EIOPA is calling for a Europe-wide Insurance Guarantee Scheme

Interview with Börsen-Zeitung with Petra Hielkema, conducted by Thomas List and Bernd Neubacher

Mrs Hielkema, what impact could the war in the Ukraine have on the insurance industry?

The first we did is to assess the direct impact. And here we see that there is very limited impact. The exposure of both insurance as well as the occupational pension sector to Russia, is low. It’s 0.1%. for the insurance sector and 0.2% for pensions. We think that in this situation it is very important that the industry implements the sanctions. At the same time, we are happy that insurers engaged in Ukraine are very active to support their people there and get them out. But they are also trying to continue as good as they can with paying claims.

How about insurers in the neighborhood of Ukraine?

Neighboring countries had to act quickly because of concerns relating to the green card insurance, and they did. In Poland, the four largest insurance companies stepped up and said they were taking the cover for the first 30 days of the green card. So both industry as well as supervisors are doing what they can.

What is the indirect impact of the war?

We have to monitor the markets very carefully. There is inflation, particularly because of oil and gas prices but also because of supply chain issues and food prices. A big risk for the insurance market in general are low yields combined with market shocks. We have pointed that out in our stress test last year. And the result was that our market can cope with that. But there is still the risk of the double hit scenario of a drop in the risk free rate in combination with an increase in the risk premia. That is a key risk we are focusing on.

Do you have to redesign the stress tests in the view of the war or is it covered by the double hit scenario?

I think it's the latter. It's good to know we just had a double hit scenario stress test in times of low yields where we did see that in the end, all participating entities had enough assets over liabilities after shock. Depending on how things develop we could identify market developments we want to include in the next stress test - in particular, inflation has our attention. There will be new stress tests for sure and then we need to discuss the scenarios we will choose. But this year the pension funds are on the agenda. So we do not expect an insurers’ stress test this year for insurance.

When is the next stress test scheduled?

Right now it is scheduled for 2024. However, there is another stress test that needs to be done: a joint stress test on climate change with the other European supervisory authorities and the ECB. The European Commission asked for that in their climate strategy. They would like to see a stress test on the impact of the many measures that are included in that strategy. I would not prefer to have an insurance stress test and a climate stress test at the same time. I think it will be either 2023 or 2024, but in the other year, we will do the other one.

Let’s go to another area of review - Solvency II. What are the major issues for you in the process of reviewing Solvency II?

Solvency II was a big change for industry, for the better. Now where do we need to look for improvement? Some of the advice that EIOPA gave is related to how would things work better in group supervision. More generally, first of all, we've done some proposals to recognize the long term character of the insurance industry. Also the European Commission is interested in finding finance for the green transition. We have found that particularly when it is long term, there is indeed some room to be more favorable yet prudent. We made some proposals on that issue.

And what is your second point?

We said you have to recognize that this framework was drafted between 2012 and 2014, and there was some delay before we started it. At that time, we did not expect that actually interest rates could go negative. We have to admit that the low yield environment changes the extrapolation. Finally, you would need to include some macro tools in the toolkit - that’s recovery and resolution and some other requirements.

Last year the European Commission came forward with its own ideas on Solvency II…

A: We were quite happy with the Commission’s proposals. They recognize market conditions and increase the macro toolkit. I think that's helpful. Of course we have some issues….

Which are…

The first one is the complete absence of an IGS, insurance guarantee schemes. I know in Germany you have that, but there are countries in Europe where there is no IGS. When is that relevant? Let me give you an example. There was a Danish insurer that was selling across Europe on the basis of freedom to provide services. When it went bankrupt only the policyholders in Denmark fell under the IGS. All policyholders outside of Denmark did not. It is difficult to explain to the public that the protection differs across Member States. We need to work on that.

What are you proposing?

I don't expect a full proposal to be included now, but I would like to have a timeline so that we continue to work on it and continue to discuss it.

What else should be changed in Solvency II regulation?

We have concerns that there is an issue between home and host that is not being solved properly. If there is a willing home supervisor and a willing host supervisor, we can solve the problem. But there are situations when not every party at the table is willing and ultimately, if the problem is not solved by the players, EIOPA can only watch and know that things are not going well.

Is there an example?

In November 2021 we published a notice of non-compliance of a national supervisor with our recommendation.

The National Bank of Slovakia did not comply with a recommendation to take supervisory actions against an undertaking with business in several countries.

What is key here is that when I say EIOPA, it is really the entire Board of Supervisors. It's the meeting that I chair with the twenty seven insurance supervisors, including BaFin. They all together established a recommendation is needed, that the situation is concerning and also that there is non-compliance.

What should be done?

If it comes this far, I would need more powers, for example, binding mediation to actually deal with the situation and ensure that consumers are protected.

In banking supervision the ECB is able take away the supervision of small banks from national authorities if it feels like. Would you mind having that power, too?

I think the question is: How much do we already want to move towards a single supervisor, like in banking. If you look at where we are right now, that is not going to happen in the near future, but it could be the direction of travel, especially in a situation in which more insurers are providing services to digital clients in a big European market out of one country. But that is not where we are right now. In the absence of that, it depends on the willingness of home and host to solve problems. And when there is no willingness, it stops and you can't act. For me, that's new because I used to be a national supervisor. Now I'm the chair of EIOPA and I have this limitation, which I think is not helpful for the policyholders or the consumers that might buy it. We need to solve it.

Anything else on Solvency II?

Yes. We want to provide a little bit more room for easing capital requirements when it comes to long term illiquid liabilities of investments. However, in the proposal of the European Commission some mid-term and short term investments and liabilities are also included. I think that is a concern because that makes the framework a lot less prudent.

Do you think that long term investments should be encouraged by lower risks charges?

