Oxbow Partners House View May 2018: Blockchain
May 2, 2018 Greg Brown
Hype around blockchain is as strong as ever: ‘Blockchain a potential game-changer’, ‘Beazley to launch blockchain registry’ are just a few of the recent headlines. Whilst we have no doubt that it is an interesting technology, we do question to what extent it solves the insurance industry’s problems.
In this post we argue that the unique advantages of blockchain are not necessarily well aligned with the challenges faced by insurers. We believe that insurers should continue to focus on their challenges and identify potential solutions – both technology and otherwise – rather than necessarily finding problems for blockchain to solve.
Benefits of blockchain
There are three main benefits of blockchain:
- Elimination of the need for a trusted third party: Blockchain is a distributed ledger with each party involved ‘owning’ their own full copy of the blockchain. Parties work together to validate the content stored on it, so there is no need to rely on a third party to perform this validation.
- Immutable and secure: Once a block has been added to the chain it cannot be changed unless a majority of parties agree. The blockchain is encrypted end-to-end and everyone has a copy so hacking one copy has no impact on the whole system.
- Automated: Blockchain is inherently computer code and you can add automated functionality to it easily, for example a command to pay a claim immediately based on an event trigger.
These advantages sound attractive to insurers, but do they solve major challenges?
1. Third party problem continues to exist
Blockchain was designed to be the core technology for the bitcoin cryptocurrency. A key design principle is that no third party is required to authenticate transactions. Blockchain tackles this by crowdsourcing validation from the other holders of bitcoin. It would take just over 50% of bitcoin users to collude for fraud to occur as the 50+% could restrict new transactions being added and reverse previous transactions. With the number of bitcoin holders estimated to be over 10m, collusion is unlikely.
This benefit does not realistically attach to insurance:
- There are generally a small number of parties active in the insurance value-chain (e.g. brokers, insurers, reinsurers) so collusion by the majority is reasonably possible
- Most blockchain use cases involve a third party which develops and owns the technology (for example Ethereum, iXledger) so you are still dependent on a third party; they may not have rights over the data but they still have a control over the parties, much like the perceived downsides of the current ‘software house’ model
2. Trust is implied
Related to the point above, most insurance use cases assume trust, rather than mandating it. In these cases, participants involved in the blockchain are not required to validate new transactions. Participants can simply add new blocks onto the chain. This requires that you trust all parties on the blockchain. If this is true, then why (necessarily) use blockchain?
3. Automation is possible in traditional systems
A lot of commentators point to blockchain’s ability to enable smart contracts. Smart contracts auto-trigger clauses based on predefined events. For example, ‘smart’ travel insurance could automatically pay a claim when it receives data that your flight has been cancelled – like Axa’s Fizzy proposition.
This functionality could also be achieved with a traditional system, using an ‘if-then-else’ statement driven off a database. Just because blockchain can run code doesn’t mean it’s the best mechanism for running smart contracts.
4. Consensus is still a fundamental development issue
One area rarely mentioned by commentators is the need for consensus. When you’re building a blockchain use case in insurance, all corporate parties (e.g. brokers and (re)insurers) need to work together to agree all facets, for example data standards and process flows.
For small POCs where all parties may be aligned on objectives this consensus is relatively simple. Building larger-scale solutions across multiple parties becomes significantly more complex and an issue the (re)insurance industry has been trying to resolve for many years. Blockchain does not to help with this.
Additionally, this is where having an independent third-party can be helpful (the very thing blockchain is trying to do without). A third party can manage and negotiate on the different requirements of the parties involved and help come to consensus. Two good examples of this are the insurance data standards organisations of Polaris and ACORD.
What this means for (re)insurers
We are not suggesting that blockchain is fundamentally poor or flawed technology. In fact, it is well suited for certain use cases such as cryptocurrency.
It may still be the case that insurers find a use case where blockchain is the most suitable, scalable solution. But this is not currently the case.
Our advice to (re)insurers is to explore blockchain and understand its unique features – but be sceptical about claims that blockchain is the only or best technology solution for any single problem.
