Over the course of his career, Gary Gorton has gained a reputation for being something of an experts’ expert on financial systems. Despite being an academic, this is in large part due to what might be described as his practitioner’s take on many key issues.
The Yale School of Management professor is, for example, best known for a highly respected (albeit still relatively obscure) theory about the role played in bank runs by information-sensitive assets.
Given Gorton’s reputation for deep and thoughtful analysis (with a twist), it shouldn’t be surprising then that the professor is now turning his expertise to studying the run-risk of stablecoins.
In a recently published piece co-authored with Jeffery Zhang, an attorney at the Board of Governors of the Federal Reserve System, the two authors make the implicit about stablecoins explicit: however you slice them, dice them or frame them in new technology, in the grand scheme of financial innovation stablecoins are actually nothing new. What they really amount to, they say, is another form of information sensitive private money, with stablecoin issuers operating more like unregulated banks.
In that vein, the eternal problem that has always applied to private money — that it remains a subpar medium of exchange that is subject to runs — continues to apply to stablecoins. To regulate them properly then, you need to factor in the broader history of private money and its information sensitivity.
The Gorton worldview emerges from a longstanding insight that the institutions which are most likely to create private money — banks — are structurally incentivised to operate in the market as secret keepers. This tendency towards secret-keeping ensures the assets/liabilities they create are exposed to suspicion and should, as a result, always be more costly to transact than “secretless” or “information-insensitive” money, which unlike the former is above suspicion.
The only way to close the information sensitivity gap is with due diligence or system-wide regulation and supervision, but this costs both time and money.
In Gorton’s framework, it is only the government that can create the riskless collateral upon which “secretless” or “information-insensitive” money is based.
In the context of stablecoins, Gorton and Zhang dub the above the no-questions-asked (NQA) principle of money, ie, a situation where money is “accepted in a transaction without due diligence on its value”. As they note in the paper:
. . . . NQA means both parties to a transaction must agree that the money be accepted at par — a ten-dollar bill should be accepted as being worth ten dollars, not a penny less. Achieving the characteristic of NQA has, historically, been very hard.
Achieving NQA status is so hard, in fact, that even conventional bank demand deposits were unable to accomplish the status without deposit insurance — something the financial system only appreciated after the panic of 1929.
As the authors note:
Few remember that demand deposits were unable to achieve NQA without deposit insurance. And, even with deposit insurance, the identity of the check-writer matters, so there are still questions asked, which is why demand deposits do not circulate as money. They remain account-based but have a partial NQA property. It is this NQA property that allows money to have a convenience yield — that is, a return which is all, or in part, nonpecuniary. For instance, individuals carry cash around even though it does not pay interest, because it has a convenience yield.
Stablecoin issuers seem to implicitly understand their NQA vulnerabilities and the related run-risk, say the authors. It’s for this reason the biggest names in the sector have opted for full transparency, publishing regular updates of the composition of their reserves.
In reality, however, this is not enough to prevent stablecoins from becoming new sources of systemic risk. The following table from the authors shows the processes that would have a much better chance of doing so:
The three options that have the best chance of endowing stablecoins with NQA, however, have their own issues. Applying deposit insurance to stablecoins initiates a broader discussion about whether stablecoins should be forced to be licensed and regulated like banks. Replacing them with central bank-issued stablecoins, meanwhile, initiates a conversation about whether crowding out the private sector by way of an overly dominant central bank is really a good idea.
What interests us most, however, is the proposition that their run-risk could be contained by being backed one-for-one by Treasuries or central bank reserves. Doing so would be largely equivalent to when the People’s Bank of China forced WeChat and Alipay to join NetsUnion. Yet, as Gorton and Zhang also note, in such a framework stablecoin issuers would not make any money, even if their stablecoins developed a convenience yield, because cash does not pay interest.
This, we think, is astute. It highlights the scale of the financial burden that such measures impose on payment companies. Hitherto to such requirements these entities would have operated as de facto money market funds / stablecoins, benefiting from the seigniorage they could earn from essentially minting their own money. In the Chinese case, when Alipay and WeChat were forced to join NetsUnion, this seigniorage revenue would have been lost, potentially threatening the viability of the companies’ business models. This could explain Alipay boss Jack Ma’s October 2020 speech, in which criticised he China’s outdated approach to regulation and urged officials to move away from the pawnshop mentality of today’s finance in China towards a more credit-based system.
