はじめに
インフレの上昇は、保険業界だけでなく、より広範な経済において懸念が高まりつつあります。PEW Research Centerによると、米国における年間インフレ率は11月に6.8%に上昇し、約40年来の高率となりました1。すでに「増加する建設労務費および原材料費による保険負担」で確認したとおり、損害保険会社はサプライ価格を増加させるサプライチェーンの問題と、利用できる建設会社および熟練労働力の不足を生む建設需要の高まりという二重の問題に直面しています。
損害保険会社は、過去にもこのような状況を見てきましたが、あらゆる損害種類や国全体にわたって見られるものではなく、大型の自然災害から発生した地域的な需要急増によるものでした。ハリケーンや地震、またその他の災害がある地域を襲うと、少なくとも短期的に増加した建設需要を満たすためのサプライ品や建設会社の不足により再建が阻まれることが多々あります。こうした状況に対して、個別の支払備金の設定や将来のレート評価において保険会社がどのように対応するのかという前例はあります。こうした前例は、保険市場がパンデミックによる長引くインフレ圧力にどのように対応していくのかに関する視座を与えるかもしれません。
一方、現在のパンデミック由来のインフレによって既に疲弊している地域で自然災害が発生した場合、どうなるでしょうか。大寒波Uri（2021年2月）、ハリケーンIda（2021年8-9月）、また2021年12月のトルネード等は、今回の事象に対してリアルタイムのケーススタディとなります。保険金支払い増加のプレッシャーを既に感じている保険会社は、今年発生する様々な災害により、さらに資金が圧迫されることになり、今後の責任準備金の積立や損害保険料率に潜在的により大きな影響をもたらします。
Introduction
Higher inflation is becoming an increasing concern across not only the insurance industry, but the broader economy as well. According to the PEW Research Center, the annual rate of inflation in the United States rose to 6.8% in November, which is the highest rate in almost four decades.1 We have already seen the impacts of increasing construction costs on property insurance: Property insurers face a twofold challenge, with supply chain issues increasing the cost of supplies and higher demand for construction creating shortages in available contractors and skilled labor.
Property insurers have seen this kind of situation before—just not usually across all sources of property loss, and across the entire country. Instead, they are used to localized demand surges resulting from major natural catastrophes. When a hurricane, earthquake, or other disaster hits a region, the rebuilding efforts are often hampered by a lack of supplies and contractors to meet the increased need for construction, at least in the short term. There are precedents for how insurers respond in these situations, with regard to setting individual claim reserves and assessing future rates. These precedents might also provide a view into how the insurance market might react to sustained inflationary pressures from the pandemic.
But what if a natural disaster occurs in an area that is already suffering from today’s pandemic-driven inflation? Events like Winter Storm Uri (February 2021), Hurricane Ida (August/September 2021), and the December 2021 tornados provide real-time case studies into this phenomenon. Insurers that are already feeling the pressure of higher claim severities will come under additional strains from this year’s disasters, with potentially sizable impacts on reserving and property rates in the years to come.
Impact of catastrophes on claim costs – demand surge
Photos of the aftermath of a major hurricane can be as compelling as they are devastating. Images show rows upon rows of flooded or collapsed houses, sometimes also featuring local residents trying to pull items out of the debris. At a glance, the magnitude of the rebuilding effort becomes apparent, adding an economic toll to the human toll that disasters inflict.
In the early 1990s,2 the concept of demand surge, the increase in rebuilding costs caused by the magnitude of a natural disaster in comparison to normal times, began to emerge in insurance literature. The phenomenon itself is not new, with examples dating back well over a hundred years to events such as the 1906 San Francisco earthquake.3
However, quantification of demand surge was relatively new in the 1990s, and events such as Hurricane Andrew (1992) and the Northridge earthquake (1994) created a field of study that has continued to develop through to the present day. A general industry benchmark for demand surge is to expect an increase in costs of between 20% and 30% after a disaster. But these estimates can vary widely by event, by material, and by location.4 What are the major drivers of demand surge, and how might they relate to the current pandemic-driven situation?
- Size of event and amount of work required: The most basic driver of demand surge is the sheer number of property repairs (or full rebuilds) that are required in a given area after a major disaster. In general, the larger the event, the more demand stress exists.
Larger events can be defined by either or both of two characteristics. They can be broader and impact a vast geographic area, leading to a large number of filed claims. Or they can be more severe and cause more complex and more costly claims. Both features contribute to the overall challenges a catastrophic event poses to the insurance claims process. - Lack of materials and supply shortages: Natural catastrophes can also put significant strains on infrastructure, obstructing the supply chain needed to get materials to the impacted area and causing prices to rise. Alternatively, the general economic environment or other market factors (e.g., a very busy construction period elsewhere) can also lead to shortages in available materials.
