The Intelligent Quarterly from the publishers of The Insurance Insider

Summer 2018

Search archive

 

Where are the missing billions?

Charlie Thomas

This year is likely to be one of the costliest on record for the international (re)insurance industry, thanks largely to hurricanes Harvey, Irma and Maria (HIM).

But how much those hurricanes will end up costing the industry is a matter of contention. Modellers seem unable to agree, with estimates ranging from less than $100bn to more than $140bn. The early net numbers published so far by carriers, meanwhile, add up to less than $50bn, around $28bn of which relates to HIM and the Mexican quakes. (see table below).

And uncertainties abound around how much demand surge could inflate rebuilding costs, how much loss adjustment expenses will go up by, and how much of the loss will ultimately end up in the relatively opaque alternative capital markets.

There are therefore three main questions that need to be answered: firstly, do we still believe hurricane-related insured losses will amount to more than $100bn; secondly, if yes, where in the (re)insurance food chain will the losses end up; and thirdly, how much confidence can we have in the answers to the first two questions?

Estimating the loss
This year has seen one of the busiest seasons for named storms. Some 17 named storms formed, 10 of which reached hurricane strength, and six of which became major hurricanes, according to the National Oceanic and Atmospheric Administration.

Click to open That means 2017 had the most named storms and hurricanes since 2005, when 28 named storms developed, 15 of which became hurricanes. It was also the first year since 2005 that hurricanes struck the continental US.

Insured loss ranges for the storms vary wildly between different parties. Data provider PCS currently estimates the total for Harvey at $15.9bn, Irma at $18bn and Maria at $21.9bn, meaning together the three events would cost just $55.8bn.

The Bermuda Monetary Authority, meanwhile, has claimed that carriers on the island will collect a $31.2bn net burden from the trio of storms, which it estimated would account for 30 percent of the global industry's losses. That would put the total at around $104bn.

AM Best, on the other hand, said its survey of rated carriers suggested a total of $90bn, while estimates from the two largest modelling firms, AIR Worldwide and RMS, range from as little as $75bn to more than $125bn.

The huge variance in estimates is in part driven by the different methodologies used by the various parties.

PCS, for example, generates its figures using data from confidential surveys of insurers, agents, adjusters, public officials and other sources, which it then analyses alongside trend factors to produce an estimate.

Loss numbers generated on this basis can deteriorate by at least 25 percent, and do not capture 100 percent of the US losses.

Meanwhile AIR and RMS, as the feature on page 18 explains in more detail, differ wildly in terms of the inputs for their models, including the assumptions they take on properties and their ability to withstand damage.

Puerto Rico and other problems
Nowhere is this more obvious than with the firms' estimates for Maria, which hit the Caribbean at the end of September. RMS estimated that insured losses would come in between $15bn and $30bn, while AIR surprised the market with a range of $40bn-$85bn, with $35bn-$75bn of that coming from Puerto Rico. It later revised its estimate downwards to $27bn-$48bn, with $25bn-$43bn for Puerto Rico.

The vast difference in the ranges is in part down to differing beliefs on the vulnerability of Puerto Rico's buildings. RMS believed Puerto Rico's bunker-style buildings would largely have been able to withstand the storm, while AIR assumed wind damage would have seriously hit the vulnerable upper storeys of the island's buildings.

AIR also believed "demand surge" was likely to hit repair and rebuild costs - suggesting as much as 27 percent of its uppermost estimate could come from that.

This confused situation from third party vendors, plus the lack of data coming out of the affected regions - particularly in Puerto Rico - has resulted in a very mixed bag of views.

Ian Beaton, CEO at Ark Underwriting, thinks a bill of more than $100bn still sounds possible, based on the assumptions published so far.

"Like all the other cats we won't know the true number for a while, but plus or minus $10bn, it's probably correct," he says, although we're likely to get closer to that level if we include losses from the Mexican quakes and Californian wildfires.

Between the estimates from AIR, RMS, PCS, AM Best and others, once you add in charges to the National Flood Insurance Program and the cat bond market, it feels like it gets towards $100bn, he added.

RK Harrison's head of claims Nick Coles, on the other hand, disagrees. Based on looking at his clients' books of business, Coles says it "looks unlikely that we'll reach $100bn from the HIM hurricanes alone".

"There are some really big losses out there, but it doesn't look as though there are enough of them, at this point, to get to that sort of number," he adds.

Another issue worth considering is that primary carriers were reportedly put under pressure by their reinsurance counterparts to come out with loss estimates more quickly than they wanted to.

"A number were uncomfortable about it at the time," says Coles. "What they didn't want to do was end up step-reserving, as it doesn't make the claim run particularly smoothly. It creates a level of angst in the claim which is unhelpful. Perhaps that's pushed some of the estimates to be over-egged and [resulted in] some higher reserves coming out."

How the losses shake out
Few commentators have been brave enough to try and estimate how much of the net loss will end up with primary insurers, reinsurers, retro providers and alternative capital, but broadly speaking, it seems most expect between 25 percent and a third could end up with alternative money - that market's first real test since it was born out of 2005's cluster of hurricanes.

