, making sure there is certainty of process, making sure there is less wiggle room in the dispute process, although that is bound to happen around any major catastrophe.

Technology will not solve the problems, but by and large following the intervention on contract certainty, following 9/11 and some of the regulatory issues Lloyd’s is currently facing, that is really where it will have its biggest impact and that is why it is being driven right now. However, funnily enough, having said that, I think the brokers with their margin pressure post-Spitzer are really the ones who are going to kick it all into place.

Eric Paire: I fully agree with your description about the role of technology, but I would just add one thing about the data. There is a huge amount of data flowing between insurers, reinsurers and brokers, and there are all these models being used and people trying to make more informed underwriting decisions. Without technology, there would be no way of sending these huge files with millions of figures in them.

The second point is the issue about the value chain, although when you say about the margin being squeezed I would probably phrase it differently. The processing of that data and the quality of the process is not why your clients pay us. At the end of the day, that is not really the key value they expect from us. They expect us to do it very well and very securely. If we can do that on the technology platform I think that is good for everybody. It also allows us to free up resources to provide more value-added services.

Alex Letts: I think it is very important. It is part of the crusade to persuade the market that technology is not about disintermediating the individuals who are doing the thinking and value add. It is about taking the paper off them so they can get on with providing a better service. Certainly, I know that is many other brokers’ view. As you know, some have implemented an electronic trading strategy. They firmly believe that whilst they can save some money on distribution, they are freeing up their people to add more value to their clients at the front office. If they can save some heads in the back office, that can help drive down their costs and drive up their margin.

Helen Yates: So you are not replacing relationships then?

Alex Letts: Except in a Woody Allen film I do not think there has ever been any example of technology replacing relationships. I think it enhances them.

Eric Paire: Trading floors in big banks are typically the place where you find enormously advanced technology, but in fact people are paid for making good trades.

Alex Letts: The intention of technology is always to get rid of the unnecessary paper and unnecessary process. In any industry I have looked at, it has never ever been successful in disintermediating people. When the dot.com big bang was happening in 2000 everything was going to go direct and brokers were going to be disintermediated in every industry. I can only think of a few industries where a successful consumer disintermediation has happened. In the business-to-business industries that I am concerned with, it has not happened.

Alan Fowler: Just picking up on the paper issues versus technology, and a bit of a plug for Alex and RI3K, for our international programme we have to coordinate data coming in from over 30 territories. We use the excellent data management tool from RI3K. Before that, we just had spreadsheets flying here, there and everywhere, which was a nightmare. Technology is very important in that data management side as well.

Alex Letts: Thank you Alan. Data collection is the Holy Grail. When we talk to underwriters they tell us the real value to them is not in getting rid of the cost of the process but in getting better data to their desks so they can evaluate the buying decision more efficiently. What we are seeing with Alan and R&SA, and in fact Ace is doing exactly the same thing, is they are starting that process with faltering steps and gathering their data better. The next step will be to ship it to the other side better, so that there can be a very transparent analysis of the risk and the risk transfer can then happen on a fair basis.

Eric Paire: On that point, in terms of the broker’s position, it also allows us to ease that process by concentrating our energy on checking whether it is the right sort of data for the underwriter. Then we can be sure we’re transmitting something which makes sense in terms of the questions being asked and not carrying it around.