I’m an Englishman and a hacker, working at Twitter in San Francisco.
Wednesday 14th December, 2011
As an experiment, I’ve recently been investigating Bitcoin and its characteristics as a form of money. I’m not a fan of the use of Bitcoin in everyday transactions; clearing is far too slow for real-time use, and the concept of a fixed pool of circulating currency with diminishing returns on ‘mining’ is subject to a variant of Gresham’s law. Besides, my initial feeling is that the intricacies of the protocol currently make it appealing only to hackers/early adopters.
However, one aspect where I do like the idea of Bitcoin is as a reserve currency in a larger digital cash ecosystem. Bitcoin’s properties make it a bad circulating currency, but as a monetary reserve it’s similar to gold (fungible, divisible, scarce), with the additional benefit that an issuer’s reserves are completely verifiable by anyone in the network. This opens the market to competition: rather than requiring a secure underground storage facility in Zurich, with expensive monthly checks by ‘trusted’ independent auditors, all it takes to start a bank is a Bitcoin address and an encrypted keyring file.
For those who aren’t familiar with the concept, digital cash is a method of semi-centralized anonymous payment based on cryptography; specifically, it relies on the properties of blind signatures. A typical transaction looks like this:
Here I’m using the concept of an account balance with an institution, but really anything that’s redeemable could be the subject of a digital cash system, including train tickets and coupons.
Theoretically, a bank would have a public Bitcoin address representing its reserves, and then it would simply issue redeemable tokens against these reserves. It’s possible to bail in simply by transferring BTC to the bank’s reserve address; all accounts held at a bank would be identified with a public Bitcoin address. This makes bailing out easy, and banks could even charge a fee for bail-out (much like how withdrawing cash from a card often incurs a fee). The fact that the transactions for bailing in/out are ‘netted’ protects the addresses from the statistical attacks on anonymity that Bitcoin itself is susceptible to.
Different banks are able to accept (and even issue) each others’ currencies at open-market rates, with interbank clearing very similar to how existing banks operate, only based on the decentralized Bitcoin protocol rather than a single clearing house. If RSA-based blinded signatures are used, it’s possible to verify that a coin was issued by a bank without having to contact the bank, and thus coins are universally non-repudiable.
It’s likely a fractional reserve system would emerge; those preferring a full reserve system would probably just carry out raw Bitcoin trades. I have to side with George Selgin on this one; I think FR systems can be very valuable. Of course, because it would be a free and unregulated market, the reserve ratio would be elastic (and likely differ between institutions based on risk appetite). So one bank could offer you high interests with a corresponding high default risk, and another would offer low interest with low default risk.
Because these currencies aren’t linked to nation-states, the phrase ‘foreign exchange’ wouldn’t really apply; nevertheless, the exchange market would be a cornerstone of any digital cash ecosystem, due to the need for currency conversion services and the desire for a carry trade. When financial institutions are truly supra-jurisdictional entities, the concept of ‘money laundering’ gets a little hazy; when you’re using a dual-layer cryptocurrency, it’s downright mercurial. The complex, chaotic, but always beautiful nature of financial markets which operate on an ‘extranational’ scale is expounded by James Orlin Grabbe in his book International Financial Markets; unfortunately it’s a little dated now, and won’t be getting any updates (since its author is no longer metabolizing). However, the underlying principles of what trade looks like when it’s not controlled by a single government are still as relevant as ever, so IFM is recommended—nay, mandated—reading if you’re interested in building the economy of the future.
A few things stand in our way before we have a full economy up and running. Despite the importance of human psychology in the field of cryptography, the user interfaces and form factors we’re faced with for securing transactions and communications currently suck. I don’t believe it’s necessary to sacrifice security for usability; it just requires a modicum of thought and a generous helping of user empathy.
It’s also necessary to develop a secure, open-source digital cash system, based on the original designs but without infringing Chaum’s software patents (*eyeroll*). Lucre seems to come some way to solving this, but I’ve not investigated it in depth yet. We’ll also need simple clearing and ledger systems for bank operations and interbank clearing, though I don’t feel these pose a huge technical challenge.
More than just the ability to issue and redeem tokens, we need serious work on the user experience—we need to think about desktop, mobile and even paper transactions. If the system isn’t usable by ordinary ‘late adopters’, it’s worthless; financial systems (especially those based on Bitcoin) are subject to an extreme network effect.
There’s a discussion for this article over at Hacker News.