“Bitcoin 2014 Panel: Economic Theory of Bitcoin” with time-based outline

It was an honor to be among the participants in this panel on 17 May 2014 at the Bitcoin Foundation Conference in Amsterdam. We addressed several issues that tend to recur in discussions of economic theory and bitcoin. The main topics were the regression theorem and bitcoin; bitcoin and the role of units of account and pricing; multiple value standards and the economics of altcoins relative to bitcoin; fractional-reserve banking, lending, and direct versus other-party control; and deflation and fixed versus elastic money supplies. I have added a time-based outline after the embedded video below to facilitate noting and locating particular topics.

Moderator: Jon Matonis (Executive Director, Bitcoin Foundation)

Speakers: Konrad Graf (Author & Investment Research Translator), Robert Sams (Founder, Cryptonomics), Peter Surda (Economist, Economicsofbitcoin.com, Robin Teigland (Associate Professor, Stockholm School of Economics)

1) Introductions, opening comments, and overview

00:00–03:05 Matonis: Introduction of panelists

03:05–07:57 Brief openings by each panelist

07:57–09:06 Economics profession and bitcoin

09:06–11:41 Matonis: Overview of topics

2) Regression theorem and bitcoin

11:41–12:12 Matonis: Introduction of topic

12:12–18:32 Surda: Liquidity, organized markets

18:32–23:16 Graf: Technical versus economic; theory versus history layers

23:16–23:50 Sams: Doubts this is relevant to bitcoin

3) Unit of account, price display, and price intuition

23:50–25:02 Matonis: Introduction of topic

25:02–27:00 Teigland: Depends on who; networks, sub-communities, generation change

27:00–27:23 Matonis: Can bitcoin overcome the existing network effect?

27:23–28:01 Surda: Uncharted area, dollar likely to remain unless deep negative event for it

4) Multiple value standards, room for 300 crytocurrencies

28:01–28:49 Matonis: Introduction of topic

28:49–31:01 Sams: Need distinct specializations; mining costs limit

31:01–32:48 Graf: Strong tendency toward one unit; only other very strong factors could counter

5) Fractional-reserve banking and bitcoin

32:48–33:41 Matonis: Introduction of topic

33:41–38:08 Surda: Money substitutes, transaction costs, price differentials, “reserve” standards

38:08–39:57 Teigland: Other non-traditional financing systems, crowdfunding, P2P lending

39:57–41:34 Sams: FRB based on an illusion, one that cannot be created with bitcoin

41:34–44:12 Graf: Bitcoin allows opt-out from all “trusted” 3rd, 4th, 5th parties. Vote with your mouse.

44:12–46:47 Sams: Who owns what? a pervasive issue; first bitcoin lending likely dollar denominated

6) Deflation, only 21 million units, number of decimal points

46:47–48:37 Matonis: Introduction of topic

48:37–49:46 Teigland: People adapt over time to situations

49:46–53:38 Sams: Deflation arguments misplaced; overheld, underused; other crypto money supplies possible

53:38–55:36 Surda: No need to change the quantity of money, but more to investigate

55:36–58:29 Graf: “Rising-value currency;” any quantity of money will do for society as a whole

58:29–59:26 Sams: Elastic supply could help stabilize exchange rate relative to fixed supply

59:26–59:46 Surda: Unit of account function depends on liquidity not volatility

7) Q&A

59:46–60:55 Q1: Banks allowed to create money; unfair playing field?

60:55–62:28 A1: Sams: 100% reserve banking; taking away private money creation privilege

62:28–62:56 A1b: Teigland: Local alternatives, experimentation

62:56–63:19 Q2: Isn’t buying and holding bitcoins already an investment in all of bitcoin?

63:19–64:06 A2: Sams: To some extent, but could be more with different money supply rule

64:06–65:00 Q3: Fixed rate of supply ignores recent lessons of monetary theory

65:00–65:27 A3: Matonis: Already addressed; Surda: May need to unlearn some of those lessons :-)

Expanded "On the origins of Bitcoin" paper with empirical supplements, other revisions

This is the 0.2 upgrade to my paper, “On the origins of Bitcoin: Stages of monetary evolution,” first released on 23 October 2013.

This expanded and revised version replaces the previous one from 11 days ago, but I expect this current version in this format to now hold steady. If you have kindly included the older file in an online reference collection, please consider replacing it with this one.

The changes are summarized in an included initial note to readers of the previous version. The most notable single change is the addition of two new sections as empirical supplements. They provide interpretations of patterns of events by “Bitcoin Year” (Appendix A) and a single five-year price-formation chart (Appendix B). Discussions in the main text of the precise timing of the first clear pattern of medium-of-exchange use have been clarified somewhat on this basis.

Download PDF:

On the origins of Bitcoin: Stages of monetary evolution (03.11.2013, expanded and revised)


Bitcoin, price denomination and fixed-rate fiat conversions

People are apparently still talking about the monetary regression theorem and its relationship to Bitcoin. There still seems to be a lot of confusion out there around both. Using a confused version of the regression theorem to criticize a confused version of what Bitcoin is does not seem like a promising recipe. I have been trying to focus on finishing up a longer work on Bitcoin and Austrian theory, but here for now are a few updated comments that came out of an online discussion today.

One newer point that has emerged in my work in progress is that the regression theorem is a theoretical explanation of how something that was at one time not money could ever become money in the first place. However, the theorem is not made to be a criterion of judgment for determining what is or is not money after the fact. Upon observing something actually functioning as a medium of exchange, the economist’s task is to explain how it came about. The role of the regression theorem was to explain specifically how something could have ever gotten started in a medium-of-exchange role to begin with. Judging and dismissing are unrelated to the function of the actual regression theorem. It is supposed to be explanatory and illuminating.

