Blockchain (distributed ledger) technology is at the height of the hype cycle. VC funding so far has exceeded
$1 billion. Banks, payment and clearing systems, government agencies and service providers have also invested strongly. Many of the largest banks are exploring, experimenting, testing and even piloting blockchain applications.
So will blockchain mirror the Internet in its revolutionary impact on banking? Let’s not get ahead of ourselves. There is no question that blockchain has genuine promise in certain areas of banking.
But even R3 CTO Richard Gendal Brown has reservations. See his recent “getting real” comments on R3’s new Corda platform. R3's Corda is
designed as a secure, reliable and fast mechanism for exchanging financial contracts. Many banking transactions are explicitly or implicitly contractual in nature. So Corda's approach is a promising one.
Parallels with Internet Payments
I am reminded of the earliest days of Internet payments (mid- to late-1990s). Financial Services Technology Consortium (FSTC) was an important player in exploring this brave new world. In 1997 I created and led the initial design phase for an FSTC project
called Bank Internet Payments Server (BIPS). This was designed to simplify correspondent banking. Its goal was low-cost, same-day payments, both domestic and cross-border. It
deployed XML-based message standards, strong authentication and encryption. It also included mechanisms for non-repudiability of payment instructions.
BIPS expressed the hope that the Internet would simplify payments. This in turn would drive out infrastructure cost. As a result, urgent and non-urgent payments would be accessible to consumers and businesses of all sizes. Although US-based, the vision
In practice the vision was hardly realized at all. Payments infrastructure actually became more complex, duplicative and expensive. And for customers, payments remained expensive for same-day, and slow for lower cost.
Through the 2000s, banks implemented web-based, and then mobile-based payment mechanisms. SWIFT initiated a switch to XML-based messaging and TCP/IP protocols. Today, initiating payments over the Internet (web and mobile) is commonplace. Yet the initial
promise of the Internet to simplify payments infrastructure was not realized. Payments initiation is certainly easier and more convenient. But clearing and settlement is still complex, duplicative, slow and expensive in most countries around the world.
What went wrong?
Was it a bad idea? No.
Was the need overstated? No.
The Internet payments infrastructure idea had many of the challenges facing blockchain implementations today. They include scalability, lack of (or too many) standards, and a variety of internal bank barriers to adoption. Above all, universal adoption of
new bank processing infrastructure is extremely difficult to accomplish.
Still, the blockchain experiments being carried out are interesting and potential useful. They will help with clearer business case statements, and identification of which projects are more viable on a large scale.
Blockchain Applications Being Tested (Not Just Talked About)
Before we go on, we should note that the examples given here are experiments, not production-ready solutions. Regulators are involved in some of them, but have not blessed them. Some are being tested with real data, but at only a fraction of required volumes.
Testing is mostly of the positive nature (will it work under good circumstances). Rigorous testing of exception and problem situations will be much more difficult. Changes will be required to operations, customer service, product development and regulatory
compliance. These are being chatted about in corridors, but not comprehensively reviewed. In other words we are still in a laboratory environment.
Any list of banking industry driven trials of blockchain applications currently under way would be out of date by the time you read it. Every week at least, there is a new press release announcing a prototype, proof of concept or pilot application for blockchain.
In a quick scan of the literature, I came across announcements for nearly 30 trials just completed, or still going on.
Application trials are heavily dominated by trading, but also several other areas of application. In the following brief sections, the links give examples but are by no means exhaustive.
In the area of securities trading, there are many publicized trials (and no doubt many more behind closed doors). Trials include the entire trading lifecycle, from issuance to clearing to settlement to reporting
and compliance. Instruments being tested include low-volume equities, bonds, commercial
paper. Exchanges conducting tests include those in Japan, Korea and Australia.
The include many banks, individually or in partnership. They partner with several infrastructure players, including DAH, Nomura, Currency Port, Chain, Fujitsu, Eris, Ethereum.
None of this is surprising. Trading activities need all the things distributed ledger technologies provide. This includes speed, non-repudiability, security, and transparency. Delays in trading cycles create significant counter-party risk. This may have
contributed to the late 2000’s financial crisis, though see below for some risks associated with trading too fast.
In the payments world, banks have been looking for ways to simplify, speed up and reduce cost of payments of all kinds for decades. The European Banking Authority has announced plans
for instant 24x7 Euro zone payments. The Federal Reserve has (rather late in the day) awarded a project to McKinsey to assess proposed ways
to modernize the US payments networks.
So it is not surprising that banks and payments providers are looking at blockchain for payments clearing and settlement. However, trials are not far advanced, other than in non-fiat digital currencies like Bitcoin. There are experiments in several areas: P2P
payments; crypto-currency settlement (into and out of fiat currencies); mobile
FX exchange; internal digital rewards currencies; and settlement of spot and forward FX contracts. All these
trials represent limited learning exercises, rather than full blown proofs of concept.
Arguably the most significant payments trial in terms of potential systemic impact is cross-border clearing and settlement by RBC and Ripple. But real implementation
even of a pilot is surely years away.
In the credit world, applications tend to have a trading flavor. For example, JP Morgan Chase and DAH are experimenting with loan sales. A small group of banks are working with
the DTCC, Markit and Axoni in a trial trading Credit Default Swaps. This could be a great example of derivative trading.
