”What about climate change?”
Following the worst year and a half ever seen for Insurance-Linked securities (ILS) and cat bonds, some investors naturally are asking if the market underestimates the risk associated with cat bonds concerning the effects of climate change. Yet, the dominating risks covered by cat bonds in the timeframe concerned do not primarily come from climate change, but from demographics, which are far easier to analyse with high accuracy. The departure of more opportunistic investors and the growth of mature investors will benefit the reinsurance and ILS sectors by limiting the supply of opportunistic capital and increasing premiums.
Human behaviour is the most important short-term driver
Climate change will, no doubt, have a profound impact on the world as we know it. Concerning natural catastrophes, the number, strength and longevity of major hurricanes is expected to increase, though the total number of hurricanes may remain largely unchanged. It is highly likely that wildfires will be more common and more severe as the atmospheric temperature increases. The frequency of extreme events, such as drought or precipitation is set to increase in the future.
Yet these events, from an insurance point of view, are dwarfed by more immediate trends affecting the insurance market. In the past 40 years (1976-2016), the number of Americans exposed to hurricane risks increased by 60 million, not because hurricanes were more severe or affected new areas, but because more people have settled in coastal regions. Increasing wealth and property values and the desire to live in coastal areas preferences are the most important drivers of increased insured hurricane losses (Hurricanes are doing more damage and it’s our own fault, Popular Science, Nov 27 2018).
Largely, the same line of argument is applicable to US wildfires. While the areas burned by wildfire have shown an increasing trend, probably ascribable to a warmer climate, a more important driver for insurance risk is the growth of “the wildland-urban interface” (WUI), i.e. areas where wildland meets housing. According to a recent study, the WUI increased both in terms of the number of houses (from 30.8 to 43.4 million, 41% growth) and in terms of land area (from 581,000 to 770,000 km2, 33% growth). Hence, smaller fires in the WUI can be much costlier than vastly larger fires in uninhabited areas. Jonathan Kinghorn (AIR Worldwide) expressed this in a blog post earlier this year as “With Wildfire, Big Doesn’t Necessarily Mean Costly”.
Even with good risk modelling, insurance is all about probabilities
Looking from an ILS investor’s perspective, this means that the most important drivers of insurance risks are factors that can be quantified and calculated with high precision. Hence, short-term risk is a far more exact science than the long-term effects of climate change on insurance. Given the relatively short term of ILS instruments of typically 1 – 3 years, the pricing of risk will reflect current scientific knowledge of climate and weather-related risks.
Insurance modelling has been continuously and rapidly developing since the launch of AIR Worldwide’s first hurricane model for use by the insurance industry in 1987. Modelling is today at a granular level, where, for example, the properties of individual structures are considered together with recent scientific findings on from all areas concerned. Entropics recently installed the latest tool released by AIR this year for modelling wildfire risks in the US.
The increasing confidence in modelling and in the asset class as such is also reflected in the fact that fewer bonds are rated by independent institutes. For liquid cat bonds, there is also independent risk modelling for each issuance, providing transparency and objectivity, in addition to managers’ risk modelling of individual securities and portfolios.
Even so insurance risk remains a multi-faceted science. Prior to 2017, we saw an eleven-year period (2006-2016) without a single landfalling major hurricane in the US, and hence moderate insured losses. This changed abruptly with the 2017 season when two major hurricanes (Harvey and Irma) made landfall in mainland US and a third (Maria) made landfall in Puerto Rico. A record-breaking wildfire season further eroded the retention amounts of some cat bonds. The settlement of claims concerning these severe events has continued throughout 2018. In addition, California wildfires in 2018 have affected several positions on the broader ILS market, but also one cat bond position in Entropics’ portfolio.
In a sense, this is not surprising. Natural catastrophes are stochastic events and will not occur regularly. We will continue to see both years without any losses in ILS markets and years with high losses, like the situation this year and last year. The ability to calculate the probability and insured losses of an event does not include the ability to determine when these events will occur.
Decreasing supply of opportunistic capital can benefit long-term investors
Looking at the wider market perspective, after the historically costliest catastrophe year ever for re/insurance and ILS investors in 2017 it appears that more opportunistic investors trying to time investments to maximize short-term gain could be leaving the market. On the other hand, long-term investors who really understand the asset class will even more dominate the future ILS market.
A more mature investor base, entails a considerably higher pricing discipline of many reinsurance contracts on January 1.