Again, we have to be evidence based and really look at the risk charges out there. We came to the conclusion that if it's really long term a lower capital requirement would still be prudent. If you look at the much debated volatility adjustment…

…which is a measure to ensure the appropriate treatment of insurance products with long-term guarantees...

.…you recognize a certain percentage of the stress in the market. We said originally in Solvency II, that you can compensate 65% of the spread for all liabilities.  Now EIOPA said in its proposal, that you can compensate 85% of the spread if you have illiquid long term liabilities because you will not sell the assets tomorrow. Our concern is that the long term requirement is watered down in the European Commission's proposal to allow the 85% for all investments and liabilities and maybe even further in the dialogue with the other two institutions.

There is also the idea of strengthening green investments.

Yes, the European Commission wants to encourage long term investments because it can be invested in green. In 2018, they adopted a proposal to reduce capital requirements for investments in infrastructure in order to encourage those investments. But in a survey we did recently we saw there is not more investment in infrastructure. So let's monitor if it's really going to green.

There's a lot of discussion if you should encourage green investments by lower risk charges.

You must have a model that assesses the risks. If the risk is lower, you can reflect that in the model.

Climate change is a major issue with implications also for the insurance industry. Losses caused by natural catastrophes will rise significantly in the future. Can insurance industry take a risk that might be impossible to calculate?

So far, the industry has coped with natural catastrophe risk, together with the reinsurance which plays a very important role here as well. But if you look at the frequency and the intensity of events, you might well say that these losses will go up.

What does that mean in terms of insurability?

I think the first question to answer is what are the risks assuming that intensity and frequency will go up? What does that mean for the actual risk? And what are the drivers of these risks and what can we do with that? Can we still find other products available? If not, can we create products or are there limitations to what can be covered? But also: Are these products bought?

That was a problem at the German Ahr Valley flooding last year…

Only 37% of the people in the region of Rheinland-Pfalz had a nat cat cover. Why are people not taking cover? Currently EIOPA is assessing that. We also do a behavioral study. Why are people taking or not taking out insurance now? Depending on the size of the risk, you might have a situation where the risk becomes too big to be carried by the industry….

…which means…

…There are two ways to deal with that. First I believe insurers have a role to incentivize consumers to adapt to and to mitigate risks. But there could be a point where you need the help of the public. 

How could that look like?

There are different forms to do that. For example in some European countries in case of a drought the first 50% of crops lost are paid by insurance and any loss over 50% is covered by a public fund. Other solutions could be mandatory insurance or a pool or a reinsurance solution backed by the government. I think this question about insurability of certain systemic risks seems to become more urgent. It also holds for cyber and for a pandemic. 

What about exclusions of insurance cover? In the pandemic we had a lot of discussion on that in Germany.

If there are exclusions, it has to be very clear to the policyholder that the exclusion is introduced.

Cyber attacks are a fast growing risk for the industry, but also for government bodies or hospitals.

Cyber risk and underwriting cyber risk is a market that will grow. I think the way the products are offered, the demand will probably outweigh the products available. That has to do with the ability to assess the risk. 

Can the industry take cyber risk without incurring high losses?

This discussion often starts from the fact that the numbers are too big and then concluding that we would need a public private solution. But the political appetite, at least so far, has not yet resulted in a lot of action in that field. In my view you should start from what can be done. There should be ways to insure this risk, at least to a certain extent. You should do a thorough assessment of what is possible, but also recognize that probably there will be a point where this can no longer be privately solved…

…and that means…

We have to think about what kind of solutions could that be? I do think EIOPA can play a role in that discussion. We can balance the need to think creatively and helping society manage the risk, but at the same time remain prudent and do not underwrite risks you can no longer actually cover as an insurer.

You have talked about huge risks private companies are not able to bear. Do you think also about the risk terror attacks in a European context being put in a Public-Private-Partnership like Extremus in Germany?

We want to offer an analysis of what we're hoping for in this case, where the limits are and what options there are to deal with those systemic risks. Indeed, terrorism could be on that list as well. I think the market can draw a lot. I sometimes think the market too easily jumped into talking to government for a backstop and then think about what they can do. So I challenge them to first think about what they can do and remain relevant and only then start discussing a back stop. I do see the point that here might be limitations to private markets, even if they pool risk and reinsurance risks. That is a discussion we will have to have in the months and years to come.

With digitalisation and data mining risks can be calculated for smaller and smaller samples. Traditionally insurance is based on great numbers and collective risk bearing though. Is that principle outdated?

The insurance industry has had to work with and underwrite on the basis of this data for decades. With data and artificial intelligence becoming much more relevant, it sometimes even feels that people are discovering only now that insurance actually has to do something with data. Concerning risk pooling: Indeed there are ways to make the pools very small and be very specific on the risks. Here it will be key to look at the overall business model. Part of the model really is a success. Because there is risk pooling, you need a pool to carry some risks. And society is interested in those risks being carried in a proper way. That is one of the reasons why I truly believe that the model of insurance actually benefits society and can also move with us into the new digital time.

There is also the issue of private data and data protection.

At the same time there will be a lot of opportunities to analyze and use data in certain ways. So ethics will be crucial. EIOPA and a group of experts came to principles of digital ethics in which we very clearly indicate what is key. And it has to do with explainability, comparability, with data management, data risk management, but it also includes non-discrimination and fairness. Being conscious of how you are pooling risks is a big part of the thinking that fits into those principles. You can be very ethical and have very good moral compass. But then there is the competition and you also need to respond to it. So I think for supervisors it is a matter of very close monitoring and probably stepping up at some point and not just relying on self-regulation. So I think ultimately we need to work on more regulation and supervision in that area. At this point, the industry is still very conscious of the responsibility they have, and as such, we were able to develop these principles together.