A guest post in two parts by Paul J. Davies, the FT’s insurance correspondent. This is part two.
So, to the really interesting question, how long will any upswing last and what might it look like? Here’s a chart from Guy Carpenter, Marsh’s reinsurance arm, that shows the impact of previous events on global catastrophe reinsurance rates. It is indexed and based on 100 at 1990:
The spikes follow Hurricane Andrew, which hit the US in 1991; the destruction of the World Trade Centre in New York on 9/11; Hurricane Katrina, which hit the US along with Rita and Wilma in 2005; and Hurricanes Ike and Gustav in 2008.
For Andrew, which cost about $25bn in today’s money, the spike is so high because this was the first loss of modern times – ie when the US was already extremely wealthy and levels of purchased insurance cover were generally high.
At the time, the insurance industry had none of the modern loss modelling techniques, while many companies were not even sure where they had written cover – the policy for a Florida holiday home, for example, was often sat in the files as registered to the policy holder’s home address, which could be way inland. The storm was a huge surprise and risks were poorly understood.
Since then a number of things have happened. US wind is much better modelled and the channels for new capital to enter the industry (and take advantage of higher insurance rates) have become much more varied and efficient.
Bermuda has become home to lots of ‘cat’ focused insurers – and its regulators still boast it is the fastest place any capital can come to set up a business. On top of this, the capital markets have created catastrophe bonds, which take extreme cat risk; side cars, a kind of segregated account for hedge funds and private equity to invest alongside good cat underwriters; and Industry Loss Warrants and other OTC or exchange traded cat derivatives.
After 2001, markets were already under great stress due to the dotcom crash and remained so due to the coming Iraq invasion – stocks only started to pick up in March 2003 once Sadam Hussein was toppled. Many of the capital market instruments noted above were also only just being developed.
This slowed the capital response to insurance prices post 9/11 and hence prices only slowed, topped off and began to decline (see chart above).
Come Katrina in 2005 and we’re in a completely different world. It was the biggest single insurance loss ever seen, yet the initial spike was relatively modest, then the drop was sharp. According to Guy Carp, more than $35bn of fresh capital came into the 2006 cat renewal season.
In 2008, Ike and Gustav helped create the third biggest loss year ever for the reinsurers – and bang in the middle of the financial crisis. However, the rebound in markets — particularly corporate bonds, in which reinsurers were much more heavily invested than in equities — meant that capital positions were anyway pretty quickly rebuilt. Hence there was even less of a post-event boom.
So what does this all tell us? At its worst outcome, the Japanese quake could have a big impact on pricing in the market in the short term – particularly because storm season in the US does not even get underway until late summer.
(As an aside, whether it’s down to climate change or not, US hurricanes are currently going through a period of peak activity, the highest since the 1940s and 50s. Last year brought the third highest number of named storms on record and it was little short of a miracle that none of the major ones made landfall.)
However, the industry is currently still significantly over-capitalised — by $19bn for reinsurers alone, according to Guy Carp. On top of which, the channels to plough money back into reinsurance risk are probably about as efficient as they will ever be – and insurance risk is seen as the last great uncorrelated play at a time when the great credit bubble and consequent meltdown put everything else in lock-step.
Markets are depressed, but past the worst and there is very little yield elsewhere. This is very likely to mean that any great uptick in reinsurance pricing either later this year or next is likely to remain short lived – a bad thing for listed reinsurers especially.
Unless, of course, we discover that there is some link between the recent quakes and volcano activity around the pacific – Indonesia, New Zealand, Japan…
What Japan’s earthquake could mean for reinsurers – part one – FT Alphaville