Why Banks cannot Replicate the Blockchain

Today's post is one of the main reasons why I felt I had to start this blog. UBS (Union Bank of Switzerland) has published a report of 36 pages on Bitcoin which can be found here). I'm going to focus on page 31 section "Institutionalising Bitcoin". An excerpt from the report says:

"Instead, technologically, Bitcoin does provide a revolutionary new payment system. Bitcoin is already used as a cheap form of international money transfer - a market for which it arguably shows the most promise. In principle, financial institutions with existing anti-money laundering systems in place (like banks) could adopt a common Bitcoin-like technology to facilitate fast and secure international transfers between end-users, with fiat currencies as the unit of account (or possibly a digital currency serving purely as an intermediary), and minimal transaction costs. These institutions could also deal with bridging the confirmation time, by taking on the credit (and possibly FX) risk involved. By keeping track of users, they can also penalise or chase up those who attempt to defraud the system. This could be facilitated by new or existing entrants, although the question then becomes, given the costs associated with running a payments system, whether existing players like Visa or MasterCard could be out-priced. We elaborate on this in the first section of the note.

With regard to technological issues, such as the 51% problem, the 10 minute transaction time, or the fact that Bitcoin is inordinately wasteful of electricity and computing resources - these are all first-generation problems that do have solutions. For instance, other mathematical processes can be used to discourage an arms race in mining, other consensus-building techniques exist to maintain decentralised transaction ledgers, and confirmation times can be reduced. Some of these are already being implemented in alternative digital currencies."

I will now explain why the distributed ledger concept, called the blockchain, will never work for banks in the way they want it.

First, it is important to understand WHY Bitcoin works. It works because there's no single point of failure. The system is setup so that if at least 51% of the participants in the network are not crooks, then the network works flawlessly. We don't even need to know who the other 49% are.

Second, suppose that we would like to copy/mimic the Bitcoin blockchain approach backed by a fiat currency, or that its money supply can be regulated. Presumably the banks themselves would like to inject cash on by demand into the system. This means that we now abandon one of the most fundamental aspects of Bitcoin; the deterministic finite money supply. Unfortunately for the banks, this is impossible, because it will break the whole system by bypassing the very first requirement; violating the principle of single point of failure. Why?

Today, new bitcoins are created as a reward mechanism to the miners who process transactions. The transaction bundle (called a block) requires a verification process which is immensely computationally complicated. The reason for why it is immensely computationally complicated is not to waste electricity (which UBS seem to think), but guarantees protection against counterfeiting. By piggybacking new bitcoins in this process means that no one can easily create new money and inject that into the system. If that would have been possible, then bitcoin would have collapsed many years ago.

If UBS wants to replicate the blockchain approach, but have a separate system to inject fresh cash into the system, then there must be an outside mechanism (not part of the blockchain itself) for doing so. Presumably, there will be some kind of "digital stamp" that validates new money injected into the system. The problem is that how do you know whether this digital stamp is on the run or not? Just the rumor that the digital stamp is on the run will collapse the system.

Some say that the centralized DNS system (the internet system to map domain names to IP addresses) has been running just fine even though it has this intrinsic single point of failure. But there's an immensely important difference between DNS and a financial system with a controlled money supply. One is money (as long as there's money there'll be scammers/robbers, no matter what regulation is put in place), the other is the scale. Here we need thousands of banks being able to inject fresh cash into the system (so they can clear their balance sheets whenever a customer wants a new loan), and thousands are orders of magnitude greater than the number of people involved to protect the central DNS system. But, says someone, this is how the current system works today, yes, but the current system doesn't have a public ledger which is precisely why transactions are so cheap in the bitcoin ecosystem. What I'm saying is that the blockchain approach is inherently incompatible with a human controlled money supply.

Therefore, there's no hope for a private institution, or a government controlled one, to build a blockchain approach to compete with Bitcoin. Such an approach has the intrinsic single point of failure which sooner or later will collapse the system. Bitcoin works because it is ingeniously designed so that there's never a single point of failure. My suggestion to the banks is to instead of inventing some crappy system themselves, is to embrace Bitcoin and start using it.

Internet is a free public network where everybody can participate. No one is foolish enough to try to build their own network and try to out-compete Internet. So my advice to our banks is to not try to fight Internet, sorry, Bitcoin, and start using it. Bitcoin will be the superior payment system and currency on this planet whether you like it or not. The sooner you jump on board this ship, the more power you'll get. The longer you wait the less power you'll get in this new world order. Are you on board yet? I am.

Datavetaren