ORIGINAL PUBLICATION HERE
Despite a multitude of other catastrophes in 2018, the single biggest loss event for reinsurers was so-called loss creep from 2017’s Hurricane Irma. Dirk Lohmann of Schroders examines what went wrong and what the industry must learn.
If there were two words to describe an investor’s perception of one of the key drivers of performance for many insurance-linked securities (ILS) managers (and indeed many professional reinsurers) in 2018 they would without a doubt be ‘loss creep’.
During the course of 2018 the loss estimates provided by Property Claims Service (PCS) for Hurricane Irma increased by $8 billion.
Of this amount, $7 billion alone was in the state of Florida. Here, virtually all the deterioration was borne by traditional and alternative reinsurance markets, given that most insurers had already exhausted their reinsurance retentions. In this context it should be noted that PCS’s figure only covers indemnity losses.
Added to this are loss adjustment expenses (LAE) that are also covered under the reinsurance protections and serve to further inflate the claims ceded to reinsurers. While the 2018 year itself was impacted by many events ranging from typhoons in Asia to two hurricanes making landfall in the US and wildfires in California, the loss deterioration experienced from Irma was probably the largest single loss ‘event’ for reinsurers during 2018 (Figure 1: Insured losses from Hurricane Irma – Florida). What happened? Was this unique? What lessons can be learned?
To understand the developments leading to the dramatic outcome of Irma, it is helpful to look back and recall the situation within the property insurance market in Florida, post Hurricane Andrew in 1992 and then the losses of 2004 (Charley, Francis, Ivan and Jeanne) and 2005 (Dennis, Katrina, Rita and Wilma).
Hurricane Andrew resulted in the insolvency of certain insurance companies and caused many national writing insurers to reduce their exposure or withdraw entirely from the market, resulting in an availability crisis for Florida policyholders.
This led to the formation of assigned risk pools and ultimately of Citizens Property Insurance Corporation (Citizens) in 2002 as a market of last resort. The activity in 2004 and 2005 led to further withdrawals and Citizens with a funding deficit of several billion dollars. In the wake of 2005 efforts were undertaken to reduce the role of Citizens in the Florida property insurance market by promoting depopulation efforts and paying private insurers bounties to assume risks from Citizens.
This prompted the expansion of certain existing companies as well as the formation of a number of smaller ‘take-out’ insurers focusing exclusively on Florida homeowners insurance. Depopulation from Citizens peaked in 2014, with more than 415,000 policies with an exposure of $117 billion being moved to the private market (Figure 2: Depopulated Citizens policy numbers and exposure between 2007 and 2019 year to date).
While the depopulation efforts resulted in a more vibrant insurance market and offered consumers greater choice, it should be remembered that it coincided with a dramatic rise in alternative capital entering into the reinsurance market, which was essential for the market to evolve.
Most take-out insurers had minimal capitalisation and were highly leveraged through extensive reliance on reinsurance. Depopulation was happening during a prolonged period of inactivity with no land-falling hurricanes in Florida for a period of 12 years. This prolonged hurricane ‘drought’ (the long-term average frequency for Florida land-falling hurricanes is one every three years) resulted in a certain degree of complacence among local insurance carriers and their management.
In many cases, more emphasis was put on growth (often in other states) than in ensuring the ability to deliver on the promise behind the policy. While all companies had their claims experts and emergency plans for catastrophes, the bulk of the claims servicing capacity was reliant upon outsourced capabilities.
A wake-up call
Irma was a wake-up call for the industry, exposing underappreciated complexities in the claims process and highlighting operational weaknesses, specifically among the Florida specialist writers of residential property insurance.
First, there were not enough third party adjusters available to fulfil the contracted outsourcing arrangements that local specialist insurers relied upon. While many have argued that the shortage of available third party adjusters was due to the untimely occurrence of Hurricane Harvey in Texas three weeks earlier, many forget that the number of loss adjusters in the region had also fallen since 2005.
This is a direct consequence of low hurricane activity in the interim, with many not seeing it as a viable profession given the lack of demand in loss-free years. Consequently, the outsourcing agreements were not worth the paper they were written on and higher adjusting fees had to be paid to secure the necessary resources.
Second, the lack of seasoned adjustment professionals can, and did, result in errors and underestimation of damages leading to many claims files being closed prematurely on the assumption that losses would be contained within the deductible, only to be reopened at a later date. With claims, time is critical for costs and without an experienced specialist unit to give accurate estimations in a timely fashion the insurer will see associated costs rising and also run the risk of the additional burden of lengthy litigation and further spiralling costs.
Another driver of claims cost related to assignment of benefits (AOB)-related costs. Problems surrounding AOB were pre-existing, often involving water damage, with the claims process open to exploitation by public adjusters, contractors and lawyers. In early 2017, Demotech, a rating agency, raised the issue of adverse development on non-catastrophe claims among writers of Florida homeowners business due to AOB, highlighting the need for companies to review their claims protocols.
