Gold standards, optionality, and parallel metallic- and crypto-coin circulations

Source: Biswarup Ganguly, Wikimedia Commons. Copper coin, 1782-1799 CE, Tipu Sultan ReignWhen one hears the words “gold standard,” it is usually either from people who think it was a horrible thing or people who think it was a wonderful thing. However, many in both groups seem to agree that “the” gold standard represents the free market money of the good old days, or the bad old days, or perhaps even the future.

However, the inclusion of the word “standard” could already serve as a warning that this may have been just another convoluted sequence of confused government programs. Looking into this more closely may suggest lessons for cryptocurrencies today.

Several different international monetary orders from 1871–1971 were based on gold: the classical gold standard, the gold exchange standard, and the Bretton Woods system. Yet these came only after a long series of previous legal interventions in money of various types. When such legal measures were absent or weaker, things tended to differ. Professor Guido Hülsmann characterizes it broadly this way on p. 46 of The Ethics of Money Production:

In the Middle Ages, gold, silver, and copper coins, as well as alloys thereof, circulated in overlapping exchange networks. At most times and places in the history of Western Europe, silver coins were most widespread and dominant in daily payments, whereas gold coins were used for larger payments, and copper coins in very small transactions. In ancient times too, this was the normal state of affairs.

One dramatic way that monetary metals were driven out of circulation was the policy of bimetallism. People we might today call “regulators” legally fixed the exchange rate between silver coins and gold coins to make the market more “regular.” The actual result was the rapid loss of a major component of the money supply from circulation. Hülsmann on p. 130:

One famous case in which bimetallism entailed fiat inflation-deflation was the British currency reform of 1717, when Isaac Newton was Master of the Mint. Newton proposed a fiat exchange rate between the (gold) guinea and the (silver) shilling very much equal to the going market rate. Yet parliament, ostensibly to “round up” the exchange rate of gold, decreed a fiat exchange rate that was significantly higher than the market rate. And then some well-positioned men helped the British citizens to replace their silver currency with a gold currency.

Hülsmann then cites similar cases in the US in 1792 and 1834. Not only did price fixing not make the market more “regular” as intended, it caused severe disruptions, with many losers, some winners, and a certain period of monopoly metal circulation.

The parallel circulation of metals may in this way have represented relatively more of a “free market money” situation than government orchestrated gold standards that arrived only after long sequences of legal manipulations—and which just happened to also channel the majority of gold into the vaults of monetary-system orchestrators.

Lessons for parallel cryptocoin circulations?

Such parallel circulation has been used as an analogy to promote parallel cryptocurrencies in a complementary monetary role. How well does this analogy hold up?

Each metal filled a different market role from the others, with some overlap. Likewise, each altcoin advertises different features. How significant will users perceive such differences to be?

The main difference between copper, silver, and gold was a large distinction in a practical characteristic, one unmistakeably clear and important to the end user—exchange value per unit of weight. A single gold coin could do the work of a handful of silver ones or a hefty pile of copper ones, whereas buying a few potatoes with gold instead of copper would have been quite a technical challenge in the opposite way.

However, this particular factor—probably the most important one from the case of metals—does not apply to cryptocurrencies, which can be divided and combined freely and have no weight. Perhaps some other factors will prove significant enough to create a similar degree of differentiation, but the final say goes to the market test, not the engineering imagination. Another significant difference among cryptocurrencies is the amount of hashing power protecting each chain. This is a factor, in contast, for which minimal significant parallel exists in the case of monetary metals (the closest thing would probably be relative differences in forgeability).

In considering a given cryptocoin from a monetary viewpoint, it is important to investigate and consider its actual patterns of use. Having the word “coin” in the name does not make it a monetary unit. What does? One sign is the extent and scale to which users are holding a unit so as to buy goods and services with it. This might contrast, for example, with an income purpose (buying and selling the asset against another monetary unit in pursuit of monetary gains), or social-signaling purposes such as giving out microtips to online commenters. Each altcoin or appcoin might fill different roles and provide different kinds of value to users, perhaps within particular sub-cultures, or perhaps in the context of particular services. Coins can apparently fill some of these functions without having to gain much traction in a more general monetary role.

