1 June 2012Re/insurance

Property cat: from the front line

In the words of XL’s CEO, Mike McGavick, in his annual letter to shareholders: “catastrophes fell in a big way in 2011”. After a relatively benign few years—2010 being, perhaps, the exception—$110 billion of natural catastrophe losses in 2011 reshaped the landscape. A turn finally materialised, with some geographies’ rate rises proving marked. Extreme events and associated losses are an integral element of the property cat landscape, but nevertheless, important lessons were gleaned from 2011 events.

The fallibility of metrics

The first major lesson of 2011 was that catastrophe models and data sets can—and sometimes do—fall short of expectations. As Justin O’Keefe, senior vice president, US catastrophe underwriting at RenaissanceRe, explained, there were surprises for some players as underwriting tools and vendor models were unable to cope with the full gamut of international exposures and losses. “Models need to be viewed as tools in the underwriting process. An overemphasis on such capabilities can result in a deterioration in results and surprises emerging post-event,” he said.

Charles Cooper, president and chief underwriting officer of XL’s Bermuda reinsurance operations said that significant losses from the Japanese tsunami and Thai floods—cats that “were essentially not priced for and were certainly not captured in the vendor models”—had led reinsurers to re-evaluate their previous assumptions and approach to vendor models. “Reinsurers need to carefully analyse exposures and price accordingly,” Cooper said—“and not just for modelled exposures.”

Stephen Young, chief underwriting officer and head of reinsurance, Bermuda at Endurance, spoke in a similar vein, arguing that “data quality was identified by a number of reinsurers as a significant issue in 2011 and there were also questions raised regarding the frequency and severity of losses predicted by the models”. Recent changes such as RMS 11 were still being fully digested by many firms, while blind spots in modelling coverage caused still further challenges.

Highlighting the Tohoku earthquake and tsunami as a case in point, Young said that the event had not been “fully considered in the event set”, leaving decided gaps in knowledge. Lessons will inevitably be learned from events such as Tohoku and the Thai floods, he said, but trust in the models, data and event sets has been eroded.

International concerns

The next major lesson has been diversification’s lost lustre. Encouraged by the rating agencies to diversify their risks and pursue business that is less capital-intensive than traditional peak zones, Bermuda reinsurers have sought international business in recent years—often at very low rates on line. However, the weight of international losses in 2011 has led some reinsurers to re-examine the value of this push for international business. As Edwin Jordan, chief underwriting officer and chief strategy officer at Tokio Millennium Re explained, “we learned that sometimes diversification is not a good thing”. Whereas the company had previously taken on a lot of high layer aggregates globally and suffered losses in 2011 as a result, its strategy has now changed to one more focused on writing business lower down on the programme in order to generate more income and make the diversification play more viable, he said.

"The pricing environment may have rebounded, but conditions have nevertheless encouraged greater take-up of retrocessional coverage, the deployment of sidecars and the issuance of cat bonds."

Jordan added that the recent push to add diversifying lines and geographies to the portfolio had led to firms pursuing business with “very low rates on line. Such business was regarded as a diversifierrequiring little capital to write”. This had made it an attractive play, but had left the industry exposed on international lines. As Cooper explained: “The reality of the reinsurance market is that it is very capital intensive. If you are going to focus on a particular area, such as short-tail business, you need to diversify and leverage that capital base if you are going to make the kind of returns that are going to attract investors. Clearly people will want to diversify going forward, but they will need to be more careful. There is increased understanding that there needs to be a minimal margin, irrespective of any incremental capital approach that they take with their portfolio.”

Meanwhile, events such as the Thai floods had encouraged the industry to seek a better understanding of international events. David Marra, senior vice president, specialty reinsurance at RenaissanceRe, said that while “off the charts on a lot of people’s probabilistic loss curves” the floods had encouraged reinsurers to ensure they are “comfortable with the aggregate risks they are taking on internationally”. He said that as an industry there are numerous risks out there such as those which materialised in Thailand, with business sometimes taken on in ways that “can be seen as irresponsible. We believe that such events are unlikely to occur, but when they do, they serve to highlight the potential inadequacies in understanding those exposures”.

