The GENIUS Act: Setting the Stage for a Federal Bailout When Stablecoins Become Insolvent

What Is a Stablecoin and What Is Its Use?

A stablecoin is a cryptocurrency token that is pegged to a fiat currency value (or sometimes to a commodity like gold). For example, Circle’s USDC token is pegged 1:1 to the US dollar, such that 1 USDC should be redeemable for $1.  The whole idea of a stablecoin is that it is a stable store of value. In other words, a stablecoin is just a token that can be redeemed for a fixed amount of cash.

Although the crypto industry likes to highlight the use of stablecoins for real world applications such as remittances and peer-2-peer payments, the real world usage is scant. As Alexis Goldstein showed in Congressional testimony several years back, the cost of sending remittances in stablecoin when all fees are included is often much, much higher than with good old Western Union; the comparisons the industry posts are never apples-to-apples of all-in costs. (And if you doubt this, ask yourself exactly how useful is it for a recipient in say, Venezuela, to receive a remittance in a form of a stablecoin? Will they be able to pay rent or buy groceries with it? There will have to be conversion into fiat, which involves fees and inconvenience.)

Instead, the primary use of stablecoins is for DeFi market making and lending protocols. This is why stablecoins account for the majority of crypto transactions, even though they are a rather limited percentage of total crypto market capitalization.

Market making. Crypto can trade either through an “exchange” like Coinbase, that’s really just a brokerage that operates a traditional order book like Charles Schwab, or through a DeFi protocol like UniSwap, Curve, or Balancer, that acts as an automated market maker (AMM). AMM’s are apps that manage liquidity pools:  liquidity providers put up a pool of stablecoins and Bitcoins/altcoins, which are locked in via an ERC-20 smart contract, and liquidity takers buy and sell the Bitcoins/altcoins from the pool in exchange for stablecoins.

AMMs use stablecoins as their liquidity medium; everything trades in and out for stablecoins. The reason: the AMM doesn’t have a bank account because it’s “owned” by a DAO, which does not have legal personhood and cannot satisfy bank KYC requirements for itself. This means that almost the whole volume of DeFi trading involves trades of stablecoins for other crypto, hence the high transaction volume in stablecoins.

(Ironically, if you wanted to start providing liquidity to an AMM liquidity pool, you’d first need to acquire some stablecoins, and you’d need to go buy them from a traditional exchange like Coinbase or Gemini in exchange for fiat, so there’s no avoiding the need to get a fiat on-ramp somewhere. In other words, there’s no DeFi without CentFi.)

Crypto lending. DeFi lending is also built around stablecoins. DeFi lending protocols use stablecoins as collateral. If you borrow from a DeFi protocol, your repayment is guarantied not by the threat of litigation in the courts or foreclosure on real world assets, but by the automatic liquidation of the stablecoin collateral you posted under an ERC-20 smart contract.  DeFi lenders are relying on stablecoins to retain their stable value, otherwise borrowers might opportunistically breach if collateral values fall below the cost of repayment.

What Distinguishes a Stablecoin from a Digital Poker Chip?

Beyond the particularized uses, what makes a stablecoin different from a poker chip in Vegas? Only that ownership is determined not by physical possession, but by control of the private key that is used to authorize transactions in the stablecoin on a blockchain, including redemptions. So it’s basically a digital poker chip.

But is a actual poker chip actually interchangeable with cash at a fixed peg? Sort of, but it’s geographically contingent. You can of course redeem a poker chip from a casino’s own cash cage, and some Vegas establishments besides casinos will accept payment in poker chips because they can readily redeem them at the casino’s cash cage. But most casinos won’t redeem more than a very limited amount of other casinos’ chips (unless they have common ownership); each casino family is a separate currency zone.

