Blockchain - Beyond the hype

The word "blockchain" is omni-present. Working myself in the financial services industry, I often get questions if particular problems should not be solved by "the" blockchain. This gives the impression as if the blockchain is a magic solution to any problem, but this article aims to demonstrate that blockchain - although an interesting and promising technology - will mainly accommodate specific problems and still requires considerable technological evolution before being viable at large scale.

Blockchain has complex cryptographic foundations, meaning that only a few persons working for blockchain companies will have a fully detailed understanding of the technology. Currently everyone tries to understand (and talk about) the technology, but once it will become a commodity, the details of the technology will be as hidden as the complex query optimization algorithms in a database.
Believing that blockchain will change the world we live in is therefore exaggerated, as all it really does is help companies reduce costs(reducing manual bureaucratic operations) and achieve regulatory compliance.
Blockchain can therefore only be considered as a success, when nobody talks about it anymore. At that moment blockchain will have become a commodity technology in the background, hidden to the business (like e.g. a database, a message queue or OAuth2).

Today blockchain is mostly known in the context of crypto-currencies and more specifically for its usage in Bitcoin, which causes a lot of people to mix up both terms. Bitcoin is a virtual crypto currency, using blockchain as the underlying technology for managing the accounting of the currency transactions and positions in a decentralized and secure way. The merit of blockchain is that it resolved the need for a digital currency, which is not managed by a central party, but still ensures that each currency is unique (avoid counterfeit), can be transferred and can only be owned by 1 person at the time. This was particularly hard to achieve in the digital world, as bits can be reproduced very easily.

The technology of blockchain (i.e. Distributed Ledger Technology) can however be used for any use case, where:

• Data should be stored in a database in a secure, non-editable and decentralized way (unique source of truth). This data can be anything, e.g. payments, financial transactions and positions, medical data, personal data, online advertisement…​
• Data should be exchanged between different parties
• Multiple parties need shared write access
• The different parties don’t trust each other and/or interests are not unified (in blockchain trust is established by technology)

In other words, blockchains are ideal as a shared database(distributed database running on multiple computers globally) in which every user is able to read everything, but no single user (or organisation) controls who can write what. As every node in the network has a copy of the Blockchain, it is impossible for any party to modify or remove data, i.e. data is immutable once it is on the blockchain.

The industry of blockchain is clearly hyping (although the fall in the price of crypto-currencies has dampened the enthusiasm). Enormous capitals have been invested in blockchain based start-ups, nearly every multinational bank is experimenting with blockchain, public blockchains are being created almost on a daily basis (Ethereum, Lisk, NXT, Ripple, Stellar, Bitshares, Tezos, NEM, Z-Cash…​) and several large alliances have been created, e.g.

• Hyperledger project of the Enterprise Ethereum Alliance: groups together 120 major firms (mainly banks and tech companies) to create a permission-based private blockchain.
• R3 project: groups together 42 of the world’s largest banks and experiments with Microsoft’s Azure-based Blockchain-as-a-Service to develop Erethreum as a bank-to-bank global transaction system.

As any data can be stored on a blockchain, the use cases for blockchain are almost endless, resulting in start-ups working on blockchain based solutions for exchanging any kind of data:

• Documents (invoices, guarantees, work orders…​)
• Public records (land titles, vehicle registries, criminal records, passports, birth certificates, building permits, court records…​), with the most interesting use cases in financial services industry for
  • Federated Identity: manage identity and personal data of customers
  • AML / KYC: according to a Thomson Reuters Survey, financial institutions spend on average $60 million on KYC and customer due diligence while some banks spend up to $500 million per year. This effort could be reduced considerably by placing AML/KYC info on a blockchain and sharing it between financial institutions.
• Private Records (contracts, signatures, wills…​), with for the financial services industry mainly
  • Trade Financing contracts:
    • Execute commercial transactions and agreements automatically
    • Enforce the obligations of all parties in a digital contract
    • Mixed with IoT (e.g. asset tracking sensors) for automating the end-to-end process
  • Payments (national or international), allowing instant payments worldwide
  • Securities trading, with real-time settlement (instead of D+3)
• Semi-Public Records (certifications, HR records, medical records, delivery records…​)
• Physical Asset keys (home keys, hotel keys, car keys…​)
• Intangibles (coupons, reservations, patents, copyrights, software licenses, domain names…​)

Important for those use cases is that the data/information becomes much more dynamic (alive), thanks to the blockchain technology i.e.

• Programmable automated actions (= rules) can be linked to the data/information. These actions can be programmed directly in the Blockchain and can be executed automatically when certain conditions are met (i.e. so-called smart contracts). E.g. discount cheques, loyalty points…​ which can only be used for certain services/products under specific conditions.
• Owner of the data is much more in control of how his data is processed (i.e. viewed, modified, shared…​). In an existing database system, the database owner has absolute control over the data held in the database. However, in a Blockchain system, ownership is maintained by the creator of the data. This is perfectly in line with new data privacy regulations (GDPR).