- Lack of providers and contractor shortages: A natural disaster will often cause contractors to be drawn to an area from adjoining states or regions, but the supply of contractors willing to travel or relocate is generally not elastic enough to keep up with the increased demand, causing labor prices to rise as well.
- Lack of claims adjusters: An insurance-specific item, catastrophes will also put significant strains on insurance claims staff. Property adjusters will not only receive a much larger caseload than they are accustomed to, but they will likely see claims that are more severe and more complex than standard claims. This has led to the rise of specialized catastrophe claims staff, as well as insurers looking to bring in claim adjusters from other locations or hire third-party adjusters. Each of these factors raises the cost of handling claims. It also tends to increase delays in the claims process, which in turn are also associated with increased claim costs over time.
The factors above may last for different periods of time—material shortages tend to be resolved relatively quickly once transportation is restored to an area, whereas shortages in contractors or claims personnel may have longer-lasting impacts.
To evaluate the impact of demand surge, Figure 1 compares the development at similar ages of incurred loss for a sample hurricane to the development of incurred loss for a cohort of standard homeowners property claims. We can see that the hurricane’s incurred losses and allocated loss adjustment expenses (ALAE) take a longer time to develop to ultimate as a result of the same factors listed above that also impact demand surge.
Figure 1: Comparison of development patterns - Homeowners and hurricane (incurred loss and ALAE)
Similarly, the emergence of paid loss is slower for our sample hurricane than for the standard homeowners property claims, particularly at earlier ages, as shown in Figure 2.
Figure 2: Comparison of development patterns - Homeowners and hurricane (paid loss and ALAE)
Additionally, we can better understand disaster-related slowdowns by studying the Construction Backlog Indicator (CBI), a forward-looking indicator measuring the amount of work contractors are expected to perform in the upcoming months.5 Over the last 10 years, we have tracked the change in CBI in the applicable region over the year following several major disasters, comparing this change to the comparable change in score in unimpacted regions of the country. While the scores are not solely impacted by natural disasters, we do see some evidence of correlation between a series of major natural disasters and an increase in backlogs over the following year. For example, the change in CBI in the Northeast following Hurricane Sandy was an increase of 1.2 months whereas the change in all other regions was 0.1 months. For each of the natural disasters shown in the table in Figure 3, the change in CBI in the impacted region was larger than the change in CBI in unimpacted regions, with an average difference of almost one month.
Figure 3: Comparative change in CBI (months)
Overall, 2021 CBI is up over 2020 as well, with the total CBI in November 2021 at 8.4, relative to the November 2020 level of 7.2. Increases are seen in all regions, with the South and Middle States showing the largest increase in the year.
Insurers’ responses to major natural disasters
Through their experience with past events, insurers have developed various ways to respond to and even mitigate the potential impacts of demand surge. We can review these techniques and responses to consider which are likely to be most impacted by the current pandemic inflationary environment.
Mitigation often starts in the immediate aftermath of an event, even before claims are filed. Insurance companies coordinate rapid-response triaging such as sending in crews to help tie down or otherwise take immediate remediation measures to properties (e.g., an actively leaking roof). The goal of these mitigation efforts is to prevent further damage by ongoing adverse conditions. In fact, these efforts can be particularly important when an area is hit by multiple major events within a short period of time. In 2020, for example, Louisiana was impacted by three significant hurricanes (Laura, Delta, and Zeta) within about a month's span. Multiple catastrophes can also lead to coverage concerns regarding which event actually caused the most material damage to a structure.
Insurers also respond to potential demand surge by making changes to their claims handling and reserving processes. Some of these measures were discussed previously (e.g., hiring third-party claims adjusters to handle the excess number of claims); others, including segmenting catastrophe-impacted claims for separate loss reserving, are handled at the actuarial level. While estimating incurred but not reported (IBNR) reserves may not have an impact on any particular claim’s final settlement value, being able to accurately account for the particular characteristics of a catastrophe (including demand surge) is essential for insurers to be able to make accurate decisions regarding the impact of the event to their companies and alert their reinsurance partners in a timely manner. In doing so, insurance companies are likely to benchmark claims expectations to previous severe events.
Finally, the last consequence of major catastrophe events is that insurers will often pursue rate increases after an updated review of their claims experience and a renewal of their reinsurance programs. In theory, any particular catastrophe event that a company provides coverage for will have already been modeled and accounted for in the premiums that they charged for that policy period. Catastrophe models, however, are continually being updated to reflect the latest experience and conditions, and each event a company provides coverage for will help reassess its potential exposure and allow it to better evaluate future potential rate changes. A similar trend occurs in the reinsurance markets, which often see even larger impacts on financial performance from large catastrophes, and thus tend to be more price-sensitive after the occurrence of hurricanes and similar large events.