One area of confusion seems to surround the relationship between Bitcoin and fiat money, specifically the idea that Bitcoin has somehow emerged from fiat money, something like the way the euro got started on the backs or the various European national currencies. I addressed this briefly in my 27 February 2013 article, but here are some further observations.

Such transitional conversions are done with fixed exchange rates set by law. The new currency takes up its value from the old one in an administratively managed process. This applies to historical metallic coin monies giving rise to paper money certificates through a fixed conversion rate (later dropping the convertibility) and it applies to retiring paper monies being used to launch a new paper money, as in the case of the euro. However, the attempt to apply this translation/transition model to Bitcoin runs into serious trouble because no such transitional official fixed exchange rates have ever existed for Bitcoin. Quite the contrary. Governmental actors are only beginning to so much as roughly understand Bitcoin years after it already entered active use. It emerged on the market from scratch as its own good, certainly not from any official fiat.

It could be objected that regardless of origins, Bitcoin is only able to keep functioning through its relationship to fiat money and fiat money pricing. It is a mere strange shadow of the existing systems. Goods and services are priced in fiat money and a Bitcoin equivalent is paid. Bitcoins can be bought and sold referencing current market pricing on the most liquid exchange, Mt. Gox. In other words, this argument implies, Bitcoin could not function without these props.

This raises a number of interlocking issues. Bitcoin is now useful for many reasons, among them transferring value that may or may not have been obtained through the sale of fiat money and that might or might not end up being used to buy other fiat money in the future. On the other hand, while there are certainly active speculative traders on the exchanges, there are also folks buying Bitcoin with fiat money with no intention of selling it again into fiat money, but only of using it to buy goods and services in the more or less distant future. There are merchants using Bitpay so they never have to “touch” Bitcoin, but there are also merchants giving discounts for payment in Bitcoin, and accumulating the Bitcoin. There are consumers holding Bitcoin ready to use and other consumers that might only obtain specific amounts of Bitcoin for some specific purpose and then return to a zero balance. There could be some Bitcoin miners who mainly only ever sell Bitcoin for fiat money, but never buy any with fiat money. Everything is possible.

One point the Austrian school has long emphasized in monetary theory is that while money is special in certain ways, it is also a good itself, not a mere veiled marker or representation of other values. It is a type of good distinguished from other goods and services mainly by its higher marketability.

It is true that Bitcoin users have benefited greatly from the existence of market economies with functioning price structures. Pricing is still done for the most part in local fiat currencies and will probably continue to be unless and until Bitcoin becomes more stable in purchasing power than the fiat money that users are comparing it to, each in his own decision-making context. Automatic software price conversion makes it possible for the system to piggyback on existing and familiar price structures in each local area with immense convenience.

Yet I do not think there is any fundamental reason that Bitcoin-denominated pricing of goods and services could not evolve from scratch if it hypothetically had to. Fortunately, it does not have to. If no money existed at all, it would be necessary to get it going. We just have the convenience of already being able to rely on existing market prices for goods and services and the further convenience of being able to reference real-time market prices from organized exchanges. An argument could be made for just taking the easy road and using them. I think this is all just to the good of contextualized convenience and not so theoretically fundamental. Still, there are already Bitcoin-priced goods and services, particularly starting within the Bitcoin economy. For example, the Trezor Bitcoin hardware wallet is on pre-order for the price of 1 BTC.

The extent to which Bitcoin users reference fiat pricing in commerce is probably what has given rise to some  conflation with what I think is the quite different process by which one fiat money is converted into another by the official declaration of a fixed conversion price. Paper euros probably could never have taken off unless the official exchange rates with their predecessor currencies had been declared by law and the predecessor currencies had also been phased out by law. Without such official (“fiat”) declarations, printed euro notes would most likely either have been worthless or negatively valued due to the need to pay to store or dispose of them.

Bitcoin never had any official conversion price (or official anything), so how could it have gotten started? Bitcoin could never have begun to function in any other roles, such as transferring value derived from paper money over distance and converting into other paper money, if some initial users were not willing to trade any valuable goods or services for Bitcoin itself to begin with. After it began to be traded for other goods and services, one could observe it functioning in various, increasingly useful roles on that basis, some in interaction with existing monetary systems, but so long as its market price remained zero, it could not begin to serve in any such trading roles.

I think the initial-value question is probably much more narrow and technical than it is sometimes made out to be when the name of the regression theorem is invoked (the name; not necessarily the understanding). That question is how to explain a movement from a zero indirect-exchange value to non-zero indirect exchange value. Reaching non-zero from zero, especially in a digital computing context, is all that is needed for the rest to follow.

Anyone still talking about the regression theorem and Bitcoin might do well to focus on detailed historical research from the year 2009 and 2010 at the latest. After that, the deal was already done, leaving room only for efforts at explanation of what had happened. The rest was up to adoption, entrepreneurship and network-effect growth.

Now in audio on YouTube: Bitcoin and regression theorem article

I just got a message today from Graham Wright with a YouTube link. So I went over and there was my first article on Bitcoin, concerning the regression theorem, already done in a great audio version with accompanying slides. What a great surprise, one made possible by the magic of Creative Commons. Here it is!


IN-DEPTH | Bitcoins, the regression theorem, and that curious but unthreatening empirical world

I attempt to account for the emergence of bitcoins in terms of the monetary regression theorem. In doing so, I argue that 1) the existence of bitcoins does not and could not challenge the regression theorem and 2) the regression theorem does not constitute any particular problem for bitcoins in terms of economic theory. That said, 3) the investment analysis of bitcoins is a separate matter from the economic-theory analysis and is a good (but separate) topic for vigorous debate.
Read More