A Quality-Ladder Theory of Digital Currencies

Ever since the first cryptocurrency, Bitcoin, was invented in 2008, the number of cryp- tocurrencies has exploded. As of July 2018, there were already about 1,800 cryptocurren- cies. Many of them were invented with various improvements over previous ones, such as a less wasteful consensus mechanism, a stronger cryptographic technology, and a more flexible transaction protocol. These innovations, just like innovations in more traditional economic sectors, are costly for the corresponding innovators. In a decentralized market of privately-issued currencies, do innovators have adequate incentives to innovate for the benefit of the society? This project makes a first attempt in answering this question.

Specifically, I plan to address this question by developing a quality-ladder model of digital currencies. In this model, a succession of digital currencies are adopted by the society, with each new one replacing an old one, and every one co-exists with the government-issued fiat money. Digital-currency innovators recoup their innovation costs by selling the initial stocks of new digital currencies instead of by collecting seigniorage. A special feature of my model is that different kinds of innovation embodied in a new digital currency can be summarized by what I call its effective nominal interest rate, with a lower rate representing a better currency. Successive digital currencies are invented with lower and lower effective nominal interest rates. Since all monies are bubbles, a newer, better digital currency needs not replace an older, worse one. The probability that such an event happens depends on how the continuation equilibrium is selected, which I assume is driven by sunspots. The fact that innovators are willing to invest resources to innovate reflects the expectation that, with a positive probability, consumers will coordinate on a continuation equilibrium where a newer, better digital currency is adopted at the expense of an older, worse one.

I expect that rich welfare implications can be derived from this model. For example, the mere expectation that some new digital currency will one day appear, replacing the existing one, may already depress the real value of the existing digital currency. The same expectation may also depress the real value of the government-issued fiat money, even though the latter will not be replaced by the new digital currency. Both would lower welfare immediately by making decentralized transactions less efficient. Ex post, when a better digital currency is invented and successfully replaces an old one, there may be two opposite effects on welfare. For decentralized transactions where digital currencies are acceptable, trades may become more efficient with a better digital currency. But for decentralized transactions that rely on government-issued fiat money, trade may become even less efficient because the real value the government-issued fiat money is further depressed.

This model hence has the potential of guiding regulatory policies towards digital currencies. I plan to fully develop this model, verify my conjectures, and obtain other welfare and policy implications. I expect that at least one journal article will be generated.