It’s not often my old profession (archaeology) and my new profession (stablecoins + governance) collide. But just down the road from me someone found one of these Ingles tokens in their yard. Akin to a 50 cent reverse-stablecoin. It’s debt owed by the holder!

This is, to me, less a story about crypto than about Big Money and Small Money. I don’t know what it’s like in Brazil, but a loan ticket this size would be mostly impossible to make in the US at any reasonable interest rate, because the underwriting and compliance likely eats up $1000 or more. I’m known for my position that USD stablecoins have the most PMF outside the US, and that same-currency stablecoins within any country are a tough sell. But if done via credit creation, I can see stablecoins getting a wedge even within their home-currency markets. Of course, all major regulatory regimes have said you can’t do that - full reserve only! - to say nothing to lending laws, which are actually quite loose in the US but not most countries.

This is admirable in concept, but there’s a reason why there’s been slow uptake. Absolutely zero shade on BW for trying this, but it is unlikely to deliver on its goal of providing reliable, standardized information. My constructive feedback on problems and their solutions below: 1) There is no clear upside to providing these public declarations. 2) If you believe there IS an upside - making your token more investable - then congratulations, you are providing statements with the express purpose of other relying on them for making investing decisions. Here be dragons. 3) Much of the liability surface in this industry revolves around not what you *do* so much as *doing what you say*. Litigious government agencies (like Biden SEC) and law firms (like Berwick) will take anything you say and use it against you. 4) …which means it takes a lot of time from your most valuable legal, technical, and financial staff to fill out these filings honestly and accurately to avoid making a misrepresentation. That’s not a small expense. You’d also need to update with new filings if you believed this was valuable. The more dynamic your organization and business, the bigger the burden becomes and the higher the legal risk of stale information being used by an investor. 5) I can’t tell that any of this info is actually verified, so I don’t think investors will at the margins make any investment decisions based on it. You’d still need to call the teams and counterparties to verify information, so it doesn’t save you a lot of effort unless you don’t diligence at all - in which case you don’t care about this anyway. 6) It doesn’t include ongoing financial metrics. Knowing how much token supply was sold and is vesting isn’t enough to make an investment, only a speculation, limiting the value of the filings. So there’s several challenges, some of which BW can address and some they can’t. * ALL LIABILITY: Legal attack surface is potentially unlimited. Nothing BW can do here. * NO VERIFICATION: The info appears to be only self-representations, and that’s just the whole ballgame. No one can rely on the filings, making them pure legal liability. BW can verify filings via sampling or direct diligence - but then they may have legal exposure. * NO INSIDER DISCLOSURES: Widen the focus from the project itself. Someone can truthfully say tokenholders control the project, but in practice that may be a single person, never mind other forms of concentration and self dealing. * CONFLICT OF INTEREST: BW itself is the wrong entity to push standardized disclosures because it sells services to these filers, making any verification lower quality to potential investors. Spin it out, with different ownership. * NO ONGOING OBLIGATION: This is the attestation problem again: it’s a specific date. Someone could one thing, file this, then go do the opposite. BW could flag which projects have committed to regular, ongoing filings.

Very deft messaging to lay down public comms that USDS is not a “payment stablecoin” which has major regulatory consequences. (I also happen to agree that DAI/USDS doesn’t meet the legal definition of a payment stablecoin, and it is good to lay down a record that Maker/Sky believes this as well.) So very good job in compliance positioning if more comms double down on that stance. Where it breaks down here is not something BD or marketing or legal can fix, though. Positioning Maker/Sky as competing as a savings product is basically throwing in the towel. Sky, to put it bluntly, sucks at making a savings product. It earns negative spreads vs sUSDS on 2/3 of its assets and has accounting standards that can charitably be described as, “at least they tried”. Only the legacy DAI supply keeps the cash flowing and the total net interest margin positive. And this is the asset Sky hates. Negative spreads, poor accounting controls, and a Top 20 holder with funds taken from CB users leave you with an asset that is financially due for a rate cut, hard to diligence due to related party accounting treatments, and a tolerance for illicit funds (which was supposed to be the main difference between USDS and DAI, other than branding) - this is not a winning formula for a savings product! And that doesn’t even touch on a yield that is below the risk-free rate. No serious institution is going to direct their users to sUSDS until these are all corrected. The bull case is that all three items are in theory fixable quickly: cut sUSDS rates/jettison negative-spread portfolio exposure, publish accounts that adhere to credible accounting standards somewhere (and get an auditor), go ahead and either remove upgradeability from the tokens or use it so users can sort by a desire for recovery or censorship resistance

