Pegcoin vs. stablecoin, a primer
An unrolled tweetstorm, originally on @ecoinomia.
Seeing yet another big player getting pegcoin confused with stablecoin, it might be time for a “why pegcoins aren’t stablecoins” thread, in three parts:
1. What “stablecoin” actually means
2. How pegcoin is at best a simulated stablecoin
3. How to stabilize coins.
1. The "stable" in stable coin refers to price stability, i.e. the extent to which price changes signal changes in the competitive environment of the priced commodity rather than changes in the underlying currency. The "marvel of the price system" relies on price stability.
Econometrically speaking, price stability is a "true value", the "observed value" is the purchasing power of a currency, usually measured by calculating the cost of a defined bundle of goods — an enormously tricky undertaking fraught with measurement problems.
Creating a cryptoasset that serves as a reliable price signal runs into the same problems, typically amplified by thin market and segmentation problems. Otoh a token that can't get stability under control doesn't really serve as anything other than a security. That's fine tho.
2. Pegcoins are a common kludge to stabilize a thin market token, which makes sense in a B2B environment where you would like the automated clearance but not the friction of short-term volatility. Makes sense but...
The big drawbacks of pegcoins are 1. that you can at best achieve the stability of the underlying asset (that's why tier 1 fiat currencies are popular pegs), and 2. you're making yourself dependent on a value creation process and issuance scheme which you cannot control.
In order to hold the peg you need to collateralize your coin ether in full (sound but suboptimal), fractionally (that's what banks do) or not at all (that's fraudulent).
There's a large twitterature out there on "why pegcoins don't work" (usually mislabeled as stablecoin), so I can skip that and talk about actual stablecoins. Short version: thin collateral makes pegs explode. (Also, pegging against fiat makes you an exchange.)
3. To create a stablecoin that's endogenous to the underlying value creation process you have in principle two levers: 1. the issuance scheme, and 2. the allocation scheme.
In practice you're running against a bunch of taboos and more technical obstacles. Taboo #1 is that you can't (and shouldn't) commit to a fixed cap bc you have to adjust issuance ad hoc acc to an external signal, which leads directly to taboo #2, the dilution/debasement scare.
But while by default new coins are allocated to in-kind efforts that secure the network (read: mining), in theory new coins could also be allocated to current coin holders as a kind of automated dividend. Some algorithmic stablecoins propose to do that.
That still leaves pretty steep technical problems: How to create (or work around) an unimpeachable external signal of purchasing power, how to withdraw coins from circulation, how to parametrize issuance and allocation, etc.
But there's already a bunch of algorithmic stablecoins out there that attempt to implement such a scheme. How well they achieve this is likely more an empirical than a theoretical problem, so there's certainly space for more.
Blockchain space like startup space is about parallelizing the taking of wrong turns.