Nakamoto's original bitcoin paper describes how to make remote payments ‘without a trusted third party’ (like a bank) that are ‘computationally impractical to reverse’. The hype around bitcoin has largely gone away after massive frauds and losses, internal politics, EBA Authority and others published their risk analyses, the volatility became too much and other opportunities for manipulation and flaws became apparent.
However, the underlying distributed consensus algorithm now popularly called blockchain still is the subject of heated discussion and massive investments of time and money.
Recently I was fortunate enough to be invited to a panel discussion at the European Payments Council General Assembly whether blockchain is a short-lived illusion or a real game-changer.
Normally I am very much in favour of innovations, but in this case I lean towards the former option being cautious about blockchain for payments (although the discussion around this may be anything but short-lived). Please allow me to elaborate why.
(Spoiler alert: contrary to most of the on-the-one-hand-but-on-the-other-hand positions, I am here consciously taking a provocative, some may say one sided stance to sharpen the debate.)
Solution looking for a problem
Innovation should always put the customer first, not the technology. People are madly trying to work out what this blockchain solution could be used for. It is a solution looking for a problem, which is generally the wrong way around.
It isn’t new
Blockchain is often praised for its novelty. Actually distributed ledgers have been around for decades. People have been building distributed systems without central control since the seventies and eighties. This started when the Ethernet began connecting systems. In the nineties, when Internet became popular, many distributed systems were built including airline systems, and, not to forget, the Internet itself. It doesn’t have a central authority but consists of millions of distributed machines that are connected to each other. Also the financial system is very decentralized; Germany for instance uses many distributed platforms for its payments infrastructure (there is not a single ledger at the Bundesbank). The same goes for Visa. Thus distributed systems (and for technology geeks: the Byzantine Generals’ Problem of decentral trust, solving the distributed double spending problem) have been solved, and very efficiently, long ago.
It isn’t a good technology
Some say blockchain is a much better system and may drastically improve cost, security, speed, user friendliness compared to current existing systems. Increasingly experts agree that there is little evidence for this - especially regarding the public blockchain that would be needed for global payments. Its major pitfall is that it does not scale well (transaction limits, latency, storage explosion), uses extraordinary amounts of resources (energy, processing power), has severe security concerns and is even surprisingly bad at privacy. Should we really base a critical infrastructure like payments on this?
Of all the potential opportunities of blockchain, large scale systems such as payments are perhaps the least likely candidate for success. I say this not because I work for an ACH, but because I have worked on distributed systems since studying this at university. Seeing the nested iterative consensus algorithms embedded in blockchain (or Ripple, Stellar, Paxos or any of the other countless fragmentations of this scene) gives people with systems knowledge of scalable systems the absolute heebie-jeebies.
Objectively we must say that payments are already very highly scaled and efficient in Europe. And with recent advances (instant payments, mobile payments, etc.) will become even more so. Those looking for more likely problems to solve should consider looking at existing inefficient, often paper-based processes (e.g. trade finance, post trade securities, health care using private blockchains). If one wanted to improve these business areas, however, one could do it well totally without blockchain, but with existing proven technologies (like DDMS and PKI). Improvement in these areas is a matter of will, priorities, cooperation and investment, not technology.
It is becoming increasingly clear (as also conceded in the Nakamoto paper) that central entities are actually required in real world blockchain implementations. Thus blockchain will largely not even satisfy those ideologists in favour of a system without central points of trust, without central banks or who distrust capitalist systems on principle. This technology may continue to be popular amongst darknet clients wishing to purchase dubious goods without leaving too much of a trace. In the real economy we need to ensure compliance, KYC/AML/sanction checked payments, regulation and must provide an infrastructure that all can depend upon.
Maybe the main benefit of the technology discussion will turn out to be that we all think more closely about whether our systems cannot be improved more. Maybe it will also lead to more technology leadership in the financial services community. This means that we make our own agenda and prioritise our own decisions on business issues and technologies (blockchain, API, NFC, quantum, identity/authentication, wearables etc) rather than relying on external consultants, media and speculators with their agendas in promoting new ‘technologies’ to us.
In summary, let us continue to base a critical infrastructure like payments on a system that really works (i.e. that is scalable, compliant, cost effective, proven, that we understand). Let us continuously improve this (less paper, less cash, more mobile, instant service, consolidate infrastructure). Let us look at the real problems customers have (and there are many – but hardly the need for a distributed ledger algorithm), set priorities amongst them, focus on the important topics and solve real business problems. That way we have happy customers and earn money for it and live in a safe and efficient society.