I am a fan of fixed rate lending, but disagree with this take. Curator incentives are still badly broken across most protocols - they are paid by squeezing borrowers and putting lenders way out on the risk curve. The old job was picking the highest yielding pools, full stop. Risk doesn’t matter because heads I win, tails you lose. It’s TradFi but with less competence. New job looks like the old job. This is a design flaw in ALL curated markets I’m aware of. Curators should have: 1) Fiduciary duty, or at least a duty to allocate to suitable products. 2) Real junior capital. Not pari passu. Not junior-in-theory-but-not-practice. No one is better able to get out early than the curator. They need to go down with the ship. 3) Hard, immutable restraints on risk profile migration. I’ve seen vaults that start out blue chip (WBTC, wstETH, WETH) and then ended up quietly adding Stream exposure 4) Full disclosure of all side deals. Curators make shit money for the legal risk, and I strongly suspect most live off of subsidies or side deals. Payment-for-lending-flow and any other arrangements should be disclosed in full. Most curators should not manage money, and fixed rates - as wonderful as they are - does not change that.

The biggest problem a GENIUS or other fully reserved stablecoin faces is emissions. $1 has to be surrendered for every 1 stablecoin created. This is tough. The two demonstrable profit centers in stablecoins - Tether and historical MakerDAO (not the current Sky, ewww) - did so without full reserves. The good news is that clever people CAN make a fully reserved stablecoin emit in other ways (although no one seems to have noticed yet, or they simply don’t want to for fear of the Law of Reflux). The bad news is that emissions is just one part of a successful stablecoin. Circulation and redemption are the others. Circulation being the most difficult.

The success of USD stablecoins vs non-USD stables is the reaping of seeds sown in the closing days of WWII. Those in the know will recognize the Triffin Dilemma in the telling of the tale. As the war came to a close, the US stood alone amongst the major economies. France, UK, USSR, Germany, Japan, China, and most of the world’s middling economies had been badly damaged by the war and the preceding Great Depression. By 1945, almost the entire world needed rebuilding. That took money, unfortunately, and the Axis, Allied, and Communist Bloc countries had little left. Except the United States, who entered the war late and had been selling weapons and food to the Allies and USSR, often for gold, sometimes for credit. Allied and former Axis countries they controlled took massive post-war loans from the US. The USSR essentially took everything that wasn’t nailed down from conquered Axis nations occupied by the Red Army, supplemented with forced labor, to finance their rebuilding. There was a problem brewing, however. The Fed was a vassal of the administration in those days, so interest rates were held down, allowing the government to inflate away much of its own wartime debt. This made lending dollars to Europe especially attractive in a world where a long-dated Treasury bond yielded <2.5%. So dollars were exported. The US was the major creditor nation in the world. But the debtor nations eventually needed to pay those loans back. Still rekt, they didn’t have much to sell the US to get those dollars. To help allies and lenders, the US airdropped USD onto Europe via the Marshall Plan. Europe began to recover, its companies & govts able to service debt, and had goods to sell the US. And they kept borrowing from the US, where the Fed kept rates low. Dollars seemed to pile up in Europe. The USD had become the main reserve currency, since other major currencies were pegged to the dollar, it made it easy to settle in dollars for transactions in lira, pounds, francs, Deutschmarks. In short, the dollar circulating supply grew! This caused a lot of heartburn in the US. Remember that the USD was pegged at $35/gold ounce. The supply of gold in Fort Knox was slowly shrinking as dollars were redeemed by foreign central banks, but the total supply of dollars outside the US was growing (remember that USD inside the US was not allowed to redeem, so could not cause a run). Imagine a stablecoin issuer who had a large reserve, but over time the supply begins to approach or surpass your reserves. That was the US in the late 40s/early 50s. Eventually the Fed stopped taking orders from Treasury in 1951, and they began to hike rates. Now dollars (and all modern money) is manufactured by commercial banks. They take an asset worth $100 and loan $50 against it. The bank gets a $50 asset (the loan) and creates a $50 liability (bank deposit). Nothing says this has to be done in America. So European banks began to manufacture USD deposits. Eventually an Asian equivalent in Singapore popped up. And Tether has a related model. So how does this promote USD stablecoins today? Because USD monetary policy is run for the benefit of the US economy, not the global economy. That is to say, a glut or scarcity of USD can occur outside the US. When this is a scarcity, Eurodollars and Tether fill the gap. Tether’s success has less to do with crypto and more to do with supplying USD liabilities where USD was in short supply. Of course, the US avoided a run on Fort Knox by letting the Fed be independent and eventually abandoning the gold standard, since the Fed doesn’t really control the money supply very well, since they’re not the ones who manufacture them. But no one ever solved the dilemma of mismatched USD supply and demand outside the US - which provides a tailwind to USD stablecoins

I hate how no one ever tries to solve for stablecoin reflux except via incentives (except Circle, and arguably not them since they use revenue share) It’s a very hard problem to solve in general, but impossible to solve on Twitter and Telegram.

