Another bump on the road: Bitcoin and bubbles revisited

Japanese “bumpy road” sign for MtGox. 路面凹凸ありOverload, delays, and the temporary closure of the MtGox exchange seem to have been proximate triggers for a sharp Bitcoin correction on April 10–12 from dizzy highs.

As trading graphs fell freely, some Bitcoin critics appeared gleeful to believe that their prophecies of the Bitcoin phenomenon being nothing more than a delusional bubble might be coming true. Commenters promptly took to the internet to gloat at the short-term losses of naïve traders and bask in their own contrasting wisdom.

In this new context of short-term sentiment, it may be useful to revisit and refine my recent critique of the dismissal of Bitcoin as being nothing more than a bubble or even a sort of Ponzi scheme. In Hyper-monetization: Questioning the “Bitcoin bubble” bubble (6 April 2013), I offered an alternative to the popular interpretation of the long-term rise of Bitcoin’s exchange value relative to fiat money. This was especially intended to address the view that Bitcoin is nothing more than a bubble. The most insistent proponents of this view elaborate along these lines: “Bitcoin has no ‘intrinsic value’ and is therefore ultimately destined to fall to its ‘inherent’ value, which is zero, completely wiping out any true believers still left around for its inevitable and welcome extinction from the universe.” Or something like that.

“Is” versus “in”

A more subtle approach to calling a “Bitcoin bubble” is also available, and has long been advanced by several people with more nuanced understandings of the system. First, Gavin Andresen, lead developer of the open-source Bitcoin Project, wrote nearly three years ago in a short post Bubble and crashes (9 July 2010) that he expected multiple recurring bubbles over the course of several years.

Bitcoin will get mentioned someplace with lots of readers, a bunch of those readers will like the idea and try to buy Bitcoins, their price will rise, which will draw even more people to “invest”, which will drive the price up even more…until people decide that the price isn’t going to rise any more and everybody rushes to sell before the price drops. I predict there will be between one and five Bitcoin bubbles (price will double or more and then crash back down below the starting price) in the next four years…I think it will be impossible to tell if a bubble & crash is “natural” or “the men in black helicopters” manipulating the system.

Second, Rick Falkvinge, who had also called a short-term bubble and a correction to $60–$65, has long identified currency exchange services as a weak link in the wider Bitcoin “ecosystem.” See his 12 April post What we learn from this Bitcoin correction. A commenter on that post wrote, “I would not call an 80% move a correction…” to which Falkvinge replied, “It is not the downslope that is abnormal, it is the upslope. A value that reverts to where it was two weeks ago is normally a mild correction.”

Finally, Peter Šurda who steadily focuses on the importance of liquidity, infrastructure development, and scaling over price, re-summarized in a 12 April Facebook comment that:

My empirical research shows a correlation between media frenzy and price, and between liquidity and price volatility, while my theoretical research concludes that the price will fluctuate more rapidly than with more liquid media of exchange (i.e. what we are accustomed to as money, or even highly liquid goods such as stocks or commodities). The fluctuations will continue until Bitcoin’s liquidity increases significantly.

Such approaches have essentially been warning that, “Bitcoin may well be in a bubble phase,” adding, “one of several large ones, just as we expected to occur along the way.” As a commentary on the price trend in late March and early April, this appears to have been a valuable assessment. These observers recognized in advance that the price seemed to be rising at a pace unlikely to be sustainable, driven perhaps by events in Cyprus and then a flood of popular media attention.

In sum, saying that “Bitcoin is a bubble” (total dismissal of the system as such) and that “Bitcoin is (or was) in a bubble phase,” are quite different claims. Now that another correction has arrived, this distinction can come into better focus.

The Bitcoin system is not the same as the peripheral trading services

The core of the Bitcoin system itself, which few people seem to grasp is something entirely different from the more visible currency exchanges and their price charts, seem to have been relatively untroubled. This includes nodes, mining pools, the blockchain, wallets, and even informal and P2P markets. Besides MtGox, with its 12-hour mini-holiday, from what I noticed, only some of the exchanges and data chart services were heavily challenged and went offline intermittently. Bitcoincharts, for example, reported a 25x spike in concurrent online users from 2,000 to 50,000, requiring “tweaking the backend” systems in response.