The Par problem
Gorton and Zhang note that when it comes to private money, issuers have an incentive to synthesise information insensitivity by adding so much opacity that the cost of producing information about the backing assets exceeds its worth. In such an environment, the private money defaults to trading at par if it is coupled with a trusted regulatory regime. However, this can only be assured for as long as market confidence in the underlying assets remains constant. If a confidence crisis hits the assets, as it did in 2008, the demolition of par can be quite sudden and ferocious.
Since stablecoin can’t rely on bank examiners, the authors argue they face a trade-off between opacity and transparency:
On one hand, it would be best if the backing for their stablecoins were so opaque that nobody would find it profitable to produce information about the backing assets. On the other hand, if the backing is not credible, then the market will want to produce information about the backing. Stablecoin issuers have currently taken the view that transparency is best, because they are not regulated and cannot rely on bank examiners and so they cannot be opaque.
It’s tempting at this point to have a debate about whether stablecoins really do represent a type of demand deposit from a risk perspective and whether that should entitle them to both bank-like regulatory burdens and privileges.
But really, as Gorton and Zhang explain, what we learned from money market funds during the financial crisis of 2008 is that it doesn’t really matter if the liabilities in question truly qualify as demand deposits according to the letter of the law. Financial instruments that are designed to perfectly mimic demand deposits have the same upsides and downsides however you structure them. “When bank depositors believe that the bank is no longer able to provide a full redemption of their deposits, they run on the bank with the hope of withdrawing their deposits before it’s too late,” they note. In money-market land this is known as breaking a buck.
This implies: if it looks like a demand deposit and behaves like a demand deposit it should be regulated as a demand deposit. Regulators belatedly came to understand this about money market funds, opening the door extensive post-crisis regulation. Nonetheless, in March 2020, it soon became clear that despite all regulatory attempts to de-risk money market funds, only direct Federal Reserve intervention could protect the funds from run-risk when Covid-19 fears gripped the market.
If that sort of thing can happen to money market funds, it seems logical to assume that it’s a question of if not when run-risk hits stablecoins and thus that regulation should be pre-emptive not retrospective.
But those who are sceptical of regulation might argue that if all those measures failed to prevent money market funds experience run-risk in 2020, then it’s unlikely they will help stablecoins in the long run either. Some believe the market would be better off leaning into the risk by openly and transparently emulating the 19th century US free-banking model. Gorton and Zhang agree the free-banking era offers the closest analogy to today’s stablecoin environment. Except, unlike with stablecoins, a key hallmark of the preceding era was that the banknotes issued by free banks often did not trade at par at all.
Here, as an example, is the discount associated with the Planters Bank of Tennessee note over time:
This discount variability was actually indicative of NQA, and how it differed both by time and location. But, according to the authors, the variability never caused an economic problem per se. From an efficient market theory perspective it could easily be accounted for by specialist dealers. The problem it presented was that it made the process of transacting hugely cumbersome since “there was constant haggling and arguing over the value of notes in transactions”.
The only way to overcome this, in the eyes of Gorton and Zhang, is probably by bringing stablecoins under the remit of central banks or by deploying CBDCs to compete them out of existence. But this too is not a perfect solution. The authors recognise that the crowding out of the private sector might force central banks to invest the funds deposited with them in ever riskier securities.
And this in turn introduces what FT Alphaville has long referred to as Gosbank risk, a point when central bank dominance is so great that it begins to influence capital allocation directly. If that happens on politically subjective grounds, it is a process that risks undermining hallowed central bank independence. Indeed, as Gorton and Zhang note:
The problem is that this would introduce distortions into the capital markets, as the private sector would over-produce the highest risk securities that the Federal Reserve purchases. This occurred in the Euro-zone. As Nyborg put it: “if central bank money is available only against igloos, or igloo-backed securities, igloos will be built. In short, the Federal Reserve would be engaging in fiscal policy with all the political ramifications that would entail and jeopardizing its independence.
The irony of such a world in FT Alphaville’s opinion is that it might — if the central bank makes particularly bad or unpopular capital allocations — risk the central bank itself losing NQA status. If stablecoins succeed in differentiating themselves from fiat cryptocurrencies and still become used as money, however, that might mitigate that risk.
Stablecoins as a collateral sinkhole – FT Alphaville
The Chinese MMF that came in from the cold – FT Alphaville
Is it a bank, a money transmitter, or a Silicon Valley shadow financier? No, it’s just Paypal! – FT Alphaville
We all become MF Global eventually, Tether edition – FT Alphaville
Cbank digital currencies and the path to Gosbankification – FT Alphaville