The doomsday sayers would have you believe that retrocession all but disappeared in 2006. But there was plenty of capacity available at the 1 January renewals, discovers Helen Yates, although it remains painfully expensive.
Retrocession is rather like a secret society - think the Freemasons or Dan Brown's Illuminati - you know it exists but you're not quite sure who its members are or what they really get up to. With only a small number of players, the providers of traditional retro have formed something of an oligopoly. Yet there has generally been plenty to go around and retrocession remains a key strategy for many reinsurance companies.
Retrocession is often highly susceptible to large catastrophes. Despite being a more robust market than it had been at the time of 9/11, most programmes proved inadequate when up against Hurricanes Katrina, Rita and Wilma. The losses incurred from Katrina burst through the upper layers of reinsurance and exhausted nearly all retrocession programmes at the time, which meant that substantial losses fell back into the laps of many of the retro buyers. But the impact wasn't felt immediately. "Everyone went to the market in December 2005 at two seconds to midnight. The losses had just happened and cedants had no choice but to buy reinsurance for their still highly volatile books. A year later, the situation looked very different," says JLT Re partner William Bennett.
According to Bennett, this time lag following Katrina meant that prices and capacity for retrocession were still very active at the beginning of 2006, although capacity for the larger worldwide retrocession programmes had fallen by up to 60% at the 1 January 2006 renewals. But it wasn't until mid-year that prices really began to soar and capacity dwindled. "In the past, major natural and man-made catastrophes did not lead to a significant reduction in available reinsurance capacity," said Hans Joachim Thoenes, head of corporate retrocession at Munich Re. He believes Katrina changed this picture. "For the first time, a real lack of capacity for a certain risk class (US hurricane) was apparent."
The anticipation was that prices at 1 January 2007 would catch up with the massive rate hikes witnessed at the mid-year 2006 renewals for the Florida book of business. "Those who still remain in the market continue to express a desire for even further rising prices - or for prices at least matching the level achieved on the 1 July 2006 retro renewals," Thoenes explains. According to John Edwards, who leads the non-marine retrocession team at Benfield, these retro sellers were disappointed. "Before the renewal season commenced there was widespread anticipation of considerable change to the availability of cover and greater restriction in contract scope," he says. "In fact, the level of change has been much more modest than anticipated. On a risk-adjusted basis, US exposed pricing has probably gone up between 25% and 35% compared with 1 January 2006 and for ex-US retro (which covers all non-US regions) the pricing increase is probably about 5% to 10%."
"Due to careful aggregate management during 2006, for 2007 the days of panic buying are not there anymore," says Edwards. This return to calm is the most significant change since last year. And retro coverage is now being limited to specific territories rather than being offered on a worldwide basis. The level of attachment is also considerably higher. "What has progressed is the sense that risk is measurable," explains Edwards. "Many years ago, retrocession provided blanket cover to a portfolio via a worldwide policy, but because information data is getting better and understanding of risk is improving, there is a narrower and more precise definition of what is and isn't being covered." Piers Cantlay, chief executive officer, Aon Re UK Specialty Reinsurance, says worldwide capacity was widely available at the renewals but that it is being placed on a different basis. "A number of reinsurers will still offer you worldwide but they will give you the ex-USA and the USA and Caribbean as two sections on the same slip," he explains. "It's a mechanism that allows them to have different attachment levels for the two different zones."
According to Benfield CEO Grahame Chilton retro capacity is still available, but at a price. "To a certain extent I'm not sure it ever went away - what went away was unrealistic negative arbitrage retro capacity. If one is prepared to pay a fair price there is capacity available. It depends on what book you want to write and what you want to buy. When you're a young driver and you want to drive a sports car you probably can get insurance but you may not like the price."
So why aren't we seeing more of a reduction in premium rates for retrocession? The 2006 hurricane season was, after all, a non-event. Nevertheless, the impact of Katrina, Rita and Wilma was still resonating long into 2006 and retro remains expensive and unavailable. "The adjustment has now taken place," says Edwards. "We've had the situation of the worst year ever followed by a remarkably benign year in 2006 - but in spite of the 'year off' all the practitioners within the retrocession market - buyers and sellers - have taken on board changed views of loss frequency and loss severity."
The perception of risk has changed and a major influencing factor was the recalibration last year of the catastrophe models, which amplified the risk and therefore the price of insuring property catastrophe. The rating agencies, having witnessed two years of intense catastrophe activity in 2004 and 2005 also made changes to their models, increasing their capital reserve requirements for those taking on catastrophe risk. "These factors made it more expensive for companies to provide retro protection," explains Thoenes. For some, the conclusion was that worldwide retro no longer offered opportunities for acceptable returns.
With anticipation that "mega catastrophes" like Katrina could become a yearly event, previous pricing and levels of attachment for retro cover, particularly US wind programmes, were deemed inadequate. Despite a benign year in 2006, catastrophe modellers are unanimous in their predictions that natural catastrophes around the world will become more frequent and more intense. And the industry is backing that view. "Over the coming years, with warmer sea surface temperatures making landfall more likely, particularly destructive storms are a likely scenario," said Lloyd's chairman Lord Levene during a speech in January on climate change. "We can expect the storm season to lengthen, and we will be at risk over a wider geographical area than ever before." Soaring insured values in catastrophe-prone coastal zones drive potential loss scenarios still higher. In the US, coastal values have doubled in the last decade to over $7trn.