Today the use cases for which blockchain is relevant are solved by a (centralised) trusted third-party (e.g. a bank, a notary, the government, clearing houses like Euroclear, a big tech giant like Google, Ebay, Facebook, Apple…​) which takes care of the secure storage, the data exchange management and the establishment of trust (e.g. by providing collaterals). Blockchains offer a way to replace these trusted third-parties with a distributed database, locked down by cryptography.

As with any new and hyped technology, there is (was) a gold rush going on. Especially in the financial services industry, people are trying to use the blockchain for any use case. Unfortunately, any use case where blockchain is used, which is outside the above context, is likely to be an overkill, as blockchain comes with a significant increase in complexity and quite some downsides regarding non-functional and functional capabilities, compared to a traditional centralized database (like Oracle or MySQL).
Non-functional capabilities like performance, scalability, monitoring, data confidentiality / privacy, fast data retrieval, integration with legacy banking systems, regulatory uncertainty, process to upgrade blockchains (as no central control) …​ are still (largely) unsolved problems. Some of those problems are even inherent to the technology, e.g. Blockchains will always be slower than centralized databases, as Blockchains have additional burdens like signature verification, consensus agreement and node redundancy (i.e. any transaction is processed independently by every node in the network).

This combined with the limited functional capabilities existing in the market (e.g. very few solutions to do analytics and machine learning on data stored in a blockchain), makes that companies should analyse very carefully if the added value of using Blockchain sufficiently outweighs the disadvantages. E.g. using Blockchain to store or exchange non-critical data internally in the company makes little sense. If trust and robustness aren’t an issue, there’s nothing a Blockchain can do that a regular database cannot.

Furthermore, even for use cases where Blockchain does make sense, it is important to consider that the technology is not sufficiently mature yet for large-scale usage. Even though Bitcoin is clearly a huge success, the scalability is still a major issue, i.e. the processing through-put of the Bitcoin network doesn’t come near to the through-puts of the SWIFT network. Add to this the expensive nature of the consensus agreement activity and the tendency for centralisation of the verification and consensus agreement steps, and you realize that the risk for banks to adopt this technology in their critical processes is still significant.

When those barriers are passed, there is still an issue about ownership of the code of the blockchain. For Bitcoin the code is open-source and maintained by an open-source community of active developers. For less public use cases than crypto-currencies, this will be more difficult. When a company wants to build out a new use case based on a blockchain, typically this company will finance the implementation and build and maintain this blockchain (and the added-value services on top). This makes that trust in this company is still a pre-requisite (even if code is open-sourced, companies on the blockchain will not dig into the code to find issues, nor will they verify if code of the blockchain is the same as the one on the open-source code repository).

But even if those technical arguments don’t convince you, you should still reflect on the operational and business constraints. Indeed, blockchains allow to remove the trusted third-party from an IT point of view, but often these trusted third-parties are providing more services than just the storing and transferring of digital data. As these other (value-added) services can typically not be replaced by a blockchain, organisations still need to put their trust in these third-parties. The question is raised then that if you can trust them for those services, why not trust them for the correct storing and transferring of data. Common examples of third-parties predicted to become obsolete by blockchain are:

• Notaries: all notarized contracts could be put on a blockchain, but then you still need someone to create (write) the contract and provide advise to customers. Furtermore even if a company would create a blockchain for these contracts, it should still be legally backed by the government, otherwise there is still no guarantee (if a court doesn’t accept the blockchain as proof of ownership, the blockchain has no real value for its users).
• Auctions: while the auction process could indeed be put on a blockchain, you still need a firm to collect the objects, describe the objects and make publicity about the auction.
• Financial clearing houses: while the clearing and settlement process could indeed be put on a blockchain, those clearing houses provide also a lot of additional services, like collateral management, covering the counterparty risk, liquidity management…​ which cannot easily be provided by blockchains.

Overall, banks and insurance companies should be aware of the enormous potential of blockchain, but also of the fact that it is still immature and has significant drawbacks. It is interesting therefore for them to start experimenting with it (e.g. in an innovations department), but they should avoid rushing into large blockchain based projects, without validating the business case. Furthermore, they shouldn’t forget that even when getting to a working prototype in a matter of months, getting to an actual bullet-proof production-ready service accepted by regulators will take much longer.

With all the challenges banks and insurance companies are facing today (i.e. regulatory pressure, switch from batch to real-time processing, cyber-security, cost of maintaining and adapting legacy systems, lack of worldwide integration…​), it is normal that managers of financial service companies look for silver-bullets to help them face those challenges. Question remains that these challenges cannot be easier solved by improving the existing systems and using traditional technologies rather than by setting up a completely new system based on a still largely experimental technology like blockchain.