The primary proximate cause of the crash, then, seems to have been the inability of a (currently) key exchange service provider to keep up with demand fed by sudden media attention and buy-in frenzy in the run-up, triggering a classic emotional wave of panic selling, most likely the corollary of the previous heat of emotional buying. The existing trading infrastructure (which is not the same as the Bitcoin system infrastructure) was not ready to scale to such a rapid demand spike. This sharp correction might be viewed in part as the rather ungentle method by which the market realigned itself with the current real-world state of scaling capabilities and business planning skills at exchanges that have been working to build themselves from the ground up.

Creation versus destruction

In the case of a classical terminal hyperinflationary event, the authorities orchestrating it are better equipped and prepared. Ink and paper are ready. Printing presses run and are up to their tasks. More importantly, printing plate engravers are standing ready to carve additional integers, a relatively simple task of creating higher and higher denominations of notes. The technical infrastructure is in place for state money monopolists to completely destroy the value of a paper currency, using “zeroes” to drive it all the way to “zero” and extinction.

Building a new kind of media of exchange for a community of all-volunteer users from scratch through peaceful cooperation, entrepreneurship, coordination, debate, and market ecosystem building would appear considerably more challenging than destroying a paper currency. After all, being constructive often seems more challenging than being destructive; it requires greater ingenuity and long-term persistence and perspective.

For additional articles on this topic, visit my Bitcoin Theory page on this site.

 

Short-term Bitcoin exchange trends as Rorschach tests of economic views

What might Hermann Rorscach (8 November 1884 – 1 April 1922) see in all this?Whatever the current excitement on any given day about short-term Bitcoin price charts, what stands out for me as a watcher of the theoretical underpinnings of discourse on these events is how they bring various economic-theory concepts out into view as the people who have them in mind use them to interpret current events. As Keynes famously put it: “even the most practical man of affairs is usually in the thrall of the ideas of some long-dead economist.”

I try to stay at least 98% positive in general; critique can be more tempting than contribution. That said, sometimes understanding can be advanced by considering contrasting examples. Here are three.

Realized/unrealized gains/losses

First, it appears that the distinction between realized and unrealized gains and losses could be kept more firmly in mind by most. The take-home point is that for all those Bitcoin market participants (whether bull or bear) who did not actually trade today or recently, nothing much actually happened during the rapid headline price changes on the exchanges (of course, the changes might lead some to adjust their future plans according to their various forward-looking judgments). What these exchange prices actually indicate is merely the current record of transactions that are occurring on those exchanges. It is only the collection of such discrete “real time” recorded exchange events that provide the data for the construction of the lines on the trend graphs.

At times of temporary disruption of access to websites, low visibility prevails. Pre-placed automated orders execute and panicking short-termers flee. The price graphs (as they become accessible after the disruption) show the actual trades and volumes at particular times during this course of events. The price graph on a given exchange indicates actual marginal activity, at given moments, on that exchange. These exchange services are specific businesses; not magical, instant oracles of “price” in general or “value” in general. In the event, the top two exchanges, it appears from preliminary reports, were most likely under active manipulation in a possibly partly orchestrated move (and do not forget that this may well have included both upward and downward elements).

What a difference a few days can make.

A spiral into absurdity

Second, the good old “deflationary spiral” fallacy (maybe market observers will get a couple days off from hearing that one now?) apparently misses the significance of the fact that BOTH a buyer and a seller are required to form any given data input for market price graphs. With no transactions occurring, which is what is posited in this imaginary world of “no one will sell when it is so valuable,” there is no “price” at all. Thus, the market “price” could not be too high, because it would be non-existent under the stated assumption that no transactions were occurring. Reductio ad absurdum. These exchange prices form an up-to-date historical record at a given time of actual recent buying and selling on a given exchange.

Look for the verb

Finally, it has been a little surprising just how many Bitcoin critics employ the concept of “intrinsic value” in making the claim that Bitcoin does not have any. It was members of the Austrian school of economics, with the so-called subjectivist marginal revolution, that had seemingly put the final nail in the coffin of this ancient economic fallacy. Beginning in the 1870s, Austrian school scholars began to re-emphasize the view that value is ultimately the result of valuation, in reference to the verb to value, which is an act that can only be performed by living people (We now understand that some of the Late Scholastics of Salamanca also had this point reasonably clear several centuries earlier). After Carl Menger, this view gained steam with Eugen Böhm-Bawerk and unmistakable clarity with Ludwig von Mises’s 1912 Theory of Money and Credit and later works. This further application of the subjective theory of value clarified that the concept of “intrinsic value” is ultimately incoherent, not only with regard to goods in general, but to media of exchange as well. Mises was rather strict on this point in TMC, pp. 61–62:

Our terminology should [help to overcome] the naive and confused popular conception of value that sees in the precious metals something “intrinsically” valuable and in paper credit money something necessarily anomalous. Scientifically, this terminology is perfectly useless and a source of endless misunderstanding and misrepresentation.

