Sometimes, "stablecoins" and variants such equally "algorithmic stablecoins" function like historical names, as they refer to projects that telephone call themselves stablecoins, such as Ground Greenbacks, Elastic Gear up Dollar, Frax and their clones.

The word "stablecoin" tin can exist used every bit a logical description for "a cryptocurrency designed to take low cost volatility" and has "stores of value or units of business relationship," or "a new type of cryptocurrency that often have their value pegged to some other asset… designed to tackle the inherent volatility seen in cryptocurrency prices," or a currency that can "act every bit a medium of monetary exchange and a mode of storage of budgetary value, and its value should remain relatively stable over longer time horizons."

On the more than metaphysically speculative stop, some have defined a stablecoin equally "an asset that prices itself, rather than an asset that is priced past supply and need. This goes against everything we know most how markets work."

Circularity is the core issue, as I come across information technology. The alleged deficiency of Bitcoin (BTC) as money and a vague definition originally inspired a host of stablecoin projects. The design features of these projects have now been incorporated back into the stablecoin definition.

Haseeb Qureshi — a software engineer, writer and famous altruist — defines a stablecoin equally simply a cost peg. Nonetheless, it is not obvious that anything with a peg should bear the name of stablecoin. Ampleforth has a "peg" and has been bucketed into the stablecoin category. The founding team routinely clarifies that it is no such affair.

And so, who is right?

Another example of merely what exactly is "stable" in a stablecoin — the peg or its value? Wrapped Bitcoin (wBTC) is perfectly pegged to Bitcoin — one wBTC volition always exist 1 BTC. Is that a stablecoin?

Co-ordinate to the original motivations for creating stablecoins, BTC is not a stable means of exchange, even though Bitcoin is the canonical "store of value" asset.

Having antiseptic the trouble — that no one knows how to define or recognize a stablecoin — the rest of this essay outlines a solution. It provides a well-divers description of value as a relational holding, namely, "value in terms of a measurement unit."

Using this description, I and so comprehensively allocate all digital avails along two dimensions — risk of loss, or the probability of realizing a decrease in value, and risk of gain, or the probability of realizing an increase in value. We can then precisely and logically define stablecoins: assets where the risk of loss and hazard of gain are both nothing.

That is:

p(gain)=p(loss)=0

I call this a adventure-divers stablecoin.

It is articulate that today's algorithmic stablecoins accept a risk of loss but no risk of gain. Thus, not only are they not stablecoins, merely they are terrible financial assets. I finish by considering whether it makes sense to expand the concept of a chance-defined stablecoin to a more full general concept centered on expected value; an expected-value stablecoin is one where the probabilities of loss and gain, weighted past the magnitude of loss and gain, are perfectly start and net out to zilch.

I conclude that the complexity and ergodicity of such a concept rule it out equally a useful stablecoin definition.

What is value?

What "value" means is not entirely clear, as evidenced past continuing debates about the "truthful" rate of inflation. Nosotros can enquire: Value in terms of what?

That is, we make up one's mind to treat value every bit a relational property between the object being measured and the thing doing the measuring. Information technology is similar asking for height — do you want it in inches or centimeters? For our purposes, can we ascertain a function that maps an asset to a set of numerical values in a chosen unit? I call it: Value.

For example, if the chosen unit is the U.South. dollar, and the particular is a bag of chips,

ValueUSD(chips)=$5.

We could only likewise have written Heightinches(table)=35in.

Risk of loss, gamble of proceeds

The value of an asset changes over time, so we can expand our Value function to reflect the idea of "the value of an asset, in terms of a unit of measurement, at a certain fourth dimension" by adding the time (\t) at which we are measuring value:

ValuetUnit(nugget)=x

We tin define risks equally the probability that, at a randomly chosen time in the time to come, the Value role would show a decrease or increase in value.

In practical terms, this means that if I convert the asset into my chosen unit, I would realize a loss or a proceeds.

A risk-defined stablecoin

We now have enough to create a well-defined clarification for a stablecoin. A stablecoin is an asset where the risk of loss and the run a risk of proceeds are both zip. That is: p(proceeds)=p(loss)=0.

