Q&A with KatRisk, global catastrophe modeler

by Jen Smoldt on July 2, 2018


KatRisk’s probabilistic multi-peril models are propelling this Cat modeler, largely known for their U.S. flood models, to the forefront of the flood modeling space. For more than four years, SpatialKey has worked with KatRisk to help underwriters and risk analysts gain a better understanding of flood risk. Our clients are finding that KatRisk’s overall score and flood depths by return period are helping underwriters understand both fluvial and pluvial flood risk like never before. In this interview, we talk with KatRisk co-founder, Dag Lohmann, who talks about the challenges of modeling U.S. flood risk, FEMA, and what’s next on the horizon for this innovative Cat modeler.

What is your data story?

We started KatRisk because we saw a need for high-quality flood risk models (not just flood maps) in the market. We envisioned that, by starting from scratch, these models could incorporate everything that we saw as important in the modeling process: precipitation from tropical cyclones, climate sensitivity, global teleconnections, fast sampling software, etc. From the beginning, we had many discussions with our clients about what type of product they wanted and on what time scale. Most clients first wanted maps, then analytics on those maps, and then probabilistic models. We did exactly that.

How is KatRisk different from other flood modelers?

We model many perils probabilistically (wind, storm surge, inland flood, droughts, wildfire, etc.). Flood is an important part, but it’s not the complete picture. We’ve also worked a lot on our probabilistic multi-peril modeling engine. I think our current clients would agree that it is the highest performance multi-peril and financial model they have seen, with much stronger mathematical rigor (i.e. coherent view of risk). Our technology stack is also easy and fast to deploy. We just like to talk about Cat models and how we can solve our clients’ problems.

Can you tell us about your U.S. flood model, and some of the challenges of modeling U.S. flood risk?

The main challenge is the multi-peril calculations, the spatial resolution (10m), and integrating our models within the client’s workflow. The U.S. has a lot of data available, and one must judge carefully what to use and what not to use. A main part of our story is that we had access to the fastest computer in the U.S. for some time. That helped us with the development of our global and U.S. flood maps.


FEMA has long been the standard for insurers who underwrite flood risk in the U.S. Why should insurers look to new sources of flood data?

FEMA has been a client of ours since last year for our maps, loss analytics, and probabilistic models. FEMA maps are a fantastic resource in many parts of the country. The engineers, scientists, FEMA staff, and everybody else who works on these is doing a great service for the U.S. taxpayers. That said, wise risk management doesn't look at just one source of data.

How is your solution helping property underwriters specifically? 

I think it is fair to say that we invented deep underwriting analytics based on flood maps. We were then able to take that to build out our probabilistic flood risk models. Currently, we are the only company in the Cat modeling space that has a very comprehensive solution for wind, inland flood, and storm surge in the U.S. and Canada, and other parts of the world will follow soon. Our overall score and flood depths by return period help underwriters understand both fluvial and pluvial flood risk like never before. Likewise, we help underwriters and risk analysts understand flood risks in areas where FEMA has not mapped, and create underwriting guidelines that provide a competitive edge to write the most profitable business.

How is climate change impacting your models?

Cat models like ours are at this fascinating intersection of science and engineering with many social aspects and human behavior. They are a great tool to compute climate change sensitivity—if the models have been built with climate change in mind. Fortunately, we have done this from the start, and also added climate variability.

In what way does SpatialKey empower your data to provide insurers with even greater insights?

SpatialKey is a great solution for gaining a deeper understanding of flood risk, and it’s really the only high-performance visual analytics solution on the market. The insights that a risk analyst or underwriter can get from SpatialKey are very valuable. Likewise, SpatialKey recognized the value of our data long before others in the marketplace and has been a collaborative partner and champion for getting some very high-profile insurers onboard and deriving tangible value from our data.  

Can you speak to climate change and the opportunity for insurers to close the protection gap, especially as catastrophes increase in frequency and severity?

The protection gap is real and needs to be closed for more resilient and sustainable societies. But, this needs to happen with or without changes in frequency and severity. I would not subscribe to a hypothesis of an increase in natural disasters for all perils and regions. We certainly saw a huge spike in 2017 in the U.S. But, separating natural variability from trends is work in progress, in my opinion; especially when humans continue to spread out, we will have more events. In some areas the signal is very clear, such as sea level rise. In others, I would caution and refer to statements the IPCC (Intergovernmental Panel on Climate Change) makes, for example, for flood and precipitation extremes. A specific area to watch out for are tropical cyclone precipitation and flood events. The IPCC strikes a good balance of what we know and where we need to spend more time. The insurance industry is watching these developments carefully, and I can see an interesting intellectual debate in almost every company we talk with.

What can we look forward to from KatRisk?

Global coverage for wind, storm surge, and inland flood data. We are working on other related perils as well, and we’re growing the company—recently moving to a new office in downtown Berkeley. Our models will cover ever more perils that are globally correlated with a changing climate in mind. From a product delivery perspective, we have been offering our location-level loss analytics as an API for some time and will expand that as well.

And just for fun…If your company was a superhero which would it be?

Catwoman (with a K though maybe?)

What would your theme song be?

The Rocky theme song...


Topics: flood data, flood risk assessment, catastrophe risk

Popular Posts