Note: I discuss value theory in simple terms with illustrative examples in Resolving the paradox of value. For a more detailed discussion, see IN-DEPTH | The sound of one bitcoin: Tangibility, scarcity, and a “hard-money” checklist.

For additional articles on this topic, visit my Bitcoin Theory page on this site.

 

Hyper-monetization: Questioning the "Bitcoin bubble" bubble

What is the opposite of this? Sweeping up in 1946 after the hyperinflation of the Hungarian pengő. Source: Wikimedia Commons, Magyar Nemzeti Múzeum Történeti Fényképtára, Budapest.

Many observers have likened the rise of Bitcoin to an asset bubble. It is so customary today to use the “bubble” word in articles about Bitcoin that there may in fact be a sort of “bubble” bubble.

Another less common word introduced in this context is hyper-deflation. Some say such a thing is horrible, others that it is great. I suggest a quite different possible interpretation of these events and a word to label them: hyper-monetization.

I first heard the term “hyper-deflation” (used in a positive sense) when Bitcoin was rising rapidly from the low thirties to the high thirties over a few days in early March (Yes, this was only a month ago). While a few specialists of a certain persuasion understand “deflation” to be a great thing for ordinary people (see, for example, my 30 March 2013 post, “A short Bitcoin commentary on Deflation and Liberty”), the word still has a public-relations problem. Along with some technical issues from its several possible definitions (price level changes versus quantity of money changes, for example), and negative interpretations in conventional economics circles, it just sounds depressing, regardless of the stated technical sense in which one attempts to use it.

The word “hyper-monetization” first occurred to me around that time as a more positive term, and perhaps a more accurate antonym for the catastrophic hyperinflations that have repeatedly killed off fiat paper monies throughout their history. A related term, “de-monetization,” denotes the process of a widely used medium of exchange ceasing to function as such.

A total hyperinflationary collapse is one way de-monetization can happen. Another type of historical example of de-monetization is “bimetallist” legal tender price-fixing schemes driving one precious metal, say silver, out of circulation in favor of another metal, say gold. Yet another historical example is when a pure fiat paper standard is created after monetary authorities permanently “suspend redemption” of their legal tender notes into the precious metals they had promised to deliver.

The opposite process of “monetization” denotes something that was not a money beginning to function as one. When euros took over the respective jobs of various European national currencies, euros monetized and the previous national currencies de-monetized. Now they are historical paper relics, but no longer function as monies.

In contrast to such a legal tender conversion/transition, however, something that gains exchange value from scratch on the open market (rather than taking up exchange value through a conversion)—and does so at a logarithmic pace—might then reasonably be described as being in a process of “hyper-monetization.”

The trouble with the “bubble” bubble

Bitcoin’s high historical and current price volatility is unquestioned. However, one problem with the “bubble” analysis is that in an asset bubble, certain fundamental matters are quite different. In a business cycle mania phase, prices of the most popular asset classes for that particular cycle are bid up as people pile their freshly printed fiat money and freshly produced fiat bank account digits into booming fields. Each party in this rush competes with all the others to acquire some of the bubbling assets. These people are misled by artificially low interest rates to bid up certain asset prices unsustainably, and this all eventually collapses, as described in Austrian business cycle theory.

However high the prices of bubble assets go, they do remain the same goods. In the case of a monetization event, though, the practical use-value of the trading unit (not only its price in terms of other goods or monies) actually does rise with the number of people using it and the depth of the market. To imagine how different this is from a classic asset bubble, it would be as if not only the price of bubble-era houses were rising, but also that their actual sought-after qualities as houses were improving spontaneously at the same time. Such houses might sprout new rooms with no one building them, with new paint jobs appearing mysteriously overnight without any painters having visited.