The issue of AOB had been building for a number of years with adverse court decisions expanding the liability of insurers and the Office of Insurance Regulation restricting the use of anti-assignment provision in the policy language in 2015. The AOB issue was accentuated through two adverse state Supreme Court decisions in 2016 further expanding the exposure of insurers to attorneys’ fees and broadening coverage by adopting the concurrent causation doctrine.
Irma further exacerbated the AOB issue, especially because of the difficulties some insurers faced in processing claims via outsourced adjustment services rather than a robust in-house solution.
While the claims and legal environment in Florida appears to be particularly hostile to insurers, this is not something that is new; Florida has long been known to be a challenging environment for insurers and consistently ranked as one of the top “judicial hellholes” by the American Tort Reform Foundation.
What becomes apparent when looking at the performance of various participants in the market is that there is a clear divergence in loss development and associated loss adjustment costs between those carriers that have significant own claims management resources and those who do not. Companies with a greater commitment to allocating resources to sound loss adjustment and claims processes have seen a translation of this preparation into efficiencies and ultimately cost savings.
Many of these are insurers with a national presence that requires an organised approach to allocating resources the right place at the right time. Covering frequency-driven events such as auto and workers’ compensation can also help hone the capabilities and streamline the claims process. We see this efficiency carried through to Florida, with fewer AOB claims and lower loss adjustment expenses (LAEs) in well-organised firms.
As can be seen in Table 1, monoline property insurers focused primarily on homeowners or commercial residential business have posted higher LAE figures.
While Table 1 (Comparison of Hurricane Irma loss adjustment expenses) is not based upon an exhaustive analysis of all insurers, we believe it is representative of the industry having covered over $4 billion in incurred indemnity losses. What is clear is that no market included in its evaluation of exposure and pricing a loading for LAE on the order of 20 percent or more, or the spilling over of reinsurance recovery shortfalls due to the Florida Hurricane Catastrophe Fund (FHCF) 5 percent cap on LAE.
One of the guiding principles of reinsurance is that of following the fortunes of the reinsured. Another is that the reinsured should settle claims and act as if not reinsured. While we remain committed to sharing the fortunes of our clients for fortuitous catastrophe events, one does need to ask if this sharing of fortunes should also extend to picking up the cost of operational shortfalls in the management of the business.
Presently, coverage is granted to most companies on a 100 percent basis above fairly modest retentions. This raises the question of whether interests are actually fully aligned once claims exceed the retention?
If one were to look at catastrophe bonds as a proxy for the capital markets it is interesting to note that only one Florida exposed catastrophe bond provided full coverage for LAE. All others had a cap or a LAE factor ranging from 6 to 17.5 percent. Even in the traditional reinsurance markets, there are exceptional cases where the ceding insurer applies a fixed claims adjustment fee instead of billing the actual incurred LAE.
In the past, it was common for ceding insurers to co-reinsure a portion of the catastrophe layers placed with the market. This ranged from 2.5 to 10 percent and was designed to reinforce the alignment of interest between ceding insurer and its reinsurer. This practice appears to have largely disappeared in the softer traditional reinsurance market, although one does still see an element of this in certain cat bond transactions. Interestingly, the maximum coverage that the FHCF offers is capped at 90 percent, probably also to ensure a degree of alignment of interests between the private sector and a state-backed reinsurer.
The severe adverse development and the highly inflated LAE related to Florida Irma claims observed at many ceding insurers has shaken the confidence of many market participants and has resulted in many investors questioning the ability of the market to properly assess the magnitude of a loss post-event. This then raises the question whether consideration needs to be given to adopting changes to ensure a more reliable outcome and better alignment between business partners.
While some may argue that the drivers of the loss experience surrounding Irma were exceptional, we would argue that many of the underlying forces were known and existed before Irma occurred. Also, a comparison of LAE experience shows a clear differentiation in actual experience depending upon the company setup and business scope.
As it stands, Irma actually proved to be a ‘near miss’ for the industry. Had it followed the earlier projected loss track over the eastern Florida coast, the magnitude of the loss and the impact on local insurers would have been far greater and many would have most likely exhausted their reinsurance protections and faced ruin.
From that perspective, it is incumbent upon the industry and the reinsurance capital providers, whether traditional or alternative, to reconsider the scope and structure of coverage provided to ensure a stronger alignment of interests between ceding insurer and reinsurer. This could be through increased retentions, including a vertical retention up the reinsurance tower to promote a more balanced sharing of fortunes as losses develop and/or introducing a LAE cap or fixed claims adjustment fee to allow for a proper reflection of this additional cost in the pricing of the reinsurance cover.
There may be other potential ways of addressing the issues raised by the Irma loss experience and we welcome additional ideas, but simply adding more premium on the slip and continuing with business as usual will not be sufficient.