Hurricane Ian: The gravity-defying Florida loss?
There was a clear “Florida fear factor” behind the wide range of loss estimates for Hurricane Ian that emerged after the late September storm, spanning low $30bn figures up to $74bn.
Behind that fear factor is the legacy of Hurricane Irma, which half a decade earlier caused losses that kept rolling over years amid a wave of litigation against local insurers.
However, market chatter around Ian four months on is increasingly – and surprisingly – optimistic that top-end projections may prove to be too high, and that the 2022 storm can withstand the state’s reputation for delivering drawn-out losses.
The evidence that is stacking up so far in Ian’s favour include:
Some insurers lowering reserves for the storm
A higher-severity, narrower loss favouring faster claims handling
Favourable regulatory changes that should shorten the storm’s tail
Delving into these further, there are few public signs of lower reserves, other than Progressive’s $400mn drop reported in December. However, according to sources some other primary carriers have either taken similar moves or their reported losses are coming in at the low end of projections.
One Floridian told this publication that, on a dollar basis, their claims are running 30% lower than expected, with lower average severity than projected. Another domestic leader said they had similar experience on lower severity, albeit without a specific projected figure, with a high closure rate on claims that was running ahead of Irma experience.
Other reinsurers pointed out that the loss has been more confined to the Florida domestic market than even the typical Florida loss. While nationwide insurers would be expected to be underweight on a Florida storm, it also helps limit the spread of claims throughout the wider sector.
Nationwide carrier losses are coming in often only within the first layer of reinsurance programmes, sometimes just under attachment, whereas reserves had been set further into their treaties based on early modelled loss information, a reinsurer observed.
On the counter side, PCS’ second-round loss pick for Ian rose 16% to $47.4bn towards year end, surprising observers. Adding on the loss-adjustment expenses and some public flood claims, this already puts its loss pick close to approaching a roughly $60bn event. However, backwards-looking PCS loss development is typically seen as somewhat of a wildcard barometer for upcoming claims trends – the increase “didn’t panic us”, a reinsurer noted.
Narrower and faster to handle
Meanwhile, there are the different characteristics of Ian versus Irma, which were already being discussed in the early months after the storm.
Ian was a much narrower, faster storm and hit an area that had been rebuilt to higher building-code standards following storms in the early 2000s.
“Generally, the housing stock down there is newer, more expensive, more robust,” one pointed out.
Gallagher Re’s chief science officer Steve Bowen said 25%-50% of building stock in the counties around Ian’s locus has been built since 2000, amid population growth in the region. The building codes they had to meet “for the most part, did their jobs”, he added.
While Ian caused severe damage to the hardest-hit areas, these are quicker to be recognised by insurers as a total loss, meaning less room for associated claims to creep up over time compared to Irma, where minor damage was widespread.
Aggressive Florida law firms seized on this after Irma to help homeowners get payouts for roof repairs and replacement that had not been expected by the industry – and the extent of roof replacements that took place may in itself limit loss development as homeowners with newer roofs are not likely to accept contractor offers for replacements paid by insurers.
Outside the residential market, one point of uncertainty for the industry in early assessments of Ian was going to be the degree of commercial losses that showed up, as these can take longer to emerge.
Some months on, London D&F sources have noted the market hasn’t really had the sort of full-limit losses on large accounts which you would often expect with a major hurricane.
One underwriter noted that the typical full-limit Ian losses which had arisen in areas such as Sanibel were made up of a lot of small hospitality, lower-scale losses with under $100mn in total insured value.
After Conduit Re released its renewal update this month, CUO Greg Roberts said questions surrounded whether hotel or BI losses would be involved. “From what I’ve seen, I don’t see much evidence of those large losses being reported,” he noted.
He added that the Lloyd’s binders losses appeared to have “materialised as expected”.
"There is a lot of SME type business in there where flood is sublimited, so they pay out and move on.”
With these emerging signs of optimism, the fundamental reason for the loading-up of the loss estimates remains.
There are still more than 18 months to run before the two-year statute of limitations on Ian claims expires. That’s one year less than Irma claimants had – the timeframe has since been brought down further to one year for future storms – but it still gives time for litigation to emerge.
It is unlikely that reinsurers at least, one step removed from the ultimate data, will be making major reserve changes at this time.
Other recent regulatory reforms could help to minimise the second-wave impact of Ian claims, but most from the year-end 2022 session only take impact on future policies or claims.
For example, reforms around how claims farm adjusters can promote their services will take impact immediately, but changes to the fee awards that can be made wouldn’t apply to retrospective claims.
Whereas with Irma, litigated losses related largely to minor roof damage, there is also the question of whether the flood vs wind debate will recur in a problematic way, although as yet there are few signs of direct pressure on this front. Local legislators are acutely aware of the pressures on private insurers, given multiple failures and exits in the past year.
As an indicator of how much litigation fears could have added to Ian loss estimates, modelling firm Karen Clark & Company’s (KCC’s) pick for this aspect of the loss reached $13bn – amplifying the loss by more than a quarter.
Meanwhile, inflation remains at high levels relative to the 2017 markets, or indeed levels that many people have experienced in their careers. This factor is highly likely to mean the “gut feel” type estimates that many in the industry have for Ian may be based on dated assumptions. Even with building cost data well down on its peak, costs remain well above the pre-Covid-19 era.
Discussing how Ian loss estimates ranked compared to Irma, KCC founder Karen Clark noted in October that inflation and Covid-19-driven supply-chain issues have led replacement values to be 25% higher than in 2017, even setting aside social inflation and lawsuits.
So where does that leave the current take on industry losses for Ian?
Despite the huge $45bn range in loss estimates that emerged after the storm, many in the market had always discarded the very top-end estimates and London market sources often put the loss around the $50bn mark.
Thus, it is difficult to measure how the optimism around the storm translates to the insured loss range.
Does it simply firm up the early sense that top-end estimates were too high, or put a stronger case for the mid-range point estimates to be lowered from $50bn towards the $40bn mark?
At the most optimistic end, one source projected that industry loss numbers could drop from the $50bn level to the high-30s range, with a few more in the $40bn range.
However, others were still sticking with the view that the loss will be an early $50bn event, consistent with the mid-range of modelled loss projections and early consensus.
Even these voices are becoming ever more confident that Ian can avoid a repeat of the whiplash from Irma’s tail loss development – pointing to the paring-back of worst-case scenarios.
On the Conduit renewals call, Roberts echoed this uncertainty when he said: “Do I think $55bn is enough as an industry loss? I would say so. Is it going to materially drop below that? I don’t know.”