Insurance companies are uniquely exposed to these sorts (climate) of changes. Tens of millions of businesses buy policies every year to protect themselves from risks. As a result the industry is vast—last year the premiums paid for property and casualty insurance worldwide reached $2.4trn, according to Swiss Re, one of the big reinsurance firms on to which consumer-facing insurers pass the risk of mega-losses. Insurance companies spent $180bn on reinsurance premiums. Extreme events becoming the norm could force insurers to fork out ever greater payouts to policyholders, as well as lower the value of the assets they hold. The best case is that insurers reinvent themselves, helping the world cope—risk is, after all, how they make their money. The worst is that some fail and, more worryingly, that swathes of the global economy become uninsurable.
Already, insurers are seeing disasters of unprecedented scale. Earlier this month Hurricane Dorian, one of the two largest storms ever known to have made landfall in the Atlantic, battered the Bahamas and then the Carolinas. In July Hurricane Barry brought the heaviest rainfall ever measured to Arkansas. The Indian Ocean basin has seen three huge cyclones so far this year, one of which caused Mozambique’s severest flooding since 2000. Last November California saw wildfires over the largest area ever recorded.
Very costly disasters are becoming more frequent. Between 1980 and 2015 America saw an average of five events causing over $1bn in damage (in current prices) each year. Between 2016 and 2018 the yearly average was 15. In the 20th century, according to AIR Worldwide, a climate-modelling firm, a hurricane on the scale of Harvey, America’s costliest ever, would have been regarded as a one-in-2,000-year event. By 2017, when Harvey blew in, that frequency was estimated at one in 300 years. By 2100, says Peter Sousounis of AIR, it will be once a century, and tidal surges that used to be classed as once-a-millennium events will be expected to strike every 30 years.
Catastrophes are also getting harder to predict. Though newer models are starting to take account of climate change, most still rely on crunching numbers from the previous few decades, which are already obsolete. And insurers struggle to handle “compounding effects”—the mutually reinforcing impact on each other of events associated with global warming.
Adding to the losses is the growing number of properties being built in harm’s way, such as on flood plains and coasts. Average annual insured losses from catastrophic events have grown 20 times, adjusted for inflation, since the 1970s, to an average of $65bn this decade, with a peak of $143bn in 2017. That excludes knock-on effects such as business disruption. Last year the global figure totaled $85bn, the fourth-highest on record, even though it was a year without a mega-disaster.
Climate losses can also come from the other side of insurers’ balance-sheets: the investments they hold in order to cover potential payouts and park any spare funds. Insurers (including life as well as property and casualty) are the world’s second-largest institutional investors, with $25trn under management. They often place chunky bets on multinational firms, infrastructure and property—which are becoming riskier propositions as the climate changes. Moreover, structural changes in the economy, such as the move away from fossil fuels, could leave insurers’ portfolios exposed.
In the face of these threats, insurers are seeking to future-proof their businesses models. Part of this is about financial resilience. Most general policies are renewed annually, meaning firms can raise premiums promptly. And since a spate of mega-disasters caught them off-guard in the 1990s they have fortified their capital reserves. According to McKinsey, the policy holder surplus (crudely, the excess of assets over liabilities) available to pay claims in America’s property and casualty sector doubled in real terms over the last 20 years. In 1992 Hurricane Andrew sent 11 insurers to the wall. All survived the record hurricane season of 2017-18.
Regulators are doing more to prod insurers to hold sufficient capital—typically they aim to ensure they can withstand losses caused by the worst imaginable year in 200. But putting a figure on this is hard, because nobody has thousands of years of data. And the worst possible year is getting worse every year. The risks will keep rising long into the future, says Paul Fisher, a former supervisor at the Bank of England. A cataclysmic year could also hit markets, hurting insurers’ investments just when they need them most. Some could be forced to sell assets to cover giant payouts, pushing asset prices down further.
Most probably, payouts will continue to rise without capsizing insurers. But that still creates a problem. To absorb bigger losses, they must charge higher premiums. According to Marsh, a broker, global commercial-insurance prices rose by 6% in the second quarter of this year, compared with the previous quarter. That was the largest increase since records began. In America property rates jumped 10%; in the Pacific region they soared by nearly 18%. The rise is to meet the demands of reinsurers, which insure the insurers. Average reinsurance rates are set to rise by 5% next year, according to S&P Global, a rating agency—and in California, after the huge recent wildfires, by 30-70%.
A few calm quarters would probably see some of those increases unwound. But there is no doubt about the trend. And it cannot continue forever without causing at least some customers to rethink whether to buy insurance at all.
The global “protection gap” between total losses and insured losses is already wide and growing. The research arm of Swiss Re estimates that it more than doubled in real terms between 2000 and 2018, to $1.2trn. Half of last year’s losses from natural disasters were uninsured. Nine out of ten American homeowners have no flood insurance despite half of the population living near water, says Erwann Michel-Kerjan of McKinsey. The share of uninsured damage is especially high in developing countries, where infrastructure and risk-mitigation measures are not keeping pace with economic growth.
Where risks become uninsurable, states and firms may work hand-in-hand. In Britain, where a sixth of homes are at risk of flooding, government and insurers have set up Flood Re, a reinsurer of last resort that enables insurers to offer affordable premiums on 350,000 homes in flood plains. Every firm selling home insurance in the country pays into the scheme, spreading the costs across hundreds of thousands of policies. To avoid perverse incentives, houses built in high-risk areas after 2009 are excluded.
Developing countries are underinsured partly because the risks they face are poorly understood. And many are urbanizing fast, which means cities change from year to year, making the value at risk hard to track. More research would help deepen insurance markets, and making models publicly accessible would also enable officials and development financiers to evaluate mitigation measures. Above all, insurers need to take the lead in publicizing the growing risks posed by climate change, and the need for cover. According to Alison Martin of Zurich Insurance, often people do not take out insurance because they think the worst will not happen. Talking of one-in-2,000-year events is not very helpful, she says, “because many people would think we’re safe for another 1,999”.