"Blockchain can bring insurance back to the basics"

Piloting new technologies to find out if there is value for Allianz, that is the responsibility of Dr. Markus Hablizel – Head of the Data Tribe @ Allianz Deutschland

With his team, he focuses on big data, analytics, and blockchain. With 400 followers in his intranet community, he enjoys discussing blockchain use cases and their implications for the insurance industry. Hablizel, a mathematician and computer scientist, believes that blockchain could become a standard for exchanging data and a standard for contracts that will make transactions faster and more cost-effective.

• What do you see as the potential for blockchain in insurance?

It’s not a given that blockchain will always be better than other digital solutions. I think companies should digitize the existing processes and then consider if they need blockchain. Sometimes, it is the best option and is helpful; in other cases, not. It can be that a centralized system is cheaper, faster, etc…

That said, blockchain could enable new business models that were not possible before. Think of peer-to-peer (P2P) business models. It can bring insurance back to the basics, where you look each other in the eyes and say: "If you have some damage, I will pay for it.” And the other way around. But there are some grey zones of applications.

This is why I wanted to start a blockchain roundtable with other insurance companies, which led to the creation of the B3i as we know it today.

• What is the potential of blockchain for insurance companies?

Blockchain technology has three layers of applications: intracompany, to achieve competitive efficiency advantages, industry-internal, to exchange crucial information, and across industries, focusing on the exchange of assets on meta blockchains, e.g. connecting energy ecosystems with smart home applications and e-mobility.

• You mentionend B3i. How is it going with the consortium?

There is already a very positive response to our discussions and requests from other insurers to join the consortium. In principle, as we deploy this, we want to make prototypes and solutions for the whole industry. There has been some progress on the Rüschlikon initiative, which aims to create a common standard for making back-office processes faster, more cost-efficient and streamlined between insurers so that customers can benefit from faster settlement times. Blockchain could possibly help here, too.

• What could still go wrong?

I see the most severe challenge as a possible loss of trust in blockchain technology which could lead to limited thinking about the opportunities it offers. In other words, the challenge is to constantly convince that the technology will deliver business value.

I am happy to say that Aegon, Allianz, Munich Re, Swiss Re, Zurich and some other insurers agreed to cooperate for a pilot project using anonymized transaction information and anonymized quantitative data in a proof-of-concept for inter-group retrocessions with blockchain technology. It's about one reinsurance company providing services to another. I’m really curious about the outcome.

What would a blockchain world look like if blockchain held all its promises?

The world would have many computing nodes that interact with each other. Every device or machine that has sensors, such as a smartphone, a fridge, or a car, could have a connection to a distributed network. It would be a world of different separate blockchains which could also be connected.

Sensors that are connected would talk to each other independently in an infrastructure of trust. But blockchain is not the answer to all the worlds' challenges.

Do you foresee a convergence of blockchain technologies or a coexistence of different types of blockchain?

There will be several blockchains for sure, many distributed networks. I imagine several blockchains per industry as well as meta blockchains that connect each of these blockchains to facilitate information or asset exchanges. One could also imagine community blockchains or private blockchains. Basically, they would emerge wherever they make sense. Hopefully, they will share some common standards which will allow for exchanges and interoperability between blockchains.

Perhaps it will be similar to standard protocols for the internet. Windows, Mac and Linux computers can still interact, even though the operating systems are quite different.

Are technical requirements higher in blockchain?

If every node has a full copy of the blockchain, then it does indeed increase your storage needs because you store it in thousands of places. But this is not yet a problem because today you could, for example, store the whole of Bitcoin on a smartphone.

From a computing perspective, there is no hurdle either. What might become a problem is the need for a standard, similar to an SAP standard software or the Stream Control Transmission Protocol (SCTP) standard. Today, there is no such standard, but it may appear in three to five years, and there may be a different one for corporate and private blockchain networks.

Today you just agree on a trusted party which gives you the standard; there is not a discussion on the standard itself. This is no limitation from a technical perspective – e.g. the algorithm, computing power or storage. But that might change in the future.

• What about quantum technology?

If a quantum computer were invented that can factor numbers very fast, you could break the blockchain cryptographies, the end of secure transactions or immutable records. There would have to be different kinds of algorithms for encryption based on quantum computing.

If someone said blockchain is 100% secure, I would disagree. Asymmetric cryptography is safe right now, but that is based on the fact that it is hard to factor numbers. Nobody has created a fast algorithm yet. Therefore, I see a risk that advanced quantum computing could mix with current private encryptions and present a security risk, or newer keys will need to be developed.

• Regarding immutability… what if you really want to change the content on a blockchain?

There are a few ideas circulating about this.

One is that you would not delete the original entry but add a history to make it transparent that there has been a correction. The original is still there, but it is amended in a way you can see.

• What about hiding transactions?

If you wanted to hide a transaction, you could utilize some white noise. The system itself will create some random set of transactions with no content, but an outsider cannot see if it is a random transaction or if it is a real transaction based on a random model. It is a concept called differential privacy. This is used to hide the existence of a transaction. The only party that can show which one is the right transaction is the one with the private key.

• And correcting?

The question is for what purpose you would want to alter the transaction? Normally, you would regard the corrected document as a newer version. The difficulty is when you want to hide that there has been a wrong transaction in the past. This is something you cannot easily alter. Then you would need to apply the concept of differential privacy with white noise. It is called plausible deniability. You can say you do not know if a transaction actually took place or not -- the case that it took place is as possible as that it did not take place. Nobody can distinguish. In a world of billions of transactions, white noise can easily hide the transaction you don’t want to be seen.

• What is your advice for companies wanting to start with blockchain?

First understand the technology. Experiment with it. Try to find the value of blockchain through use cases. Ask how it could apply to your industry, and how it could add value? We have about 50 use cases at Allianz, and we are testing more cases throughout the company. We are considering which ones would be easy to implement and the effects on the insurance industry as part of our prioritization of projects. Processes where intermediaries are involved are potential targets for a blockchain use case (e.g. when roadside assistance is paid for by a driver’s insurance, a smart contract could be used to make sure the right party pays). Anything that is managed today on a platform could potentially move onto blockchain.

I encourage companies to remember that blockchain is not only about technology. It is a mindset and culture. It requires openness to explore challenging opportunities.

Interview by Sebastian Steger and Stephan Janssens, Roland Berger, Munich, October 2016.