In contrast to this, a central function of holding cash and other liquid balances is to address the uncertainty of the future and this is a general function—the more general, the better fulfilled. For example, we may know that we will want to buy some things in the future, but not necessarily know exactly which things, when, where, and at precisely what prices. Cash balances, due to their flexibility, enable us to adjust to such constellations of uncertainties. In this sense, a unit that is more widely accepted is likely to come in handy in a wider range of such future situations than one that is less widely accepted (there are also other factors to consider besides generality of acceptance, such as whether the units are expected to tend to gain or lose value while being held in balances).

I suspect that only significant traction in such a general monetary use, such as bitcoin has begun to gain, could sustain a large increase in a given unit’s purchasing power over the longer term through the network-effect process I have termed hyper-monetization.

There is a strong tendency in a trading network toward the use of a single monetary unit. This theoretical insight has sometimes been extended to the historical claim that this is the natural role of gold, or the forward-looking claim that gold should fill this role in an ideal future. However, other factors also push back in the opposite direction toward parallel circulations and multiple options. Such factors could be natural, such as we saw with large practical differences among different monetary metals, or political, such as the legal favoring of some monies in combination with the geographic sectioning off of the total trading universe.

One option is not really an option

Finally, adaptive systems and species that survive for a very long time tend to have some redundancies in critical systems. There is no single more critical system for the functioning of civilization than indirect exchange using money and other monetary units. A repeated theme in the history of money, however, has been actions by rulers that have the effect, whether intended or not in any given case, of removing alternatives and opt-out paths for money users, leaving them highly vulnerable to whatever happens with the remaining monopoly unit.

If a society has a single dominant monetary unit for whatever reason, it would seem favorable from this larger vulnerability assessment or antifragility perspective for its members to have other viable options at least waiting in the wings in parallel operation. Use of a single money certainly has strong advantages, but while network effects and broadness of acceptance are very large factors, they should not be mistaken for being the only ones.

In particular, use of one unit with no alternatives available does not address the need for adaptation to unexpected events. The complete absence of freely chooseable and ready alternatives makes a society more vulnerable to the effects of large-scale shocks. Points often lost on central planners of all schools are that redundancies and parallel options tend to have unexpected very long-term survival value, that more options are often better than fewer, and that having only one “option” is similar to having no option at all.

Recommended related books:

Jörg Guido Hülsmann, The Ethics of Money Production (2008)

Nassim Nicholas Taleb, Antifragile: Things that Gain from Disorder (2012)

The helpful fable of the "bitcoin": Duality models revisited

Bitcoin is many things, all referenced under the same word. Confusion about its nature and valuation naturally arises from insufficient differentiation of these facets, combined with a general human tendency toward “either/or” thinking. Often, the situation is more “both/and,” which becomes clearer after looking through first impressions and simple or even misleading analogies.

A short section of Francis Pouliot’s 17 May 2014 post on the Bitcoin Foundation of Canada blog caught my attention: “The currency and the network, although conceptually different things, cannot be separated. Bitcoin the network is valuable in itself because of its characteristics and, because you need to obtain bitcoins in order to use it, so is Bitcoin the currency.”

This reflects the kind of unit/system duality approach that I have found helpful, and it started me considering some further implications (I discussed the application of unit/system duality and economic/technological duality concepts to Bitcoin in “On the origins of Bitcoin” (3 November 2013)).

Discrete tradable bitcoin units are one of the integral aspects of the Bitcoin network, which in turn is a live instantiation of the Bitcoin protocol (language/convention/consensus system). The value of the units is what enables the distributed financing of the entire network; the existence of this network enables the existence, security, and value of the units.

Along another conceptual axis, economic theory supports the interpretive understanding of what people do. What things are is addressed in this case as what I call the technological layer. These layers interact, but the methods appropriate to studying them differ. One is the domain of action theory, with concepts such as ends, means, and preference; the other, in this case, of computer science, networking, and cryptography.

Still, the technological layer of Bitcoin (the system) can give hints toward economic theory interpretations of the value of bitcoin (the tradable units). Additional economic insights might at times be inspired by checking back to see what is “really” going on in the technological layer, and then clarifying the relationships between the layers.

