The back-to-back devastating California wildfires of 2017-2018 caught the property-casualty re/insurance sector by surprise with the intensity and frequency of the losses and challenging the sector's understanding of this hazard. Nevertheless, in view of most re/insurers' robust capitalization, these wildfires in conjunction with other catastrophe losses had limited impact on their creditworthiness.

Historically, the re/insurance sector has mostly focused on the primary perils such as U.S. hurricanes, tornadoes, and earthquakes, which in the past have been major causes of property-catastrophe risk and losses. The events of 2017-2018 highlighted the increasing risk from secondary perils such as California wildfires, which have increased in frequency and severity. Eight of the most destructive fires occurred in the past two years, and five of the seven largest fires and 10 of the top 20 most destructive fires occurred after 2009. However, it took the events of 2017-2018 for the industry to start paying the kind of attention this peril deserves.

The modeling for California wildfires has been challenged by a number of factors. Climate change is one but not the only factor contributing to the increase in risk, with increasing frequency and severity of dry weather and extended droughts heightening the risk of wildfires. In addition, the level of urbanization, and population and economic asset density, which are close to or encroaching on the wildlands (commonly referred to as wildland-urban interface [WUI]), have been growing, which makes for a catastrophic event when these high-density areas, potentially with expensive properties, are hit. The recent updates to the model targeted a higher level of sophistication for the primary causes of wildfires, resulting in higher frequency and severity of estimated losses. However, challenges persist in understanding this type of peril.

With back-to-back above-average catastrophe years, reinsurance and alternate capital are smarting from the losses--especially those from the California wildfires. Several re/insurers are not comfortable with their understanding of the risk, which has led to the sector taking a cautious approach, with many curtailing or stopping their underwriting. This has constrained the available capacity for this risk. Furthermore, retrocession capacity, which in the past provided a cheaper form of capital and enabled the players to pass on the property-catastrophe risk in general, is constrained as well. Whatever capacity is available is much more expensive after two subsequent years of double-digit rate increases. S&P Global Ratings expects significant rate increases for wildfire reinsurance between now and next year's renewal seasons. This may not be as apparent because this peril is usually combined with the other reinsurance coverage for primary perils; hence, the impact of pricing changes for wildfires gets somewhat lost in the aggregated pricing. We expect this dynamic to influence the primary pricing as well, though more so in commercial lines than in personal lines.

2017's Losses Were Significant And Widespread

The resultant losses from two years of back-to-back wildfires were widely spread across the re/insurance sector. Most of the insured losses were paid by five to six large national primary insurance companies with substantial homeowners business. This risk was well-insured on both the personal and commercial sides, with the majority of homeowners covered by their insurance policies. Insurers in turn used reinsurance and capital markets to mitigate the risk, which ended up propagating the losses throughout the value chain. As the severity of events became clearer and as the losses spread beyond just property-catastrophe, they caused turmoil in the broader re/insurance market and not just in the alternative capital space.

The 2017 California wildfires were a major event and although hits to insurers and reinsurers were expected given loss levels, they also affected alternative capital providing retrocession covers. Although wildfires are considered a secondary peril and modeling is not as developed as that for, say, U.S. hurricane and tornado exposures, the property-catastrophe funds were taking a lot more wildfire risk than perhaps they realized or priced for, likely because wildfire exposures are usually part of the property-catastrophe or aggregate covers. The losses from 2017 events were deep in the tail based on the market understanding of the risk at that time, so the sector largely considered it a low-probability occurrence.

And In 2018, They Were Even More So

However, after the 2018 wildfires, not only alternative capital but traditional re/insurers started questioning the understanding of this peril and their ability to appropriately price it. The losses stemmed not just from the property-catastrophe risk business but also from the casualty lines, including utilities, which went counter to re/insurers' expectations. In California, a utility company is responsible for paying property damages from wildfires linked to the company's equipment, and does not necessarily have to be found negligent to be held liable for the losses. With aging infrastructure and increasingly hot and dry weather, the risks have grown significantly in recent years. A recent example is Pacific Gas and Electric Co., which had to file for bankruptcy because of the extent of liabilities from the wildfires. According to the utility company, its liabilities could exceed $30 billion.

Models Let Re/Insurers Down

The first generation of California wildfire models came out in the early 2000s and were refined over the years; however, they fell well short of the loss experience in 2017-2018.

A 30-40-year historical data set suggests increasing trendlines for wildfires and weather data, particularly for large fires. The scale of wildfires continues to expand, with an increase in WUI interface and density, natural drought cycles, climate change, forest management, and ignition sources all contributing to increased wildfire risk. The 2017-2018 wildfires didn't happen in the main fire seasons for both Northern and Southern California. The notable high-damage fires happened outside of the peak fire count or area burned periods, when winds are usually subdued. Recent extended drought conditions have reduced the wet season, extending the dry period into periods with higher wind. These conditions exposed unrecognized risks and higher severity for fires, with wind getting recognized as a major hazard for severity.