Now try paying someone with a Vegas chip in Chicago or DC. If the chip is even accepted for payment it will be accepted at a steep discount. So much for the fixed peg. The discount exists because of five related problems:

1. Counterfeiting. The payee will be concerned about whether it is even a legitimate poker chip or a counterfeit. How does a DC denizen know whether something is a legitimate Caesars’ chip?
2. Transaction costs. The payee cannot readily redeem the chip for cash—which is needed for paying certain transactions, such as tax bills and judgments—without traveling to Vegas, which is a huge transaction cost.
3. Custodial risk. The chip is easy to lose and vulnerable to theft in a way a bank transfer is not.
4. Issuer insolvency risk. There’s a chance that the casino will go bankrupt and won’t honor redemptions of the chips, so there’s credit risk of the chip issuer.
5. Uncertainty of value; network effect. The payee will be worried about whether other, downstream payees will accept the chip, and with what sort of discount, which in turn depends on the extent of the first four problems. In short, we have a network effect problem, which cannot be solved until there is enough critical volume of persons willing to accept the chip at par. It’s not enough if they will accept at a discount because that discount will have to be negotiated in every transaction, rendering the chip of uncertain value.(The reason some Vegas merchants will take poker chips as payment is because they have a customer base that might have greater willingness to spend: I might be willing to spend a whole $100 chip on a $90 item rather than shlep back to the casino cash cage to redeem it and get cash so I can pay $90.)

Now consider how that compares with a stablecoin, which fixes the counterfeiting problem and changes the nature of the transaction cost issue, but does not fix the custodial risk or issuer insolvency risk problem, and that in turn results in an uncertainty in value, at least outside of the crypto-ecosystem, such that there is little real world demand for stablecoins.

1. Counterfeiting. It is easy enough to verify if it is a legitimate stablecoin; counterfeiting isn’t a concern with a stablecoin.
2. Transaction costs. The stablecoin might be readily redeemable (or not depending on its terms) from any geographic location. But there might be transaction fees on the redemption. The mining necessary to validate the block ain’t free, particularly if you want it done fast, so transaction costs still remain that exceed those faced by a payor on a ACH, debit card, or check transaction. (And please don’t start telling me about merchant fees–I’ve been writing about those for 20 years now, so I know that payments are not free, but they are paid by the payee, not the payor, in the first instance.)
3. Custodial risk. A stablecoin does not fix the custodial risk problem. The stablecoin still needs to be stored somewhere, as the wallet where it is stored could be hacked (Bybit) or the custodian could embezzle the funds (FTX, Celsius) or the custodian could simple go bankrupt (Celsius, Voyager, BlockFi). The details of the custodial risk change by going digital, but they still exist.
4. Issuer insolvency risk. A stablecoin does not address the credit risk of the issuer. A stablecoin issuer might not be able to honor its redemption requests because its own investments go bad. USDC, for example, dropped from a $1 peg to 87¢ in the wake of Silicon Valley Bank.
5. Uncertainty of value; network effect. Here’s where things get weird….Because of factors two through four, the willingness of other downstream payees to accept the stablecoin should still be uncertain; the network effect problem should remain. And for real world users, it absolutely does. Few parties will accept payment in stablecoin for real world transactions, much less at par. The only folks who do for real world transactions are those who are willing to take stablecoins at par do so because they are willing to subsidize stablecoin payments to satisfy their ideological priors about crypto. (This could change–one could imagine Amazon or WalMart issuing their own stablecoins to save on payment costs because they have the scale to make it worthwhile for consumers to bother having their coins. )

   In the crypto ecosystem, however, the network effect problem has been overcome–there are enough folks who are willing to take payment in stablecoins at par that the uncertainty of value issue has disappeared. In part this is because if you’re doing DeFi transactions, payment in stablecoins at par is cheaper/easier/faster than the alternative of conversion in and out of fiat. But one would still expect discounting to reflect the risk.

   So what gives then? Perhaps investors are subsidizing the risk because of their ideological commitment to crypto. But perhaps investors do not understand or simply cannot measure/price credit risk on stablecoins. Stablecoins have two distinct types of credit risk—custodial and issuer—and both are incredibly hard to measure. Investors have no real way of verifying the security of custodial arrangements or the solvency of stablecoin issuers, and ByBit, Terraform, and FTX should underscore the exposure of the market to sudden catastrophic events. There’s no way to price that risk, so maybe it just gets ignored, especially on the theory that it’s rare and couldn’t possibly happen to me (some behavioral economic spicing here…). But I think another piece is that many investors simply don’t understand the credit risk and just what could happen to them if either a custodian or an issuer runs into trouble. And that brings us to the GENIUS Act, which I fear is going to lull investors into a false sense of security about the credit risk on stablecoins.