Marra argued that the industry needs to take a more cautious approach to such non-peak exposures. Cooper added that events had encouraged reinsurers to take a more geographic-focused strategy, with greater care being taken when writing worldwide treaties “particularly given the numerous exposures that go into those”. Despite 2011 losses, diversification will nevertheless remain a component part of the reinsurance landscape. As Marra outlined, events have resulted in a “more sceptical view of risk, but the benefits of diversification—if done properly—are very real. The movement we have seen is more of a refocusing on risks that are non-peak, rather than a move away from them”. And as Mike Van Slooten, head of Aon Benfield’s international market analysis team, was keen to highlight, “we are dealing with a pretty unprecedented sequence of events”, with such an accumulation of international losses hardly a fair reflection of the usual risk-return profile of diversification.

Finally, Van Slooten added that further concerns had been raised in 2011 by the troubled investment environment. As he outlined, “the earnings of the industry are heavily driven by the investment side of the business”, with recent volatility in the capital markets generating “significant headwinds” for the sector. This had meant underwriting losses have been all the more marked, with investment income stymied by global economic conditions. Conditions had created a realisation that investment returns would not always be there to buoy up difficult underwriting years, encouraging greater emphasis on technical pricing internationally.

And the upside ...

After a year to forget, it seems that the profits of Bermuda reinsurers have bounced back, with glowing results achieved in the first quarter. Reflecting on the improved pricing environment for property cat, Young said: “Property cat has become more attractive following 2011 losses. Although it is a volatile and capital-intensive business, it can boast strong margins for disciplined underwriters. It’s risk that we are comfortable taking.”

Pricing has inevitably reflected losses, with cat-hit lines registering the most significant rate rises, but it seems that there is now a general upward development on the line, with price increases finally ushering in a healthy dose of good news for the reinsurance sector. Nevertheless, “cold spot areas persist, where pricing is pretty cheap compared to the US”, said Jordan, with appropriate rates requiring time to build into the international underwriting environment. Nevertheless, further rate rises appear on the cards as we head deeper into 2012. This can only be good news for Bermuda reinsurers.

It seems that it is not just pricing—the view of risk has also changed following 2011 events. O’Keefe said that much as occurred in 2004 and 2005, losses—particularly those internationally—had led to a change in pricing, “but much of this is being offset by a changed view of risk”. Cooper spoke in a similar fashion, raising the question as to whether “rates are going up enough to compensate reinsurers for their elevated perception of risk”, with events in 2011 likely to weigh heavily on the mind. O’Keefe added that such questions need to be considered carefully by underwriters from a capital allocation standpoint. The question becomes “Are you getting paid more or less than you used to for this new view of risk, and to what level are you prepared to leverage your balance sheet?” he said.

Alternative plays

The pricing environment may have rebounded, but conditions have nevertheless encouraged greater take-up of retrocessional coverage, the deployment of sidecars and the issuance of cat bonds to satisfy calls for both capacity and wider capital plays. As Young outlined, “there was a significant uptick in retro purchasing following 2011 events and we expect to see even more interest as we head into the wind season”. He said that additional cover was not only being sought in the retro space, but also from primary insurers concerned about the prospects of a busy hurricane season. Florida and Texas accounts, in particular, were generating additional demand, he said, with the industry paying close attention to new risk models when considering their exposures. Marra was more conservative in his estimates of additional retro take-up, indicating that RenaissanceRe had not observed as much of an increase in retro purchasing “as you would have thought after $100 billion plus of losses, and from some pretty unique zones”. Market capacity had been enough to fill any gaps, he argued, with few dislocations brought on by 2011 events.