AM Best was one of the few to try and put a number on the various groups. In a report released in November, it suggested that based on conversations with rated entities some $45bn of the losses from HIM and the Mexican quakes would sit with the primary market, another $20bn-$25bn with the reinsurance market, and a final $20bn-$25bn with alternative capital, mostly in the form of collateralised retrocession.

The high level of primary net loss might look surprising, but a number of sources indicated that retentions had been increasing over the past decade, particularly in Puerto Rico.

As detailed on page 22, our own interrogation of the insurance-linked securities (ILS) and retro markets seems to mirror this outcome.

Hurricane risk is the ILS market's greatest area of exposure, but the HIM storms each presented such different challenges that they impacted ILS managers in a variety of ways. Retrocession specialists bore the brunt of the losses as aggregation claims mounted, while the cat bond market appears to have got away relatively lightly.

Off-the-record comments from the market suggest retro writers may have lost up to half of their money this year, although it's challenging to get confirmation, given that these players sit at the end of the value chain and may have difficulty establishing their losses at such an early stage.

Looking at the industry loss warranty (ILW) market, the Micrix index fell 12 percent after September, implying a $720mn loss from the $5bn-$7bn market, although when aggregate ILW triggers are added in this figure is likely to be higher.

Interestingly, this event also seems to suggest that retro strategies are far more exposed to a 1-in-10-year aggregate loss than a 1-in-100-year single hurricane.

Mitigating factors
Underlying warnings from the likes of Willis Towers Watson's Bill Dubinsky, who said it is "still far too early to close the book on the exact allocation of losses among insurers and reinsurers, let alone between traditional balance sheets and ILS", are a number of issues complicating the market's ability to get its arms around the HIM losses.

Chief among the complaints of those seeking clarity on the market's HIM losses is the growing question mark around Maria's impact.

Over the past few weeks, anecdotal evidence has pointed to a slightly better picture than first predicted for Harvey and Irma, with adjusters able to get on the ground relatively quickly after the events and begin their assessments. Cargo losses didn't look as bad as first feared and flood wasn't covered by much of the market.

Interestingly, Coles notes that adjusters were being brought into Houston to generate full damage reports so that, if they needed to reject a claim on the basis that the coverage did not include flood, they could point to independent evidence that the damage was caused by water and not wind.

Puerto Rico, however, is a different ball game. Getting on and around the island in the days after Maria struck was nigh on impossible, as Vince Cole, US CEO of Charles Taylor Adjusting, explains.

Huge amounts of infrastructure damage meant there was no power, no telecommunications, no traffic lights, nowhere to eat and nowhere to stay, he says. The time taken for simple tasks such as navigating around San Juan suddenly went from 20 minutes to four hours.

"The first month I had people describing Puerto Rico to me as a bit of a warzone," says Cole. "We had two guys there who were staying in tent and a rented home. The tent obviously had no power and the rented one only had a generator that worked for a few hours at a time.

"Meals were beef jerky and bottles of water. Another guy found a Starbucks after a few weeks and survived off eating their food for three or four weeks. It's not like they get a vacation in the Caribbean."

The terrain also caused problems, he says: "Getting around the cell towers in a mountainous region, finding them and all the bits and pieces and parts has been - and will continue to be for a year or so - a challenge."

Convincing adjusters to go to Puerto Rico was also a problem. Stories of poaching were rife. Cole agrees with other sources that fees for adjusters ramped up due to the quick succession of natural catastrophe events.

"It's not like we don't have enough resources as an industry, there were enough adjusters available, but when the events get stacked like this, you have to pay a little more for each subsequent event. So yes, the expenses are higher."

No more 3-for-2
The worrying thing as far as the (re)insurance market is concerned, is we're nowhere near obtaining a full Puerto Rico loss number.

"We don't have a handle on it yet, and the number of loss adjusters out there isn't as great as it should be for an event of this size," says Cole. "At this stage, I don't think we're even up to half of the total expected loss notifications... it'll be another 12 months before we discover what's fully going on there."

Elsewhere, there are also rumblings of potential legal disputes brewing - and not just in super-litigious Houston.

Several sources said debates were already happening around Irma and Maria, with disputes over coverage and aggregation issues and whether buildings and infrastructure were hit by Irma, Maria or both.

Finally, there is a big unknown in the form of how big a problem business interruption and contingent business interruption might prove to be.

It's too early to ascertain the impact this year's cats will have on rates in the upcoming January renewals, but there is a growing consensus that coverage levels will become a key talking point.

As Ark's Beaton explains: "Has anyone really been pricing in flood risk at anything more than marginally above zero? Bushfire - another huge issue this year, not just in California - when people were pricing that, were they putting in anything other than something approaching zero?

"Effectively what's happened is there's been a lot of 3-for-2 shampoo sold and people are now realising they can't do 3-for-2 anymore and need to put a proper price next to it. If you just want the coverage just for named windstorm, then let's give that restricted cover again.

"There's definitely a coverage conversation about what people want to buy and what people want to sell," he concludes.

Click to open

This article was published as part of issue Winter 2017

Euromoney Trading Limited - 3rd Floor, 41 Eastcheap, London, EC3M 1DT, United Kingdom. The content of this website is copyright of Euromoney Trading Limited 2018. All rights reserved Euromoney Trading Limited actively monitors usage of our website and products and reserves the right to terminate accounts if abuse occurs.

Π