The Emerging Risks Interview: Tom Johansmeyer, PCS

In this first part of a two-part interview with Emerging Risks, Tom Johansmeyer, head of Property Claim Services (PCS) at Verisk – insurance solutions, talks about catastrophe outliers, insured flood losses and the right way to utilise models.

-From your perspective, has there been a clear trend over the past 15-20 years for higher nat cat losses in the North American market?

Looking specifically at the data, the answer is yes. But what I find more interesting is what’s going on under the hood. What I remember most distinctly from almost a decade of working at PCS is that it’s the stuff that everyone says doesn’t happen is problematic. I remember that when we entered Turkey everyone was saying that it’s a quake and flood market. Ok- but the first catastrophe we did after we went live was political violence and it was one of the largest insured losses in Turkish history. And then after that we had two hail events in Istanbul!

Also, people were saying that wildfire is never a big deal. Then we had Fort McMurray in 2016 in Canada, which at the time was the largest insured loss we had in Canada or the US for wildfire, but now it’s not even in the top five. And $3-4 billion hail has become routine, while a massive hurricane hitting Puerto Rico is a huge insured loss which nobody saw coming. And we had a massive political violence event in the US last year – that hasn’t happened since ’92, and ’92 wasn’t as big. So what I find interesting is that not only is the number of catastrophes in North America increasing, they are also becoming more diverse and more troubling. When you look at the sheer scope of cats that can occur, we’re seeing that tested more, and we’re seeing the upper limits of those tested as well.

 -To what extent do you think there is a correlation between higher insured losses and climate change?

I haven’t driven anything with a motor since 2007, and I recently became a weekday vegetarian to decarbonise my diet. I can’t give up meatballs and lasagne at the weekends, though- I’m half Italian and I have standards, right? I’m trying to do my part, and I’m a true believer on climate change – but I’m really uncomfortable with the discussion of whether there is a correlation between climate change and larger insured losses.

It looks like there is some amount of correlation between climate change and the risk of large losses, but does climate change lead to more – and more intense – catastrophes? Yes, we’ve seen that all over. But does it lead to greater insured losses? Not necessarily. For example, you need to have an event that makes landfall, and that event has to make landfall in a place with a lot of insured value. And then on and on… it’s just not as simple as climate change = more insured losses.

I’m always careful about trying to use PCS as a measure in there, because it’s not the right fit. And the one thing I really worry about when we pick our measures is whether we are picking metrics that are easy for climate change deniers to pick apart?

But are we fully supportive of efforts to understand/ quantify and to prevent/resist/ mitigate climate change? Absolutely. We just don’t think our data is the right tool for it. 

-What has been the main driver of nat cat losses in recent years in North America- wind or flood?

For insured losses, flood is becoming more of an issue, although it has always been there in the background. But you’ve got to be careful when looking at flood as insured versus not… flood is not always just flood. Wind can create an opening that water can get into, for example. Also, hurricanes are becoming more of an issue, and I think that Harvey was really when the light switched on for a lot of folks.

It’s interesting. We’re seeing from another Verisk database called Exact Analysis that there were almost as many claims from Ida in Brooklyn as there were in Alabama. So one borough in one city versus an entire state. Granted, I know that Brooklyn has a couple of million people and a whole lot of ‘em packed together, but it’s still sobering to look at that when you’re trying to understand a cat.

When you look at flood, a conversation I’ve been having a lot down here in Bermuda is with the reinsurers, who look at the impact to their book and it’s a numerical exercise. So I tell people: use your head an put yourself into the cat. So I ask them if they’ve ever been to Manhattan? Yes, of course I’ve been to Manhattan. You ever take a client to a nice restaurant in Manhattan? Yes, totally. Midtown, right, where you see the restaurant at ground level and go down a set of stairs from the street? Yes, totally. Well, where do you think the kitchen is for that restaurant? It’s down under the ground. Now a flood comes- what do you think happens to that restaurant? It’s washed out- and that has insurance implications. You’re going to tell me there’s no claim on that?