Cooper, for his part, said that retro take-up very much depended upon the level of losses suffered. For those that had taken a battering in 2011, they “certainly don’t want to enter US hurricane season with theirbalance sheet exposed, as that then raises questions of survivability”, while others were better positioned to weather any impending storm. Retro coverage would appear a must for some players. Cooper added that corporate governance guidelines had encouraged some reinsurers to seek retro cover as surplus, capital or equity thresholds were breached by 2011 losses and changes to the vendor models.

Sidecars have also proved popular, with a number emerging in 2011 to satisfy calls for additional capacity, much of it in the retro space. But perhaps the most marked interest has been in the insurancelinked securities (ILS) market. Cooper said that “from a relative return perspective the ILS space looks quite attractive right now because of the low interest rate environment”, with events in 2008 having helped to prove the non-correlation of ILS with the wider capital markets.

Marra agreed that sidecars and ILS were an “efficient way to take on property cat risks”, and he predicted that the growth that had been experienced in the past few years would continue. Looking ahead, he said he expected “most of that capital in-flow will go to existing sidecars and those players that have shown that they know how to manage third-party capital and deliver returns to their investors”. It seems likely that interest will continue as capital entering the market looks to take advantage of perceived opportunities and protect troubled balance sheets.

Model behaviour

2011 ushered in significant model changes, with RMS 11 being perhaps the most substantial. Developments had considerable ramifications for the industry, both in the US and Europe, with the new models acting as drivers of upward rate movements in both regions. The changes led to a new-found wariness among reinsurers, with the industry increasingly insistent on the need for a multi-model approach. “Changes in the two main vendor models prompted a healthy scepticism,” said O’Keefe. “It has encouraged risk-takers to use models as tools and guiding factors within their business decisions, rather than the ultimate answer.” This was a healthy development, he said, although he accused the industry of having a short-termmemory. “You continue to see post-event, post-model change, with people making dramatic decisions from a cat model standpoint, wavering between going back to more traditional exposure-rating and exposure-view methodologies and staying with specific probabilistic cat models.” It remains to be seen which way the industry will go.

"Strengthened enterprise risk management helped the industry manage its exposures, with the proof of its durability evident in the limited number of negative rating actions taken following 2011 losses."

Jordan concurred that a certain level of healthy scepticism as regards the models is a good thing. He said that reinsurers were looking for “more transparency and more options on how to run their models. Reinsurers want to be able to take their own view on risk and tailor the models to particular cedents and perils”. Cooper spoke in a similar vein arguing that “no-one wants to use new models straight out of the box”. Rather, they are seeking a customised view of risk. “The industry should not be using the same piece of software to price our business, particularly considering the volatility of the model changes,” he said. Such an approach would create dangerous levels of convergence. Instead, reinsurers should establish their own viewpoint. Such an approach would “create more divergent views between reinsurers”, said Jordan, promising the brokers a headache, but greater variability in rates.

Finally, touching upon suggestions that AIR Worldwide and RMS were working towards data convergence, Cooper said that such a move was a positive development for the sector. “Data convergence is good for the industry—results convergence is not,” he said. Young concurred, arguing that “modellers should not be competing on the data itself, but rather on the models”. Creating united data sets should mean reinsurers will have more time to establish and analyse their own particular viewpoint of specific risks.

Bermuda bounce

Most reinsurers would be happy to forget 2011. Losses in certain territories were marked, far out on the probability loss curve. Balance sheets took a battering and consolidation once again became a component part of the landscape. Nevertheless, the industry held up remarkably well. As Van Slooten indicated, “had this run of losses happened 10 years ago, we would be looking at a very different market environment”. Instead, strengthened enterprise risk management had helped the industry manage its exposures, with the proof of its durability evident in the limited number of negative rating actions taken following 2011 losses, with some reinsurers even having benefited from positive rating developments, he said.

2011 certainly served to highlight the pitfalls of property cat, but at the same time underlined the invaluable role Bermuda reinsurers play in meeting the world’s risk needs and its ongoing strength. After each major loss-year Bermuda emerges stronger and wiser. 2011 appears to have been little different.