Thread by @GeorgeSelgin

I ended my first thread on Gorton and Zhang's new working paper (papers.ssrn.com/sol3/papers.cf…) by observing (with a nod to work by the late, lamented George Kaufman) that bank runs aren't necessarily a bad thing: papers.ssrn.com/sol3/papers.cf…

That makes for a neat segue to my next criticism, concerning GZ's table 1 (p. 5). Here they consider various "Options to Address Stablecoins," asking of each whether it (1) would eliminate runs on stablecoins and (2) would make it unnecessary for their users to scrutinize them.

Based on those assumed goals, they narrow down acceptable options to three: treatment like ordinary banks, 100% reserve (or Treasurys) backing, or replacement w/ CBDC, that is, outright prohibition.

But should we accept this procedure? If stablecoins are indeed "like" banks, as GZ argue (albeit not consistently), Kaufman's strictures should apply to them. And if they aren't like banks, it's even less clear that they need to be absolutely run-proof.

As Kaufman notes, and as I noted in my previous thread, runs can be an effective means for shutting-down bad banks quickly. They are to be regretted only when they pose systemic risks, because panic spreads like a contagion, or because institutions are "interconnected."

Work by Kaufman himself and by Calomiris and others shows that bank run contagion effects have been much less common and extensive than is often supposed. "Interconnectedness" has been a more common cause of spillovers, but it has rarely been such as posed a systemic thread.

Are stablecoin issuers different? Would a run on one necessarily pose systemic risks? GZ themselves supply an answer 2 pages later, in a footnote where they refer to the sudden, spectacular run on the Iron Titanium Token in June that caused it to become worthless in a single day.

Of course holders of those tokens took a beating, including some who, having first touted it, made their losses a reason to call for more regulation: news.bitcoin.com/mark-cuban-iro…

But economists (usually) look for spillovers before calling for regulation. Where other markets disrupted? Was the run contagious? Did were systemically important Iron Titanium counterparties threatened with failure? Nah.

Instead, some crypto enthusiasts who thought they'd found a way to make a fast buck while others "had fun being poor" got their cumuppance. These weren't people who had no choice but to deal with stablecoins because they had to go shopping: they were investors placing risky bets.

Why should public policy strive to rule-out any chance that such people will incur losses, any more than it strives to ban dealings in penny stocks, gambling, or (and almost certain 100% loss) lottery ticket sales. (Oh, sorry: it has dealt with the last, via GZ's solution # 3!)

If you ask me, the best solution to the "problem" of stablecoins like the Iron Titanium Token is instead GZ's solution 1: do nothing. And by all means don't do anything that suggests that the government is "watching over" them!

What about GZ's 2nd regulatory desideratum, viz., that regulation must help stablecoins achieve "no questions asked" status? Here, the problem is circular reasoning.

GZ maintain that the "no questions asked" (NQA) condition must be met by any decent "money." They then characterize stablecoins as "private money." And so they conclude that regulators must see to it that stablecoins satisfy NQA.

But who, besides GZ, says that stablecoins are or have to be "money"? The standard definition of money is any "generally accepted medium of exchange." No stablecoin today meets that definition. Instead, most are used only for very limited types of transacting.

Tether, for instance, is pretty much used only for cryptocurrency purchases and sales, and then only because most crypto exchanges aren't plugged into the formal, Fed based USD payments network.

Those of us who care only to have some decent "money" to transact with don't bother with stablecoins. Why should we when we already have plenty of NQA alternatives at our disposal? So, who cares if there are non-NQA conforming stablecoins out there?

Perhaps GZ imagine that there is some risk--purely hypothetical at this point--that one or more non-NQA stablecoins will succeed in displacing good-old NQA dollars in ordinary payments. Perhaps they worry that Facebook's Diem will do so.

But for them to suppose so begs the question: if NQA is "the most obvious" property any decent money must have, why would the public abandon established NQA-conforming media for non-NQA alternatives?

If GZ have a reason for thinking it would, they should explain it before they conclude that we need potentially Draconian stablecoin regulations to prevent it. In the meantime, so long as stablecoins aren't _really_ money, they simply don't "need" to be NQA.

That's all for now. Thanks for your attention.

Sheesh! So many typos. Sorry!

Sorry for wrong link. Correct one to Kaufman paper is here: fraser.stlouisfed.org/files/docs/his…

Note: This is not an argument for zero regulation. It is an argument against the assumption that the goal of regulation should be zero stablecoin runs.