Take a US $8 billion dollar insurance business that serves more than 7 million customers annually regarding residential real estate assets valued at more than US $5 trillion, that’s working in a regulated, politically hyped environment and one might see opportunity for innovation. Throw in significant exposure to regional wildfires and urban area earthquakes, and consider- where does InsurTech go next in the state of California? Or any other disaster-prone area?
Patrick Kelahan is a CX, engineering & insurance professional, working with Insurers, Attorneys & Owners. He also serves the insurance and Fintech world as the ‘Insurance Elephant’.
Clearly there is evidence that natural disasters’ frequency and severity is growing, whether hurricane, typhoon, flooding, or wildfire. Throw in potential earthquake occurrences and weather-related agriculture losses and the need for the insurance industry (primary and reinsurance) to be attentive and apply a sense of urgency is heightened.
The California wildfires are an apt study for how InsurTech/insurance players and innovation efforts are or can react to the industry’s needs- property owners, carriers, agents, adjusters, regulators, repair industry, response industry, and politicians. Not through individual efforts to innovate pieces of the industry, but macro approaches to managing before, during, and after efforts.
To date there have been many industry reactions to the wildfires Californians have experienced during the past two years, even the inclusion of at least one property carrier that has failed due to unexpected volume and severity of homeowners claims. Setting aside the tragic loss of life, one can say that every other aspect of risk sharing and management has been a victim of the depth and breadth of the fires. This suggests that each aspect is an opportunity for review and potential revision of how disasters are planned for and responded to, and how (or if) the insurance industry can integrate disaster planning into regular business.
Underwriting and Recovery
Wildfires pose an interesting challenge to traditional underwriting as there is a higher percentage of probable maximum losses within wildfire areas, and wildfire zones are growing in projected size and frequency of being affected. Hazard Hub (industry experts at scoring physical risks within the US) indicate that 16% of California locations are rated ‘D’ (high) risk score or higher. That’s 1.3 million dwellings in the state with that level of wildfire risk. During the past two years (2017-18) it is estimated that 30,000 structures have been lost to wildfires, with many of the 30,000 being dwellings.
The past two reporting years have also shown homeowners insurance to be an ‘upside down’ business, with incurred losses for multi-peril and fire policies equaling $29 billion, and earned premiums during the two year period of $15.7 billion, a loss ratio for the period of 191.2 %. Yes, there were many prior years where LR were lower and it could be said that surpluses were acquired, but considering the average HO policy premium in the state runs around $1200 and the median value for houses in the state is just over $500,000, it’s easy math to run comparing $8 billion annual premium on a housing stock of $4 trillion in value. And- if wildfires continue to occur close to San Francisco and Los Angeles, the relative value of potential losses increases and perhaps appetites to remain a homeowner line carrier begin to wane.
How can innovation/technology step in to adapt underwriting to these imbalances?
The easy answer is to raise deductibles (like EQ or windstorm), but that doesn’t eliminate the outcomes of the high probabilities of maximum losses, just reduces the dollar effect a little. Can high and extreme fire risks become unwritable for carriers? If an ‘F’ risk per Hazard Hub suggests a one in fifteen probability of wildfire risk, then a carrier can calculate that risk by property into a premium, and there will be some premium shock. So yes, but regulators may push back. Can other perils become policy victims in order to rescue fire, e.g., convert water or other perils to parametric outcomes to provide a sublimit position for that peril- more predictable financial outcome and potentially less proposed premium. The depth and breadth of claim and risk data could surely produce underwriting rules to reflect the new risk paradigms in disaster prone areas.
The California legislature has been active in enacting laws that increase policyholder policy benefits due to the fires, including extension of statutory lawsuit periods, enhanced periods for Additional Living Expenses (ALE), and reducing the burden of personal property inventory preparation when a total loss occurs, all reactions rather than ‘pro-actions’. Encouragement of disaster-response industry sandboxes might be a proactive option to consider, unless the state thinks this year is the last for occurrences.