Disasters in a pandemic-driven inflationary environment
Given what we know about standard demand surge, how might we expect the insurance industry to respond when a higher inflationary environment is compounded by a significant loss event such as Hurricane Ida, which some estimates expect to produce more than $30 billion of insurance industry losses?6
In at least one instance, we already know the answer. The catastrophe modeling firm Risk Management Services (RMS), when producing its initial industry estimate of $31 billion to $44 billion for the event, noted that it expects many Hurricane Ida losses to be associated with what it called “post-event loss amplification” arising from a combination of standard demand surge factors, as well as COVID-19-related inflationary pressures.7 The combination of these factors is a major driver of RMS’s estimate, which is approximately 30% higher than those produced by other peer catastrophe modeling firms.
In other regards, it is too soon to know exactly how demand surge will factor into Hurricane Ida payouts, which are just beginning to be processed at the time of this article. However, we can track four metrics over the upcoming months to help inform these projections:
- Claim closure rates: We can compare the rates of closure on Hurricane Ida claims to prior catastrophes, to see whether there is a slower rate of closure due to delays in obtaining materials or beginning projects. Further, to the extent that insurers utilize third-party contractors or resources, it is reasonable to assume companies’ ability to provide these services might be hindered by the current economic environment, which would be trackable through claim closure rates.
- Paid and incurred loss rates and average size of claim (claim severity): We can track the incremental paid and incurred rates (as a percentage of estimated ultimate loss) and the average severities for various entities in the market against prior hurricane losses. In a demand surge environment compounded by the pandemic, we might expect to see longer tails on the development patterns, meaning that paid and incurred losses will take longer to reach their ultimate values compared to a standard environment. In addition, we might also expect to see higher claim severities overall compared to past events.
- Filed rate changes in the impacted markets: As insurers file for rate changes in 2022, we can note the various assumptions they are using, and potentially identify how much of the rate change is driven by recent events and revised provisions for catastrophes, versus how much is attributed to overall economic conditions or trends. In particular, because rate filings are a prospective pricing exercise that won’t begin to deliver additional dollars of premium to insurers until several months or even a year into the future, insurers must answer the key question: How long will this current inflationary environment be expected to last?
- CBI: If past trends hold, we will likely see a greater change in CBI in the areas impacted by Hurricane Ida (primarily the South, with some smaller impacts in the Northeast) compared to the rest of the country. Assessing the magnitude of these CBI changes compared to past events could give input into how much the pandemic environment is heightening the existing stresses caused by Hurricane Ida.
With evidence that we currently in an era of increasing natural disaster impacts (due to populations gathering in high-risk areas and potential climate risk impacts), the topic of demand surge will become increasingly relevant, no matter the underlying economic conditions. In an inflationary environment such as the one the United States currently faces, determining the interaction of both phenomena may be tricky—but it is an effort that informed insurers will certainly benefit from having completed.
1 U.S. Bureau of Labor Statistics (December 10, 2021). Consumer Price Index Summary. Economic News Release. Retrieved December 26, 2021, from https://www.bls.gov/news.release/cpi.nr0.htm.
2 Examples including an Insurance Services Office publication (1994), “A ‘demand surge’ in materials, labor, and living costs…increases costs much more than if each loss had occurred separately.”
3 Reid, W. (April 1908). Address at Luton Chamber of Commerce Annual Dinner. In the Bancroft Library, University of California, Berkeley.
4 Olsen, Anna H. & Porter, Keith A. (January 2010). What We Know about Demand Surge. Structural Engineering and Structural Mechanics. Retrieved December 26, 2021, from https://www.sparisk.com/pubs/Olsen-2010-CU-WWKADS.pdf.
5 Associated Builders and Contractors, Inc.
6 Scism, L. (September 22, 2021). Ida Storm Damage Expected to Cost Insurers at Least $31 Billion. Wall Street Journal. Retrieved December 26, 2021, from https://www.wsj.com/articles/ida-storm-damage-expected-to-cost-insurers-atleast-31-billion-11632303002.
7 RMS (September 16, 2021). RMS Estimates US$31-$44 Billion in Total U.S. Onshore and Offshore Insured Losses from Hurricane Ida. Retrieved December 26, 2021, from https://www.rms.com/newsroom/press-releases/press-detail/2021-09-16/rms-estimates-us31-44-billion-in-total-us-onshore-and-offshore-insured-losses-from-hurricane-ida.