In this way, quite unlike the case of an asset bubble, the more people “pile into” a medium of exchange, the more valuable it actually is in its function as a medium of exchange from the point of view of its users. This is a separate matter from its price, as a few astute observers out there have so far already been noting.

This type of value has been likened to the use-value of a language rising the more people there are who can speak it. Another analogy would be to the use-value, from the point of view of each user, of a given social networking site rising the more people join it and the more they use it.

These are called network effects. In this case, the exchange value of the unit for each holder is directly related to each holder’s expectations of being able to use the unit in future exchanges (much like the value of knowing a language relates to one’s expectation of being able to communicate with it). This is in turn related to how many people accept the unit, how readily, and for what. It is important here to note, due to long-standing and common economic misconceptions, that the “future” in this sense is any future time—from five seconds from now to however many vaguely numbered years into the future a particular acting person might happen to have in mind.

When it comes to network-effect growth, the more the merrier. An analogy can be made not only to the rising stock price of a growing social networking site, but also, and more importantly, to the number of users of that site and how much it is used.

Check this box for a perspective shift

Yesterday, I saw a tweet from the insightful Bitcoin watcher Jonathan Waller. He wrote (enthusiastically, I think) that, “The bitcoin all-time chart is not even slightly sensible,” and linked to a chart [showing logorithmic growth shown on a linear scale].

This tweet got me thinking (yes, this is also a possible function of tweets). How can we make sense of this trend? Might taking some other perspective help?

This chart struck me as looking quite similar to a hyperinflation. However, instead of the exchange value of a trading unit plummeting toward the abyss as in an archetypal fiat paper-money collapse, Bitcoin has been doing the opposite.

Checking the log-scale box on the bitcoin price chart reveals a different picture. It shows a (so far) intuitively ascertainable long-term historical course with a large bump or two and some curves in the road. In this longer-term view, the exchange rate has been growing, not so much from one to two to three to four, as on a linear scale, but from 0.1 to 1 to 10 to 100. It has grown by several orders of magnitude during these couple of years.

Of course, the usual caveats must be quickly noted. “If present trends continue” can and often is infamously followed by them not doing so. But what might nevertheless be observed about this trend?

If one were somehow witnessing a phase in the first “hyper-monetization” in history, is this not more or less what one would expect to see?

Mark my words

The value of a paper money at the tail end of a hyperinflationary event is mainly the direct value of the physical paper (burning, wall-paper, etc.), but there is a more gradual build-up before the final collapse. The following chart is the price of Goldmarks in terms of Papiermarks from 1918–1923 in the Weimar Republic. This includes a steady logarithmic trend from 1918 to mid-1922. The exchange rate also moves from roughly 1 to 100 during those few years.

After that, however, the 1923 portion looks incomprehensible even on a log scale. As monetary authorities run the presses full speed and add new zeroes to denominations, a point is reached toward the end when the primary objective of market participants is to rid themselves of paper as quickly as possible before the last shred of exchange value evaporates.

The USD/BTC trend shows the price of Bitcoin against (also steadily depreciating) US dollars. This bears a certain similarity to the pre-1923 phases of the Weimar Papiermark/Goldmark chart. One difference is that the trend for Bitcoin from autumn-2010 to spring 2013 is the inverse of the trend for the ill-fated Papiermark from 1918 to mid-1922. In other words, for the years in question, the rise of Bitcoin’s relative exchange value shows a statistical pattern with similarities to the decline of the exchange value of the paper mark. Of course, the specific factors behind these events are quite different. In one case, the destruction was driven by ever increasing, arbitrary production of more units. In the other, the growth appears to be driven by voluntary adoption (with all its various motivations) and network effects.

If we were now actually witnessing early stages of an unprecedented hyper-monetization event, what might the top of such an event look like eventually? This is a fantastic and entirely speculative question and certainly invites the ever risky “if present trends continue” types of thinking. Looking toward the future should never be confused with looking into the past.

That said, during such a singularity-like event, were such a thing to be occurring, one might at some fairly early stage expect to see an Epic Rap Battles of History installment called, “Bitcoin vs. Fiat Money.” The key question would then soon become:

“Who won? You decide.”

For additional articles on this topic, visit my Bitcoin Theory page on this site.

[UPDATE: Seven months later, a new article including revised highlights of this article along with new material appeared: Hyper-monetization reloaded: Another round of bubble talk (7 November 2013).]