According to a Swiss Re Sigma report, 2018 was the most deadly and destructive wildfire season in California, with record insurance losses, followed by 2017. The loss events provided some important lessons necessitating a revamp of the pre-existing models. Wildfire models typically involve four factors: ignition, spread, suppression, and damage. All of these factors were revised to take into account the recent lessons, human activity, community influence, and additional forces for the spread of wildfire footprint. Although vendor models differ in terms of event frequency, size, severity, and other factors, the outcome of these updated models is ultimately higher estimated losses (both frequency and severity) than predicted by the pre-existing ones.

Reinsurance Pricing: Renewal Timing Was Off

Reinsurance pricing for wildfire risk did move in the past six to 12 months, but not as forcefully, as the sector appeared to assume 2017 was a one-off event. In addition, this risk is not priced on a stand-alone basis for the most part; rather, it is combined with other treaties including property-catastrophe and aggregate covers. Reinsurance pricing for wildfire risk moved by double digits in 2018, more so for commercial lines business, primarily utility companies. 2019 reinsurance renewals for treaties covering California wildfires didn't provide any encouraging news from a pricing perspective, despite the second year of extensive wildfire insured losses. This is because many insurers buy multiyear treaties, so not all coverage is renewed at the same time. Furthermore, a few of these treaties are renewed about six months in advance of the January renewal season, which for 2018 also pre-dated the California wildfires in the second half of that year. As a result, despite the 2017-2018 losses, 2019 reinsurance coverage was already in place, and hence, the pricing for this risk didn't move as much in January renewals of this year.

The Insurability Of Risk Is Being Questioned

There is no consensus in the re/insurance sector on the insurability of wildfire risk. Reinsurers' comfort level with the updated models for this hazard is not that high and not consistent. While some reinsurers have become comfortable with the updated view of the risk and are willing to underwrite it--albeit at higher prices, others have either withdrawn from the risk entirely or have cut back significantly. In addition, the retrocession capacity is just not available to the same extent as in previous years, more so for the aggregate covers, and whatever capacity is available is at much higher prices.

This dynamic extends beyond reinsurers; primary insurance companies are struggling not just with understanding the risk but also with their ability to raise rates, especially on personal lines. Proposition 103, passed by California voters in 1988, mandates insurance companies to require "prior approval" from the California Department of Insurance before implementing property-casualty insurance rates. Any proposed increase of 7% or greater, regardless of rate indication, must go through a resource-draining process that many insurers try to avoid. Furthermore, the Department of Insurance does not permit the use of catastrophe modeling in premium rate filings. This frustrates the primary insurers in achieving the level of rates they want, as the pricing wasn't adequate to begin with considering modeling deficiencies. In addition, the transfer of risk to reinsurers and capital markets (through insurance linked notes), which had helped in the past, won't be cheap. Stuck between limited flexibility on primary rates and the rising cost of reinsurance, insurers are increasingly staying clear of this risk wherever feasible.

A Hefty Rate Increase Is In The Offing

With reinsurers in slight disarray and given their lack of comfort with the California wildfire risk, pricing will inevitably increase. Reinsurance pricing could rise 30%-70% between now and the January 2020 renewals in view of higher expected losses under the updated models. And given inherent difficulties with the modeling, we expect a healthy risk margin to be built into those rate increases, given the uncertainties involved. We also expect tightening terms and conditions, with reinsurers pushing to make the definitions for loss occurrence narrower; currently loss occurrence can have different insurer interpretations. With California approving a $21 billion fund to cover the cost of wildfires for the utilities, reinsurers may see a piece of that business, though details aren't clear yet.

We expect primary insurance rates will rise as well. For the most part, the California market is served by large carriers, mostly nationals. Considering their large, and diversified, books, the impact from wildfires on the national business was not as severe. While primary carriers would like to take higher rates, considering the importance of the California market, insurers will keep writing the risk, although they will be more selective in certain fire lines, leverage the residual market to maintain a presence in the state, and work on the rates over time. At the same time, primary insurers will raise rates as much as they can in the commercial lines market or for high-value homes, constrain their capacity if they can't get the rates, or withdraw from the risk entirely. High-risk policies either end up with California Fair Plan Property Insurance (FAIR Plan; association property insurers underwriting in California) or potentially make their way to the excess and surplus (E&S) lines market. However, there are disadvantages to both: the FAIR Plan can constrain an open insurance market and it's unclear how much volume it can handle; whereas the E&S market is pushing back as it is not set up to deal with this type of business or policies through the wholesale distribution channel. In addition, Lloyd's is an E&S player in California, but is reducing its capacity due to the loss experience.

Re/Insurers Tread Carefully As They Reassess Their Risk Appetite

Re/insurers are in the business of risk-taking and these risks can be indeterminate given the nature of the business. Nevertheless, we expect re/insurers to take a disciplined and measured approach considering the risk-reward trade-offs and to use their well-developed risk management practices to mitigate the risk, including development of risk measurement techniques, models, and tools to manage risk from secondary perils both from a frequency and severity perspective. The past two years have clearly highlighted that these secondary risks are not to be taken lightly. Indeed, reinsurers have reassessed their risk appetites in view of recent experiences. Considering the limitations of the wildfire catastrophe models, if re/insurers were to underestimate this risk, they may end up taking outsize exposures that could result in a capital event and ultimately hurt their credit worthiness.

GRH2019