The GENIUS Act and Insolvency Risk with Stablecoins

The GENIUS Act would create a federal regulatory framework for “payment stablecoins,” that is stablecoins used for payment or settlement. The entire GENIUS Act is built around addressing credit risk in stablecoins. Most of the GENIUS Act is devoted to creating an upfront regulatory system for stablecoin issuers. I won’t go into the details, but the basic idea is that by having a standardized regulatory system, users of stablecoins can have confidence that the coins are in fact backed by the reserves claimed. (Basically this is just taking a move out of the 1863 National Bank Act, in which national banks were authorized to issue bank notes, but only in accordance with their holdings of Treasury debt, and the Office of Comptroller of the Currency was created to make sure they were complying.) So a lot of the GENIUS Act is trying to address credit risk ex ante and assure investors that bad things won’t happen.

But two provisions of the GENIUS Act deal with what happens if things do go wrong. There are two insolvency risks that arise with stablecoins. First, there is custodial risk, which is a risk that exists for all crypto. And second, there is issuer insolvency risk, which is unique to stablecoins.

Custodian Insolvency Risk

The GENIUS Act attempts to deal with custodial risk by declaring the stablecoins to be property of the investor and requires it to be segregated (but it may be commingled in an omnibus account held by a bank or trust company). What does this property status mean if the custodian ends up in bankruptcy? First, no one should assume that a bankrupt custodian has in fact been complying with its segregation requirements. If they haven’t, the there’s going to be a hot mess.

Second, even if the coins are segregated, they might just not be there. The custodian could have been hacked (ByBit) or the coins could have been embezzled (FTX…). If the coins aren’t there, the investors just have an unsecured claim for their market value as of the date of the bankruptcy filing. No one is actually guarantying that there will be coins for the investors to get back.

Third, even if the stablecoins have been segregated and are still around, it does not mean that the investors have immediate and unfettered access to their stablecoins. An investor cannot unilaterally transfer its coins out of a custodian’s possession without the custodian’s consent (and even more so if the coins are commingled). The investor doesn’t have the full set of keys to the custodian’s wallet.

The bankrupt custodian has little reason to facilitate transfers out of its custody; it’s already lost every customer who wants to take its coins and go elsewhere. Instead, the custodian’s bankruptcy estate will probably freeze transactions, at least temporarily, so it can figure out whether (1) it even has enough of the stablecoins to meet all transfer requests, (2) whether the exceptions to the property rule apply, and (3) whether it has any claims against the investors that it might want to try to exercise through a setoff. Accordingly, the bankrupt custodian will likely take the position that the automatic stay applies to all attempts to transfer the coins. And if the custodian is in a free-fall hot mess, like FTX, where there’s no telling that they will actually even have your coins or have operational functionality even if they wanted to release your coins.

So you’re likely jammed up by the automatic stay, even if it is your property (and smart contracts are not necessarily a work around–they might well be stay violations and subject to avoidance as post-petition transfers). You can move to lift the stay, but again, that’s not automatic. There will have to be a hearing and the court might not rule immediately. In the best scenario, you’re probably not gaining access to your stablecoins for a good month and possibly much longer. Meanwhile, you are exposed to market swings. If the coins drop in value in the interim, well, that’s on you bro. So just making the coins your property doesn’t actually eliminate the custodial risk problem. It only lessens it. Yes, it is better to have the coins as your property than to be an unsecured creditor of the custodian, but it doesn’t mean that you are unimpaired.

Put another way, the GENIUS Act doesn’t entirely resolve the problem of coin ownership that bedeviled the 2022 round of cryptocurrency exchange bankruptcies, but even if it did, that doesn’t actually get the coins back in the hands of the investors immediately.

Contrast this with the fate of deposits at a failed bank. The FDIC probably does a whole bank resolution–the bank is sold as a going concern to a buyer that assumes all of the deposit obligations. The depositor has nearly uninterrupted access to its funds, whereas the stablecoin investor has to wait and possibly fight to get access to its coins and faces market value risk in the meanwhile.

Issuer Insolvency Risk

Custodial risk is a problem that exists for all crypto, but stablecoins have an additional type of credit risk, that of the issuer. The attraction of a stablecoin is that it is (in theory) redeemable at a fixed peg. That requires the issuer to have sufficient liquid assets to be able to meet all redemption requests. In theory, a stablecoin issuer should just be putting its reserves into very, very safe assets, like insured bank deposits, Treasuries, commercial paper, etc. But we know from recent history that sometimes that is not the case. USDC was trading at 87¢ on the dollar when investors realized that its issuer, Circle, had billions in reserve in uninsured deposits at the failed Silicon Valley Bank. Paxos’s BUSD has also found itself massively short on reserve assets in the past.  Stablecoin issuers make money off their reserve earnings, so they are always incentivized to try to chase higher yield if they can get away with it.

