For the January 2010 renewal, rates are expected to hold at around 2009 levels. Given that 2009 saw only a few headline catastrophe events, what’s the reason for that?

Firstly, yes headline losses have been few and far between, but there has been a frequency of losses affecting insurer retentions –  for example windstorm “Klaus” in France and Spain earlier in the year, hailstorm in Austria and Switzerland, typhoon in Asia, earthquake in Italy, Californian bush fires and windstorm in Canada. And we’re still not through 2009, with ongoing earthquake risk and the European windstorm season continuing through November/December. These frequency losses are maintaining the demand for reinsurance capacity and therefore holding rates firm.

Secondly, counterparty financial security remains paramount - it pays the claims and ensures continuity of offer. Insurers are prepared to pay to ensure that they have financially strong reinsurance companies supporting their portfolios,  and that means doing business with reinsurers with a technical, informed view of risk that set an appropriate price.

Thirdly, solvency regulation is leading to a restructuring of insurance company balance sheets. That’s again maintaining the high demand for reinsurance capacity as well as the call for secure reinsurance partners.

So in terms of overall outlook. In the absence of further 2009 headline losses, demand continues to be high and there’s no sign of that changing. Catastrophe reinsurance capacity is adequate but not overly abundant. Previous loss affected markets may see some slight easing of rates from high levels, but generally rates can be expected to hold at current levels for the reasons just mentioned – the 2009 attritional losses, the need for financially strong reinsurance partners and evolving balance sheet requirements.

2009 saw several “reversions” of commercial catastrophe models and PartnerRe has just updated its CatFocus® earthquake and European windstorm models. Can the industry now confidently rely on model outputs?

 At PartnerRe we don’t price our business purely to the model output – even as they evolve and confidence in them correspondingly increases - rather we use catastrophe models as a support tool, a valuable component in the underwriting process.

Updating our models is inherent to their use and has a significant positive impact on their reliability. Models make certain assumptions and are based on currently available data. As new data, improved modeling techniques and new scientific knowledge come to light, it is incumbent upon all model developers to review the models and to challenge the assumptions to ensure that the model view of risk and expected loss remains as accurate as possible. As the models become more sophisticated, we can indeed increasingly rely on their outputs, but only if they are taken for what they are, models, and used appropriately within the underwriting process.

What can you tell us about the new version of the CatFocus® earthquake model?

The earthquake model was one of the earliest CatFocus® models and although, as with all models, it has been continually evolving, this is its second major update and we’re already using the new version to evaluate risk. The model is highly advanced, building on the latest scientific findings and engineering expertise on seismic hazard assessment and vulnerability research. Amongst other improvements, the model now includes a state-of-the-art modeling methodology to disaggregate aggregated exposure data, as well as a unique approach to vulnerability curves based on a methodology that delivers empirical vulnerability curves from a worldwide database for a wide range of different building types, occupancy classes and regions.

We believe that it’s important for all parties – not just modelers, but also brokers, risk managers, insurers and reinsurers – to understand in principle how a model evaluates a portfolio of risk and more importantly what the numbers do and do not tell us. For this reason, we will soon publish a new PartnerRe technical publication that gives a lot more detail on the important components and assumptions in our updated CatFocus® earthquake model.

And the CatFocus® European windstorm model?

Here we wanted to do two things. One, to extend the storm event set that underlies the hazard component of the model, and two, to do this in a way that gives us an especially robust view of the largest windstorm events. The normal approach is to create new storms by adjusting the physical parameters of historical storms; that’s a good approach, but it has its limitations in terms of what could be possible in the future and the appropriate return periods of events.

What our researchers did instead was to use fourteen different climate models to generate the windstorm events. This is a unique approach in itself, but the number of climate models used makes it even more impressive. Our event set now includes events that the climate models anticipate, together with their return periods, and does that with minimum model bias. We’re also using the climate models to assess future climate change risk.

In updating both these models, we have substantially improved the confidence in our understanding and pricing of earthquake and European windstorm exposures.