RBC decided to look for a safe and easy way to introduce a blockchain-based product to the market. They are considering trials of a card loyalty platform.
Export and import transactions can be very complex. They are unwieldy, prone to errors and delays, and vulnerable to use in Terrorist Financing and Money Laundering. Banks provide financing to the import or export party in the form of Letters of Credit.
These bridge the funding gap between manufacture and delivery.
In an international trade transaction, many contractual documents are exchanged. This includes (among others) invoices, bills of lading, and insurance certificates. Digitization these documents, and exchanging in a distributed ledger environment, could help
with several existing challenges.
Several banks are experimenting with and testing the use of blockchain for portions of the trade life cycle. Standard Chartered Bank and UBS are partnering with an initial focus on invoices. Bank
of America and an unnamed bank are also starting experiments yet to be described.
One interesting question is whether distributed ledger platforms can help with financial inclusion. This is the global drive for relevant, accessible and affordable financial services for the very poor. In general, driving down infrastructure and transaction
processing costs will help. But there are specific areas of focus as well.
Digital currencies are interesting in economies where the poor do not have easy access to brick and mortar bank branches or even ATMs. Several companies are exploring possibilities, but generally outside formal banking.
An acute challenge in many countries is the difficulty of proving collateral ownership, particularly land. Many countries do not have a reliable land registry. As a result, fraudulent claims of ownership are rife, with the result that banks are not able
to trust in loan collateral. Factom is a startup experimenting with use of a distributed ledger to record land title. They are starting with a national
land registry in Honduras in partnership with Epigraph. Property title searches are complex and costly in most countries. This is especially true. with commercial properties. A distributed ledger approach could have a
significant impact on real estate lending.
There is not much in the way of real world compliance trials yet, but this area will become hotter. In particular the KYC and CDD phases of AML regulations are more and more burdensome and
error-prone. As an example, a global distributed ledger for financial transactors would be tremendously helpful. SWIFT has initiated a KYC
registry service, which is in use by a number of banks. It is likely they will consider blockchain as an underlying technology in
the near future.
An Exercise in Realism – Barriers to Adoption
R3’s Richard Brown stresses a key element of overcoming the hype, saying
“The reality is that solutions based on selecting the design first and then trying to apply it to arbitrary problems never work out well.”
The worst possible outcome would be huge investment in solving the wrong problem.
Even if we have the right problem, and a great technology solution for it, we have a very long way to go. Several barriers have yet to be overcome, as several organizations have
expressed. Here are some of the key issues to be addressed:
- Business case. The most likely reason for failure is that there is not in fact a compelling business case. There may be a real business problem that is genuinely solved by blockchain technology. But it still may be difficult to demonstrate a sufficiently
high Return on Investment. This will be essential to warrant switching large investment from some other compelling need. Universal adoption will be especially hard. All participants will need to make an investment, and they will all have different business
- Universal Adoption. In a low volume, specialized trading environment, players will include the exchange, regulators, and a limited number of marketmakers, brokers and investors. For them all to use a common platform is not too great an expectation.
But replacing something like a global correspondent banking network would involve massive change. It would look rather like implementing the SWIFT network from scratch.
- Transaction volumes and complexity. Transaction volumes and complexity. If the blockchain is to maintain a permanent record of all transactions the shared ledger will grow continually. If the ledger has to handle high volumes (millions or
even billions of transactions per day) then performance will be very difficult to maintain. This is the scalability issue, applying particularly to a public blockchain (e.g. the underlying infrastructure
of Bitcoin). But it will also be true of private distributed ledgers running high volume applications (e.g. equity trading platform). Solutions are being proposed (e.g. BigChainDB)
but unlimited size is a tough nut to crack.
- Integration. There are few potential problems that a blockchain can solve in a vacuum. Most of the time the solution will need to be integrated with existing systems, some of them quite old. This isn’t just a matter of building interfaces. Systems
have been built with old style solutions to the business problem. The current solutions may be inadequate, but technology and business processes will need to change significantly with new approaches. In addition, secure connections between backend or middleware
systems and a permissioned blockchain may present challenges.
- Standards and Interoperability. The recent history of the banking industry is full of examples of competing standards, and networks that don't co-exist well. In the payments world, for example, the first XML-based message standards were proposed
in the late 1990s. Yet only now is there starting to be universal adoption of their successor ISO
- Regulation. Some applications of blockchain may operate inside existing regulations. In others there will be legal questions about validity of smart contracts in particular contexts. There will be concerns about AML/KYC implications, and digital
identities. There may also be questions around the applicability of legislation designed for older operating models. This is somewhat similar to the challenges that arose when banks first started to use digital signatures for documents. For example, questions
will arise along the lines of “Where is the digital asset? Who is the asset’s custodian?” If new legislation is required, the delays could run to several years.
There are others, but you get the idea.
Wrapping Up the Blockchain Hype
Blockchain, including the distributed ledger, smart contracts and associated technologies, is an exciting idea. There are, in fact, many major business problems that result from inefficient infrastructure. Blockchain may offer viable solutions for some of
The experimentation makes a ton of sense. It will lead to a much better understanding of the capabilities and limitations of blockchain in all its varieties. It will also encourage the beginnings of thinking about what it will take to put these solutions
In the end there will be a few compelling business cases. The best of these will be compelling to all parties, and these are the ones that will be implemented. Which ones will they be? Sorry, I’m not a gambling man!