The reinsurance market at mid-year 2019 – diagnosis and prognosis

Ross Nottingham manages the North American reinsurance teams for the Hiscox Group in both the Bermuda and Lloyd's platforms. This business is capitalised by Hiscox's own balance sheet and a range of third-party partnerships and funds.

In this interview, Ross explains the recent market dynamics and the opportunities in 2020 and beyond.

Click here for mid-year 2019 data

What have you seen happen to the supply and demand for reinsurance over the last 12 months?

Supply has shrunk, largely down to ILS markets pulling back for a host of reasons. One is capital that has been paid out or locked up or due to loss creep, specifically Hurricane Irma and Typhoon Jebi.

It also seems the investors are taking a watchful ‘wait and see’ approach to rate movements before deciding whether to redeploy further funds. In 2018 there was a promise of more rate, whereas actually, a trifecta of bad news followed; stale rates, increased existing loss and new accident year losses, which explains the hesitancy to reload again so quickly.

And climate change seems to crop up in most investor meetings which we can’t deny. Science points us to a worsening loss picture and yet also steers us towards acceptable returns for the risk we take.

How has that affected US reinsurance pricing at the key renewal dates (January 1, June 1 and July 1)

There is still enough capacity for the right risks. Largely the US property treaty rates for non-loss affected business remain flat and at the lowest point in the cycle since 2014. Loss affected areas are pricing higher: The areas of California for wildfire; nationwide aggregate covers; low-end risk excess treaties or Florida business. It somewhat flies in the doctrine of insurance, where the premiums of the many should pay for the losses of the few. Now it is just the localised few that are paying increased rates.

Florida clients paid more for their reinsurance on June 1 - increases of around 8% to 50% and more. The increase depended on the individual losses, the management response and the level of the layer. Although Hiscox only supports a subsection of these clients, we could tell the capacity had been withdrawn across the whole market as markets where you would normally expect to be signed back to 70% or thereabouts, suddenly you’ve received a full signing.

California mid-year renewals went one further and there were repriced contracts that simply weren’t getting home. Finding new wildfire limit is extremely tough as management has restricted underwriting teams to go into wildfire season with less exposure.

At July 1, capacity for big renewals was stable and flattish. New programs, however, were tough to place as capacity dried up which may bode well for January 2020.

Do you think that reinsurers need to update their models following the experience of 2017/2018? Are you satisfied that they adequately capture risks such as global warming and demand surge?

Reinsurers need to continually update models and views of risk with the arrival of new loss data points. At Hiscox, we have a team of data scientists and weather specialists pulling apart vendor models and refining our view for our clients and their experience.

Ripping apart a new vendor model for wildfire we think the expected loss it generates needs lifting up for, already out of date, data ranges in respect to acreage burned and where people live (wild-land urban interface). Adding drought as a modifier, the next step is seasonal prediction.

In Florida, the Hiscox view of risk has incorporated a set of what we call ‘Social Inflation’ uplift factors, where we are trying to establish who may suffer from the kind of model miss we have seen in Hurricane Irma from 'assignment of benefits' or public adjuster inflation. These factors, vary per wind speed and need to be applied on a differentiated basis. We tier cedants according to a host of qualitative scores, such as management experience and claims team quality as well as quantitative data pointing towards model miss and locations of values.

This insight and understanding, offering different views of the vendor model gives us the confidence to take the material bets we do.

What is your outlook for the market at 1/1 2020 if there are no major losses?

The retrocession market has already moved in 2019 and will drive further hardening this January. With the reduction of pillared retro at the bottom end of programs and the repricing of aggregate covers, which would normally step into that level, we predict buyers will be reminded to look closely at gross aggregates again, squeezing the supply of primary reinsurance for the same level of demand.

As well as the losses and the withdrawal of capacity from certain large providers, trust in the collateralised product is down from its peak further to well-publicised collateral release shenanigans. Buyers are looking for a different blend of retro with the balancing swings back in favour of rated paper.

What is your outlook for the market 1/1 2020 if there losses similar to 2017 or 2018?

Harder still. Insurers are getting rate in the US commercial and E&S markets and, as mentioned above, retro is getting rate. If we see more losses at the scale of 2017 and 2018 the whole reinsurance market moves, clean or loss affected.

Hiscox can access risk and capital using multiple platforms. Where do you see the future of the market (Lloyd’s versus Bermuda; own balance sheet versus funds).

All of the above. The beauty of having ILS funds behind a rated vehicle is that you can offer clients exactly what they want: the underwriting expertise and relationship-driven shop window; the ability and willingness to pay valid claims; and access to traditional rated paper supported by investors that have access to a range of different return appetites to offer to the cedants.

What are the most exciting opportunities for reinsurers to grow their market?

Cyber is a growth area. Hiscox is a lead market for non-proportional cyber treaty business.

We are all still building our models on nascent loss data so the chance that we haven’t seen the worst or can’t predict an accurate expected loss is high. The Lloyd’s push to exclude Silent Cyber from property treaties seems perfectly sensible. If the exposure doesn’t exist, let’s exclude it just in case. And if it does, let’s measure it and charge a premium for the cover.

We all owe a duty to try to close the protection gap be it in developed or developing countries and economies. With the largest US earthquake in 20 years in the last few days, we remind ourselves of the excruciatingly low take up rates for earthquake. Why wait until the next loss? We saw in Hurricane Harvey that the flood cover was woefully inadequate and despite a small uptick in NFIP sales post event, the gap is enormous. At Hiscox we have been designing products for flood and earthquake to make the cover more readily available and affordable and yet there exists an “it won’t happen to me” mindset and uptake of a private solution is minimal. There is real reticence to buy more insurance that isn’t mandated.

Our industry needs to continue to try and to innovate. insure-tech is no silver bullet but incremental gains which will benefit all insurers, not just the start ups.

Posted: Monday, July 8th, 2019