-Axa recently said we should expect more severe and frequent flood risk in Europe as the century progresses- do you think the same is true for North America?

 We’re not predictive, but would I would say is that anyone in this business should be looking at recent events, at climate conditions, at projections for climate change, and act appropriately and figure out what could be coming. This recent flood in Germany to me is the same issue as Canadian wildfire or Turkish hail, and all this other stuff that was not supposed to happen – at least not with the quanta involved, right? Who saw a German flood at that magnitude coming? Nobody. So the answer is not to figure out how to see German flooding coming, but how to figure out what risks we could be exposed to but aren’t taking seriously enough.

 -To what extent is the NA property cat market adequately served by appropriate loss models? Is there still room for improvement?

 Is there room for improvement? Absolutely. Of course there is. Has cat modelling progressed to an impressive degree over its lifetime? Yes. Where cat modelling gets an unfair rap is how it’s often used, and I’ve seen this a lot around Ida. After Ida you look at the AIR or RMS number, then you look at the PCS number and you scratch your head and you can’t figure out why it’s so different? Well you’re doing it wrong. Heading into an event, cat models can help you understand what’s coming. Can you supplement it with additional data, and should you? Definitely.

But post event? From that point you have your model, you have PCS, and maybe you have some other stuff as well such as gossip from the broking sector  – which can be useful. Now you have this big collection of insights. That’s when you sit down. And your cat model is an important part of it, but it has to be part of a constellation that creates your view. I think the users have to figure out how they are using those kind of tools in order for the sector to further realise its potential.

For insured losses, flood is becoming more of an issue, although it has always been there in the background. But you’ve got to be careful when looking at flood as insured versus not… flood is not always just flood. Wind can create an opening that water can get into, for example. Also, hurricanes are becoming more of an issue, and I think that Harvey was really when the light switched on for a lot of folks.

Follow us on twitter: @risksEmerging

Owned Weak RSA from Hack The Box! #Cybersecurity #security via

These people get it. Allowing global capital to replace one destructive industry with another is not a realistic way forward. We need an alternative to the growth imperative.

Witness the Power of Graphs - USC Viterbi | School of Engineering

@JimHarris @JolaBurnett

#CNNs #Matplotlib #5G #NLP #SciketLearn #AutonomousVehicles #ArtificialIntelligence #LinearAlgebra #MLOps #DataScience #CyberSecurity

Connecticut Democratic Sen. Chris Murphy said he continues to work with his Republican and Democratic colleagues to pass gun safety legislation, especially after the deadly school shooting last week in Michigan.

#tech #technology #AR #vR #AugmentedReality #AI #Algorithm #innovation #reinvent #cloud #IoT #5g #data #DataScience #BigData #CyberSecurity #automation #CloudComputing #Algorithm #StaySafe #wfh #ArtificialIntelligence A definite game changer !! 👇 Thanks @Ronald_vanLoon

Ronald van Loon at @AWScloud #reInvent@Ronald_vanLoon

#AugmentedReality windshields could soon be hitting the roads
by @usatodaytech

#AI #ArtificialIntelligence #VR #AR #VirtualReality #MixedReality #MR

Cc: @rethinkrobotics @davidsmith4324 @maxjcm @ronald_vanloon @xbrlstandard

"The GDPR brought in the concept of Security by design."- Pooja Bansal, @Fareportal Speaking live at #PublicPolicyForum on Security & Data Protection

77th #SKOCHSummit #Caii #SKOCHParivar #CyberSecurity

@SkochSameer @oscilatting @DhanjalDr @abmindia

Buy Google Voice Accounts
Google Voice Number Best Contact Site. Connect anywhere with Google Voice account.
#GVoiceNumber #MachineLearning #DataScience #Python #AI #100DaysOfCode #IoT #flutter #CyberSecurity #tech #DDoS #data #RPA #IIoT #5G

Load More...