The exit of some key retro players during the course of 2006 took even more capacity out of the market and kept prices sky-high. Some notable departures included PXRe, Royal Bank of Canada and White Mountains. Swiss Re's decision not to renew GE Frankona's book of retro business came as a blow for the market, but it was not unexpected. Unlike rival Munich Re, Swiss Re has historically shied away from providing or purchasing retrocession believing it is "too incestuous", according to board member Martin Albers.
But any dearth of capacity caused by retiring retrocessionaires was soon filled with opportunistic catastrophe writers, explains Edwards: "In 2006 perhaps 20% to 24% of traditional retro capacity exited the market. Having said that, 2006 was a year when many other reinsurance players' balance sheets expanded and new capital sources started to emerge." Describing the situation at 1 January 2007, Cantlay said it was a case of out with the old and in with the new: "We had less worldwide players who were willing to offer coverage on a worldwide basis, but then we had a fair amount of new entrants like Aeolus Re. Blue Ocean had more capacity to sell, Everest had more capacity to sell and more people were just generally looking."
For those reinsurers with hurricane exposures, the dearth of traditional retrocession capacity in 2006 forced many to consider alternatives and the capital markets became "a lender of last resort", according to Thoenes. Last year saw a boom in alternative risk transfer vehicles such as sidecars, catastrophe bonds and industry loss warranties (ILWs). These different vehicles all provide a means of transferring insurance risk to capital market investors, particularly hedge funds. According to Benfield, over $30bn of new capital has been raised since Hurricane Katrina in various forms. In 2006 over $3.6bn of equity and loan capital went into 18 sidecars, cat bond placements increased from $2.4bn in 2005 to $4bn in 2006 and industry loss warranties also took over $4bn of limit.
"It will be interesting to see what the long-term play will be for that capital," says Cantlay. He points out that the level of take-up of ILWs, index-based products which boomed last year as an alternative to retrocession, had begun to taper off at the 1 January renewals. Bennett believes it is likely that some of the increased capacity provided by capital market investors is opportunistic and will only remain available in the short term while rates hold up throughout 2007 and probably 2008. "Some of this new capital is likely to remain as a feature of the reinsurance/retro market over the long term," he added.
The arrival of over $30bn of fresh capital market capacity since Hurricane Katrina (including the amount raised through new startups and existing companies) wasn't quite enough. According to Benfield, despite the rapid growth of the cat bond and sidecar market, the amount of capital raised in 2006 represents only 3% of the estimated $330bn capital of the global reinsurance market. "Even with the influx of capacity from the capital markets, the demand for hurricane risk capacity still far exceeded supply," explains Thoenes. The other issue with capital market solutions, according to Edwards, is basis risk. "Cover based on indexes whether it is model loss or ILW or parametric triggers always give the buyers varying degrees of basis risk and this is innately unattractive," he explains. "Therefore there is a tendency to look towards traditional retrocession cover first, if it is available."
One tactic in 2007 for retro buyers who remain uncomfortable with the current pricing is greater risk retention. "High expected profits in 2006 and fresh capital looking for high-yield investment opportunities make this a veritable option," says Thoenes, predicting a number of reinsurers could look to increase their capital bases. "We expect a number of reinsurance companies to reduce their aggregates in a way that they ultimately could survive without retrocession." He believes the continued influx of fresh capacity and a growing reluctance of traditional buyers of retro to accept inflated prices could cause pressure in 2007 on prices to come down. According to Piers Cantlay, this has already had an impact at 1 January 2007. He thinks one reason rates for retro did not match the "hysterical pricing" witnessed at mid-year 2006 was down to buyer's tactics. "Most of the seasoned buyers of retro observed what happened in the middle of the year and recognised that some of the retro pricing that was purchased was uneconomic. So they had a plan that allowed them to either have more capital within their company or to reduce their original exposures - to not be so reliant on retro."
According to JLT Re's recent renewals report '2006: A year of plenty', market observers believe current levels of retro pricing are unlikely to be sustainable in the long term. It points out the coverage restrictions seen at the renewals, which are intended to make the programmes more attractive to reinsurers and suggests that late entry into the market by new retro writers may cause price softening during the year. "A benign season inevitably leads to some lessening in prices," says JLT Re's Bennett. But he thinks there will always be a place for traditional retrocession. "Demand is always there for the retro market. It's just a matter of price and so there is a delicate balance. Some people will have to buy it at any price but they might buy less."
For John Edwards, the popular assumption that the retro market disappears following a large catastrophe is a complete misnomer. "Our market exists because there are willing buyers and willing sellers and it performs a very necessary function for many reinsurers," he says. "The idea that market forces disappear as a result of a major loss is unfounded. In fact, it is a market that can very quickly attract new capacity." The key, he says, is that clients continue to feel that they're getting something of economic value.
- Helen Yates is editor of Global Reinsurance.
- Is the industry overcapitalised?
- Retro squeeze throwing market off balance
- Retro too incestuous for Swiss Re.