Does the Investor Even Have a Claim?

So what happens if a stablecoin issuer ends up insolvent and files for Chapter 11 bankruptcy. First, it’s not clear that all stablecoin holders would even have a claim in the bankruptcy. Some stablecoins given redemption rights only to a limited subset of institutions, such that most holders do not have redemption rights. Without redemption rights, a stablecoin holder probably doesn’t have a claim. Instead, it would have to sell its claim to one of those institutions with redemption rights, which could then have a claim. Those institutions are going to extract a serious discount, if they’ll buy at all. (And if the stablecoin is locked up in a smart contract, there are further questions about who would have the bankruptcy claim…)

Super-Duper-Duper Priority Still Doesn’t Make You Top Dog

If the stablecoin holder has a claim, the GENIUS Act provides that it has super-duper-duper priority per a new section 507(e) of the Bankruptcy Code. The Bankruptcy Code’s priority system is somewhat opaque and needs to be pieced together from disparate Code provisions. Here’s the short version. Sitting at the top of the tree are claims not subject to the automatic stay, such as those of repo and swap counterparties. They are entitled to grab whatever margin has been posted to the transactions. After then come secured claims (section 725), but only from their collateral. Then section 726 takes over. It proves that section 507 claims have priority over general unsecured claims. Section 507 provides that section 507(b) claims having superpriority over section 507(a) claims, such as the administrative expenses of the bankruptcy or certain tax claims. But even 507(b) claims get trumped by super-duper priority claims of DIP financiers under section 364(c)(1). So what does new section 507(e) do? It would say that stablecoin claimants have priority over the administrative expenses of the bankruptcy and employee claims and tax claims, but not over DIP financing claims, secured claims, or swaps and repos.

Bankrupt Stablecoin Issuers Are Likely Insolvent

As an initial matter, that’s just unworkable. The administrative claims–the lawyers and other bankruptcy professionals—-need to come first or they won’t do the work: you gotta pay the gravedigger. But once you do, notice the problem: there might not be anything left by the time the stablecoin claimants come up for payment. Remember that a stablecoin issuer isn’t likely to file for bankruptcy unless its peg has broken the buck. That’s doubly bad news for the investors. First, if the stablecoin has broken its peg, then it probably does not have enough assets to pay all of its creditors. So the order of priority really matters. If the issuer’s reserves include lots of swap and repos positions, the issuer’s assets could be cleaned out by counterparties. At that point the DIP financier and the professionals will gobble up what’s left. Bankruptcies aren’t cheap: FTX has had nearly $1 billion in professionals fees.  So yes, priority is nice, but stablecoins investors aren’t getting enough priority to really protect them and giving them more (or even what they currently have in the GENIUS Act) starts to make bankruptcy unworkable. (And just so it’s clear, if there’s no bankruptcy process, every stablecoin investor is in a race to the courthouse with all the other investors to try to get a judgment and levy on whatever assets remain of the issuer. Good luck with that.)

Stablecoin Investors’ Claims Might Be Dollarized at Market Values at the Time of the Bankruptcy

Second, their claims are not for a stablecoin, but instead get dollarized as of the time of the bankruptcy petition. There’s an argument that they get dollarized at the market price of the stablecoin, rather than at the redemption value. If so, then they have already realized the market value loss at the time of the bankruptcy filing and will not get it back.

Putting Some Numbers on It

Let’s suppose, however, that there are enough reserves around to pay all of the prioritized stablecoin investor claims. Even if the investors get their claims paid in full, they don’t get them paid until the effective date of a bankruptcy plan, which might be years in the future, and they don’t get any interest on their claims. Let’s put some numbers on it. Imagine a stablecoin issuer that fails after breaking the buck and that its plan does not go effective for 2 years. What happens to an investor who has a $1 million stablecoin claim? He gets paid $1 million in 2 years. If we used a 7.5% discount rate, continuously compounded, then the present value of that claim is less than $800,000. The delay alone will eat up a fifth of the value of the claim.

