Lead Forensics


Stefan Golling: If we don’t achieve our margins now why should we stay?

04.11.21 AdvantageGo

This week’s episode of The Voice of Insurance podcast sees Stefan Golling, Member of the Board of Management of Munich Re, chat with Mark Geoghegan about a wide variety of topics ahead of the January 2022 renewals. From models to the use of algorithmic underwriting, ESG and data standardisation, this week’s episode covers quite a lot.

Stefan Golling serves on Munich Re’s Board of Management and until recently was its Chief Underwriter. He also has responsibility for Global Clients, the North America Division and oversees HSB and American Modern as well as the Lloyd’s and Bermuda markets.

As the conversation touches on the upcoming 1.1 renewals, Stefan and Mark chat about those unknown and significant events, such as Covid, and if current models are adequate to deal with another unexpected event. Commenting on models, Stefan says: “We want to use models, we want to use the scientific input, the expertise of scientists. Not every underwriter is a geological expert, or is a weather expert, so we want to make use of such models to form the best possible view of risk that is possible, but at the end, you also, of course, need to apply common sense.”

Mark asks about cyber modeling and if the industry can get a handle on such a dynamic risk. Stefan explains that cyber risk is much more dynamic than static risk, and that the insurance market has a role to play in preventing losses and helping customers recover from such a situation. However, with much less historical data to use for modeling, even with historical data, the dynamic nature of the risk could raise questions around that data. Commenting on this, Stefan says that the industry’s history of finding solutions to dynamic risk and providing adequate structures to deal with them is what sets the industry apart: “As the insurance industry, we have proven again and again that we can cope with such type of dynamic risk.”

Speaking further on the topic of models, Stefan comments: “We partner with professional modelling companies, with startups, so I think one thing is important, you need to accept that you need these partnerships, you will never be cleverer by yourself than maybe if you work together in an ecosystem where you bring different strengths, different access to data together.”

The Art of Underwriting

A lot has been written about the art of underwriting, with its special alchemy of tacit knowledge, experience, and data. Mark and Stefan discuss the use of data in today´s underwriting process, with Stefan commenting: “So, the art of underwriting, we need to take into account where do we have uncertainty in the data, where do we have uncertainty in the models. I don’t think that our problem is the models; I think we rather need to concentrate on doing proper underwriting.”

How far can algorithmic underwriting go up the value chain, asks Mark? Stefan replies: “I have been an underwriter for my first 20 years in this industry, and I think I’ll always stay an underwriter. Therefore, I am actually convinced that there will always be the role of an underwriter needed. At the same time, of course, it’s clear that those in insurance and reinsurance – additional data sources, artificial intelligence, algorithms, they will incrementally lead to a certain automation of at least parts of the underwriting process. Generally, I would say we expect a higher degree of such automated, algorithmic underwriting as you call it in primary insurance rather than reinsurance.”

Stefan also shares his views on the impact of Artificial Intelligence on the insurance value chain. “I think where artificial intelligence will help us is also maybe in sourcing additional data into our underwriting. We very often deal with very large and complex risks in reinsurance and assessment of those risks they require a lot of domain knowledge and a lot of experience, but maybe the algorithms will help us support underwriters to structure risk information, to observe the computer systems, to see the signals in the underlying data flow to then maybe enable a better risk assessment to develop, maybe, also new databases to augment the decision base of underwriters.”

Enjoy the podcast.

Latest Insights