Interview with Torsten Leue on natural disasters, cyber attacks and GenAI

Kristian Kudela: Mr Leue, for centuries, the business model of the insurance industry has been to calculate risks. Are these risks becoming more difficult to calculate today due to climate change?

We protect our customers from financial risks. This means that we must have excellent risk calculation capabilities. In addition, the so-called tipping points will certainly increase. Ultimately, the question remains: Can a business model withstand the potential future burdens of climate change and other external factors? We believe that we are very well positioned for this: on the one hand, we are well diversified, and, on the other hand, we are resilient.

The combined ratio (loss-cost ratio) of Talanx AG is falling, but the frequency of natural disasters is increasing. The devastating forest fires around Los Angeles in January cost Talanx 624 million euros. Can the insurance industry develop sustainably?

Risk is our business – we are there for our customers in difficult situations. We can afford to cover large claims because we can balance the risk across our portfolio. This means that we must charge premiums that are commensurate with the risk and be broadly diversified. The frequency of earthquakes around the world does not increase when there is an earthquake in Chile for example.

Calculating probabilities means that we usually only have limited data available – and data is never sufficient. Of course, there could be more hurricanes now, but the decisive factor is where the insured primary risks and secondary risks lie. I believe that the insurance industry is well equipped to deal with higher risks in the long term.

Will it be necessary to consistently increase risk premiums in the long term?

I don't see that happening, especially not across the board. Every industry has insurance cycles based on claims experience. Looking at the Talanx Group's major loss budgets and the actual major losses over the past ten years, I can say that we have calculated well.

The question of the protection gap is, of course, also relevant. If we take the approximately €320 billion in losses caused by major natural disasters as an example, under 60 per cent of these were uninsured. This raises the question of which risks are insurable and how both traditional and alternative market capacities are provided.

Munich Re announced back in 2020 that it would reduce its risk appetite for secondary risks, such as forest fires. Will reinsurers raise prices to such an extent that they become uneconomical?

I am certain that, firstly, there will be cycles in the markets and, secondly, that sufficient capacity is available. Large events in particular require a lot of primary and reinsurance, and there is also a lot of alternative capital available. Cat bonds (investors use these products to bet against the occurrence of extreme natural events) in particular have established themselves in certain areas as a good complement to traditional reinsurance, especially where capacity is tight. We do not see a general risk that reinsurance could become unaffordable. However, I also assume that the protection gap will continue to exist or could increase in extremely vulnerable regions of the world.

Economists fear tipping points at which entire regions or neighbourhoods are no longer insurable. Do you share this assessment?

That may happen, but we have not observed it so far. For example, California is said to be a high-risk earthquake zone. An earthquake could happen next year – or not for another 100 years. We continue to offer insurance cover in California.

The actual tipping points, i.e. when they occur and how clear the patterns of natural events are, cannot be reliably predicted in terms of frequency or intensity. It is obvious to think that extreme events would increase, but the world is far more complex than that.

You mentioned the growing protection gap, i.e. the gap between total losses and insured losses. Are people losing confidence in the insurance industry?

I don't see people losing confidence in the insurance industry. We can compensate for financial losses, and we do that very well in the event of a claim. Some people may be surprised by the amount of the premium in exceptional cases. However, people understand very well when you explain to them where the sum comes from.

Take, for example, an area that is highly prone to natural disasters and, statistically speaking, could be affected again soon. If you rebuild your house there, you have to expect it to be flooded or damaged by storms again – and there are plenty of such regions around the world. What is important to me is that we have a major claims budget of €2.8 billion for 2025 within the Group. If our customers are affected by insured major claims, we pay accordingly. This applies not only to major claims, but as a matter of principle.

Talanx is active worldwide. Are insurance companies perceived differently in different regions?

In general, insurance markets depend not only on the mentality of a region, but also on the economic maturity of a country. These levels of maturity vary greatly around the world. That is why non-life insurance penetration in established markets such as large parts of Europe is around 3.2 per cent of GDP, while in emerging markets it is around half that.

This is precisely where we see enormous growth potential. In Latin America, for example, there is a market potential of around 100 billion US dollars over the next 15 years. We want to participate in this with our strong presence there. The decisive factor is when a middle class emerges in a society. Only then do opportunities for higher insurance density open up.

Hugh Terry, founder of the online platform Digital Insurer, takes a holistic view of the future role of the insurance industry: to support customers in reducing and managing risks. In other words, not just as a pure payment mechanism. Do you see it that way too?

I absolutely do. We can support our customers better simply because of our decentralised structure we are closer to them. Our customers benefit directly from our cost leadership in over 90 per cent of our portfolio.

If you look at corporate insurance, we work intensively with customers on preventive measures to make insurability more attractive. More prevention also flows into premium pricing. In dialogue with customers, it's all about building trust – precisely the kind of close exchange you mention.

Another holistic example: During the pandemic, it was unclear across the industry whether property damage caused by coronavirus was covered. We therefore decided to interpret these unclear cases in Germany in favour of our customers. This creates a lot of goodwill and trust among customers in the long term.

A holistic role for the insurance industry can also be supported by artificial intelligence – for example, in the areas of claim handling, underwriting or 24/7 customer interaction. How are you implementing ‘data-driven insurance’ at Talanx?

We have trained a large number of senior managers and employees in recent months. The key question here is how we can achieve scalability in our core processes, which have the greatest leverage. This is where our industry currently stands.

We are not a transactional industry where customers wake up in the morning and immediately want to know something about their insurance. It's different with banks, where they check their accounts every day. Our touchpoints – the ‘moments of truth’ – are rather limited throughout the year: for example, when they receive a bill, report a claim or look for a new contract. It is precisely at these points that we need to be available and offer an intuitive, simple and comprehensible process. AI helps us to do this.

AI, together with satellite images and 3D models, could be particularly helpful in the area of natural disasters.

We have an internal system called Argos, in which we map five million location data points alone. Our underwriters can import their locations into the platform using AI-supported import and then assess the overall risk before a contract is signed. Such tools are continuously being refined, including through AI. Ultimately, data is the new currency, as they say. The challenge for us is not only who has the better data, but also who can evaluate it correctly, i.e. intelligently and efficiently, in terms of stress analysis.

A few years ago, you emphasised that cyber security would become a major issue in the insurance industry. Do companies now understand that they need to invest more resources?

I see more need and increasing demand for cyber insurance. Nevertheless, there are unfortunately still too many companies that do not take the risk seriously enough, even though cyber-attacks have also increased dramatically in Europe. Imagine if a cloud infrastructure from one of the major providers failed – that would be like a hurricane for companies and insurers worldwide.

On the other hand, there are many small and medium-sized enterprises that are often underinsured, even though they are the ones who need cyber insurance the most. How can we offer them affordable insurance cover while they invest in firewalls and prevention? We want to participate in this growth potential in particular.

Interview with Torsten Leue (German)