This ways that if I sell the stablecoin asset in the futurity, I volition neither experience a loss nor proceeds in value, equally measured in my chosen unit.

The Boston Consulting Group's famous matrix was invented past the company's founder, Bruce Henderson, in the 1970s. With some rearrangement, we tin can repurpose the Boston Consulting Group growth-share matrix to classify all digital assets by their risk of loss and risk of gain. The four categories are however stars, dogs, unknowns and cash cows.

A star investment, with no adventure of loss but a risk of proceeds, is rare nowadays but abundant in hindsight, such every bit when one regrets selling Bitcoin back in 2010. Stars also exist in the imagination. Such was the case with the investors in Bernie Madoff's fund. Just those kinds of investments quickly reveal themselves to be dogs. Dogs are sure losers — in that location is no run a risk of gain, but if yous hold them long plenty, the take chances of loss becomes an actual loss.

Star investments are most abundant in hindsight when we can no longer buy them:

Unknowns are your regular investments — y'all could be up or down in terms of value, depending on the day. Most digital assets, fifty-fifty Bitcoin, autumn into this category. Lastly, cash cows are investments that accept minimal risk of loss or gain. They are dependable. Nosotros tin can now take those projects that accept been named as stablecoins to see which truly fit.

Let'southward put some major digital assets and stablecoins into the gain-loss matrix.

Projects chosen algorithmic stablecoins are stablecoins in name only. Because of their multiple token designs, they have no risk of gain — every bit all of the new supply is given to investors — but holders retain a risk of loss.

Toll peg is not plenty. The expected value of owning an asset could be positive or negative, but information technology is non zero. Another lesson is that it is important to specify a unit when discussing value. If our measurement unit is the U.S. dollar, then wBTC is non a stablecoin. Only if we are defining value in terms of BTC, then wBTC is the perfect stablecoin.

Lastly, risk cess is hard. I've received pushback about classifying Tether (USDT) equally a stablecoin, given its counterparty risk.

These are all valid points.

Except nether boggling circumstances, no stablecoin is truly free of the risk of loss. Perhaps Tether is a cross between a dog and a cow.

Nevertheless, it should exist clear that certain projects egregiously advisable the term "stablecoin" in a bid to grant investors a hazard of proceeds while saddling holders with a risk of loss. Since no sane person would concur these assets on their books, withal, it is near certain that these dogs will go extinct.

An expected-value stablecoin?

Astute readers volition have noticed that expected value is not just a part of the probability of loss and gain — the magnitude of losses and magnitude of gains is merely as important.

For instance, assume I have a fair dice. If I roll a six, I win $60. If I roll whatsoever other number I lose $6. The expected value of rolling the die is:

EV(die)=$lx∗p(gain)−$6∗p(loss)=$60∗(i/half dozen)−$6∗(5/half-dozen)=$5

Only tin nosotros expand the concept of a run a risk-defined stablecoin into that of an expected-value stablecoin? In other words, would it suffice for the expected value of holding an asset to be zero? Using the dice example higher up, this condition would be met if I won only $30 instead of $lx. So, whatever time I try to convert this "DieCoin" into U.S. dollars, there is a five-sixth chance I will realize a loss in value, and a 1-6th chance I will realize a proceeds. Simply considering the gain is then much larger than the loss, these cancel out.

I call up this could be a clever approach that tin be realized through a gear up of derivative contracts. Nonetheless, it would lose the property of assuasive holders to exit their position with minimal impact to their portfolios.

This should remind us that, ultimately, definitions are artifacts of a community of speakers. And I find it doubtful that more than than a few people will find an expected value definition persuasive.

This article does non contain investment communication or recommendations. Every investment and trading move involves adventure, and readers should bear their own research when making a decision.

The views, thoughts and opinions expressed here are the author'south lonely and exercise non necessarily reflect or represent the views and opinions of Cointelegraph.

Manny Rincon-Cruz is a fiscal historian and a researcher at Stanford. He serves as an advisor to the Ampleforth project, and is a co-author of the protocol'due south whitepaper. His research focuses on various aspects of budgetary history, Chinese history, and network science. Whatsoever opinions expressed hither are his alone and exercise non stand for the position of Stanford University.