The helpful fable of the “bitcoin”

In applying economic-theory concepts to interpreting actions, the interpreter references the more specific constructs that the people in question use in their own acts. In this case, among Bitcoin users, this construct is the operative image of “bitcoins” or other such units as interchangeable, tradable digital objects.

Yet when dialing the technology layer up into a higher presence in awareness and overlaying it on the action-interpretation layer, “bitcoins” begin to look like something of a made-up image, albeit one that enables people to interact with the technology layer in a meaningful way. The image makes it intuitive for people to use the system to accomplish their own objectives—to create, hold, and adjust balances and to buy and sell products, services, or monies out of such balances.

The tradable units on the network are not bitcoins, and are in a sense not even satoshis (100,000,000 to a bitcoin). Satoshis are an abstract unit of account within the network, whereas the elements held and traded are “unspent outputs” of all possible sizes denominated in this abstract unit (or more convenient multiples thereof). Satoshis are not now generally useful in the form of a single unspent output of one satoshi. Unspent outputs, each defined in part as some number of satoshis, are assigned to an address in a state from which they can be reassigned to other addresses (including to change addresses as needed), provided the specified signatures and other transaction data are relayed to the network.

All of this can work for a general population of end users because none of them needs to understand any of it to use the network for their own purposes. Even those who do understand such details do not have to think in such literal terms when interacting with the network in the role of end user themselves. The fable of the existence of “bitcoins” helps facilitate the human-network interaction at a practical level.

So long as the practical effect of such an image fills this role without causing errors or deceptions, it is a purely pragmatic and instrumental issue. For example, it does not matter at this level if a car’s steering wheel turns the wheels on the road mechanically or sends electronic control signals to electric motors that actually steer the vehicle—provided that the practical result in either case is that the vehicle actually turns as intended in response to the human-generated directional signals.

A dualistic valuation

In this way, combining the unit/system duality and economic/technological duality approaches can lead to additional insights about the way people value Bitcoin/bitcoin. The network is only in a loose metaphorical sense valued “as a whole.” The principle practical way for users to value it is via their own possession of and ability to transfer specific tradable units. Such units are an integral characteristic of the system. Viewed together as a social phenomenon, this could suggest the superficial appearance of a mass user valuation of the system in general. However, an idealistic “in general” valuation or mere widespread sentiments of technological appreciation could not support a functioning monetary system; only individual user valuations of discrete units can do that, and it is from there no surprise that this is precisely what Bitcoin “the system” enables.

Unspent outputs denominated in satoshis and multiples of them form a key part of the end-user interface of the protocol/network. Users value these and incorporate them into their respective structures of action. The units (or rather, the interface construction of the units) cannot function as they do in this role without the system; nor can the system exist as it does—or be entirely self-financed in a distributed way as it is—without the scarce and discretely valued tradable digital objects denominated in the system’s own abstract accounting unit.

Bitcoin weekly weighted average price, July 2010 through April 2014

Here is an update through April 2014 of a chart I published at several different points last year. It takes several measures to enable a longer-term overview next to the more typical focus on short-term trends. Notes on the reasoning behind these measures are included after the chart.

Notes: The log scale enables comparisons that reflect percentage changes across large differences in relative scale. Whereas a doubling from $1 to $2 would disappear next to a doubling from $100 to $200 on a single linear chart, this chart is able to better show both doublings at a similar relative scale of effect in its own context.

Second, weekly weighted averaging helps show price trend by general significance. This gives less weight than a daily high/low scale to headline-catching numbers that, as a practical matter, only impacted a relatively low volume during the most extreme moments of a mania peak or panic bottom, and dispite their general relative irrelevance, become the eternal darlings of all headline writers.

Third, the year-over-year comparison format shows that at least up to the present, the price has been significantly higher YoY throughout its entire history, with one famous exception, which resulted from an extreme upward and then less extreme downward price movement in 2011.

While the earliest price information begins to appear in late 2009, it is relatively sketchy and it is unclear how many trades actually took place at those published prices, so this version begins with the opening of the Mt. Gox exchange in July 2010 at $0.05. I switch to Bitstamp for the beginning of 2013 onward to factor out the impact of withdrawl irregularities that began to impact Mt. Gox data already in spring of 2013 and later worsened.