Now let’s make it even worse. Let’s suppose that the stablecoin is trading at $0.87 and the court says that the investor’s claim is for the value of the coin, not the par redemption amount. Under section 502, the investor’s claim is locked in at 87 cents on the dollar. So the $1 million investment is now a $870,000 claim. If it is paid in 2 years, then the present value, assuming a 7.5% continuously compounded discount rate, is around $750,000, and if it is paid in 3 years, the present value would be down to under $700,000.

Now let’s make it even worse and say that swap and repo claims and professional expenses have eaten away half of the reserves before the plan goes effective. At that point, the stablecoin investor is getting a nominal 43.5¢ on the dollar in 3 years, so the present value under the prior assumptions would be around $350,000.

You might dicker with my discount rate assumption or with the question of whether the claim will be allowed only at the market value rather than face or how long the bankruptcy will take or even if the reserves will be insufficient. Maybe the loss won’t be as bad as in my scenario. But there is no avoiding that (1) there will be a present value loss and (2) whatever the payment is, it will be delayed. It’s either a bit bad, or really, really bad, but there’s no scenario in which the investor doesn’t take a loss.

Bankruptcy Ends up Very Badly for Stablecoin Investors, Any Which Way

The GENIUS Act tries to mitigate credit risk on stablecoins by declaring their property of the investor vis-a-vis custodians, and prioritizing the investors’ claims vis-a-vis creditors of the issuer. But neither is really a fix, and the truth is that absent a government guaranty, there’s no way around this problem. No matter how much stablecoins are prioritized in bankruptcy, bankruptcy is a slow process, and time is money. And there are limits to how far stablecoins can be prioritized in bankruptcy without rendering the bankruptcy system unworkable. (Making stablecoins the property of the investors is no help in an issuer bankruptcy because all the investor gets is a digital token that is worthless without the redemption right, and that’s just a bankruptcy claim; this is different than in the custodian bankruptcy scenario.) The only real way to ensure 100% timely repayment is a government backstop, but that’s not something the industry wants (because it goes with regulation).

At the end of the day, even with a 1:1 peg, a stablecoin is not the equivalent of a dollar in a bank deposit account. If the bank fails, the FDIC steps in and ensures uninterrupted access to the deposit. If the stablecoin custodian or issuer gets in trouble, investors are going to be impaired; the only question is by how much.

So this should raise the question of how a stablecoin pegged to the dollar at 1:1 can clear at par in the market. Perhaps there is some sort of discounting that cannot be observed, but short of that, the only answers I can provide is that investors either do not understand the credit risk on stablecoins, hyperbolically discount it to zero because they cannot readily measure it, or are just willing to subsidize stablecoins because of ideological priors.

Any which way, there’s some sort of unpriced risk here, and the GENIUS Act is likely to lull investors into thinking that stablecoins are equivalent to deposit accounts in terms of credit risk, and they ain’t. That’s not just bad for stablecoin investors. It’s also bad for taxpayers who want nothing to do with crypto.

The GENIUS Act Creates an Implicit Government Guaranty of Stablecoins, Meaning a Subsidy for the DeFi Market

By creating a regulatory regime for stablecoins, the federal government will “own” any problem that arises in the market. And here’s the pernicious operation of its ineffective insolvency provisions:  they promise to have created safety for stablecoin investors at no cost, but because it cannot deliver on that promise, it sets up a situation where the government has to deliver safety otherwise, on its own dime. In other words, it sets up a bailout. When there is another crypto crash and stablecoin owners realize that their going to incur major losses, they will come crying for a bailout, noting how critical stablecoins are for the whole DeFi world and how they thought their investments were safe because of the GENIUS Act.

What do you think will happen then? After Silicon Valley Bank can one really have confidence that they won’t get a bailout? Will banks be allowed to support their insolvent stablecoin issuer subsidiaries? Will the US Strategic Cryptocurrency Reserve (if created) be used to bail them out by buying their stablecoins at 100¢ on the dollar?

The GENIUS Act creates an implicit government guaranty of stablecoins. That means that taxpayers will be implicitly subsidizing the DeFi transactions that rely on stablecoins and that generally sit outside of the reach of anti-money laundering enforcement: taxpayers are going to be implicitly subsidizing money laundering. Is that really a desirable policy outcome? I fear the consequences of the GENIUS Act haven’t been fully thought through.