Municipally sponsored parametric benefits can be devised to assist when mandatory evacuations are ordered, a benefit to all property owners before any claim is filed. This is an easier tech adaptation (just ask the good folks at Jumpstart) that would have immediate effect at the onset of a regional fire, and would serve to preempt the requirement that an emergency or disaster be declared. Classic parametric trigger- mandatory evacuation, and tech exists now that could help development of an action database, or even (shudder) blockchain handling of address data and payment actions.
ALE services such as Temporary Housing Directory (THD) could open its database to evacuees to facilitate relocation- temporary or post-loss- or dovetail its services with those of forward-thinking carriers such as Hippo Insurance. When Hippo reaches out proactively the message could include evacuation routes, potential lodging options, connection with emergency authorities to communicate evacuation progress, and so on. These are tech advances that might not generate premiums but would surely lead to retention.
Recovery and Rebuilding
The very fact that laws had to be enacted to change policy benefits from one or two years to three suggests that wildfire recovery is daunting. Consider the city of Paradise, where little rebuilding has occurred even after a full year. Contamination, consternation, and coordination issues abound.
Tech advances in property location, ownership information (reduce the reliance on ledger books at the town hall), automated code compliance, virtual transmission/approval of building plans, virtual confirmation of insurance for the owner and contractor, and overlay maps of utilities (buried and other) would set the foundation of recovery. Mortgagee resolutions based on immediate virtual dwelling insurance settlements would remove the paper chase for mortgage releases.
Rebuilding brings a new set of concerns with shortage of contractors, materials and oversight. A clear recognition of rebuild cost spikes could be adapted by primary estimating platforms like Xactimate and Symbility Mobile Claims to reflect real-time materials and labor costs and reduce the urge to pad projects. Confirmation of contractor licensing through automated means, online data for total loss dwelling calculations, and use of drone surveys of damage areas to confirm scope of loss (this is happening now) would remove many admin barriers to recovery. Hardware/firmware innovations can make online an expectation rather than an exception after a major event. What if major recovery efforts simply did not need cohorts of catastrophe adjusters to be on site, but allowed virtual handling of much of the work? These improvements exist now, would simply need to be scaled and coordinated for macro use.
And worst case scenario- if rebuilding is simply not a practical option due to site or risk issues, identification of non-viable areas can be handed off to FEMA for potential property buyouts without the years of proving each case.
Reinsurance and Backstopping
If severities continue at the current pace the unknown for reinsurers will be reduced- the treaty thresholds will be expected to be met or simply be moved higher, and rates will harden. Of course primary carriers will have few options since gross severities cannot be expected to fall. Rei price then rolls into premiums, and the process repeats until the next occurrence. ILS vehicles could evolve and be parametric program partners (say that three times), with a reduction of some unknowns for those investors. Sure that would raise the cost of ILS or cat bonding but would be more predictable than current handling. Ironically as major events occur the nature of associated risk data would become more robust and predictions more fine tuned, and reinsurance triggers easier to gauge.
These are some thoughts focused on the California wildfires, but thoughts that could in most instances adapt to other natural disasters in any jurisdiction. InsurTech resources have been searching for collaboration partners- disasters are the testing grounds for pretty much anything InsurTech wants to do. There may not be a financial quid pro quo from coordinating one’s efforts with others in a disaster scenario, but it’s certain- any efforts taken to improve how the industry and all its partners plan and react to significant events will cause improvements that can be implemented in mainstream insurance, and disaster efforts help restore local economies where all insurers want to be.
If the reader thinks this message is focused on US disaster prep/recovery only- the same principles apply to plan for wildfires outside of Athens, flooding in Tewkesbury, typhoons in Honk Kong or Japan, earthquakes in Indonesia, flooding in Brazil, hurricanes/landslides in Central America, tsunamis in Chile, typhoon flooding in Mozambique, or monsoon effects in Bangladesh. Planning, coordination, reaction and risk management.
And as all the disruptors can tell you- that’s where the real money (and benefit) is.