Though severe enough to impact earnings, the flooding in Texas is unlikely to hurt property & casualty insurers badly enough to force rates higher.
At first glance, record rainfall from Hurricane Harvey and images of Texas streets resembling rivers suggest the damage could swamp the insurance industry. Yet current estimates for Harvey call for total insured losses of about $10 to $20 billion, well below the totals from Hurricane Katrina in 2005 and potentially comparable to 2012’s Superstorm Sandy.
Given that $65 billion of insured losses from 2005’s trio of hurricanes (Katrina, Rita, Wilma) were not enough to create sustained pricing power for property & casualty (P&C) insurers, we do not believe Harvey will create a “hard market” for P&C companies that would enable them to charge higher premiums.
As estimates are published, it’s important to separate total insurable losses from flood-related losses. Press reports indicate insured losses could reach $40 to $50 billion though approximately 60% of total insurable losses will likely be flood. Flood is primarily a liability for the Federal Government’s National Flood Insurance Program (NFIP).
Half a million cars "at risk"
Auto insurers are one area where P&C insurers are exposed to flood, assuming the policyholder holds “comprehensive” coverage. About 75% of all auto plans in the U.S. have “comprehensive” coverage. With an estimated 500,000 cars potentially “at risk,” Harvey’s estimated $3 to $6 billion loss for auto insurers could prove the largest event on record for auto insurers. However, we would be more constructive on Harvey’s impact to auto insurance pricing if the industry wasn’t already in a very hard market (due to rising loss costs from distracted driving). That said, we prefer insurers with auto exposure because we expect a long period of improving margins for auto insurance after the Harvey losses pass.
Even at the upper estimates of insured losses ($20 billion), Harvey is not a huge event for P&C insurers. Indeed, even the trio of 2005 hurricanes (Katrina, Rita, and Wilma) and their $65 billion of industry-wide losses were not enough to create sustained pricing power.
To estimate losses by company in the Texas marketplace, we make an assumption of how insurance industry losses will be spread based upon historical experience (30% homeowners, 45% commercial P&C, 25% auto) and apply those to various industry-wide loss scenarios based on each company’s share. Our conclusion is that even the most exposed companies, from a market share perspective, are unlikely to lose money in the third quarter as a result of this storm.
In sum, Harvey remains an “earnings event” but not an event that destroys capital for the industry.
But we're barely into hurricane season....
The peak of hurricane season is mid-August thru mid-October. As such, the possibility of further major hurricanes is real. Katrina, Rita, and Wilma (KRW), for example, occurred in August, September and October 2005, respectively. Additional hurricanes making landfall in the U.S. could change our view on P&C insurer pricing trends.
However, we believe insured losses from hurricanes in 2017 must exceed significantly the $65 billion of KRW losses to make a sustained impact on pricing. The industry is in a dramatically different position today due to a “flood” of capital willing to invest in this relatively uncorrelated asset class. Investing in non-correlated products such as catastrophe bonds is now a well-accepted form of diversification for pension funds, hedge funds and insurance companies. One landfall hurricane in nine years is not going to change this concept. Capital needs to be destroyed and afraid to re-enter the marketplace before primary P&C or reinsurance pricing is going to “harden.”
We believe it would take a “clash” event to create a hard market. This is an event that cannot be modeled. Such an event could drive fear in the marketplace about unforeseen or underpriced risks and drive out alternative capital. Events like 9/11 or the Fukushima Daiichi nuclear disaster qualify as “clash” events. It would take a huge amount of “model-able” insured losses - in excess of KRW’s $65 billion - before we would view alternative capital providers as being scared away from the marketplace.