An underappreciated risk of GENIUS to USDC and similar centralized stablecoins is escheat risk. In everyday English, this to say the risk from unclaimed property laws, which send abandoned assets to state governments. Circle is famously loathe to freeze tokens without a lawful order from the government. This is understandable, because that kind of taking and seizure exposes them to litigation risk if the holder surfaces and Circle has no iron-clad justification. However, USDC was born in 2018, which is plenty of time for tokens to be viewed as abandoned. Now, unclaimed/abandoned property laws vary by state and asset type, and how centralized asset issuers handle self-custodied has not to my knowledge been tested yet. Normally, simple possession defeats escheatment. But for accounts there is an inactivity clock, usually 5 or more years. Furthermore, for payments instruments there is a *presumption* of abandonment at 7 or 15 years from issuance, depending upon the instrument. While I think redeemable stablecoins are likely at risk of presumed abandonment-style escheatment, the fungibility of ERC20 tokens makes that nonsensical as soon as the USDC commingles with its younger brethren. The blockchain, however, provides really good evidence of inactivity - 5 years of no outgoing transactions or signed messages is easy to demonstrate and explain. Currently, Circle, like CEXs, banks, and custodians, only escheats balances from inactive onboarded users. This is because the obligation falls to a “holder” of the assets. A self-custodied wallet belonging to an unknown person is arguably not someone Circle has a relationship with, and so not holding assets on their behalf. Until GENIUS, USDC lived in a kind of gray area that likely fragments the payoff for a state treasurer, lowering the motivation to pursue escheatment claims. GENIUS, however, clearly establishes that all token holders are contingent creditors of the issuer through how it handles bankruptcy and claims priority on the issuer’s reserves. That makes the anon, cold storage or dormant address with USDC arguably an issuer-holder relationship that is familiar to unclaimed property law. Notably, there’s is no safe harbor or federal preemption over state escheatment laws in GENIUS Secondly, there’s a real motivation for two states in particular to sort this out in court as the supply of escheatable USDC grows: New York and Delaware. If USDC is determined to be equivalent to a money order or traveler’s check, then New York (as the principal place of business of Circle) is lined up for a yearly windfall as inactive USDC supply ripens like grapes on a vine. If USDC is covered by more general property classification, then Delaware (as the legal domicile) is going to want that free money. An analogous fight recently played out in the Supreme Court in 2023 where Wisconsin and Pennsylvania sued Delaware for a share of escheatment of MoneyGram checks. Delaware had to cough up $191m and forgo part of an ongoing escheatment income stream. So states will be motivated! Circle, of course, is motivated to avoid this question coming up, as it would rob them of free float, which is what produces their income. It also gets messy about how to administer escheatment on a self-custodied address. GENIUS requires 1:1 backing, so simple freezing may force Circle to pay out of pocket. Burning is the cleanest answer for them, or a safe harbor that excludes frozen stables from supply. @JBSDC @millercwl @amandatums may also have views on GENIUS collision with state escheatment

It is a curious fact of history that both the English language and the dollar have enjoyed two periods of being globally dominant. The English language did so through the consecutive hegemonies of the UK and US. The dollar did so through the Spanish silver dollar and the US dollar today (the latter - like the yuan, won, yen, most pesos, most dollars - descended directly from the former), with a brief interregnum for the sterling pound. Both were channeled through the US for round 2, but given the extinction rate of both languages and currencies not descended from a silver coin in 1500s Bohemia, it feels like escape velocity for both has been achieved.

It never ceases to amaze me how people say they need more yield to use a protocol, and then when that yield appears, they say yield is a red flag. It’s as if people only want guaranteed yield via subsidies that disappears in a schedule. Anyway, it does suggest there’s demand for fixed rate lending!

The Amazon delivery guy has been pretty standoffish ever since he caught me berating the deadbeat tomato plant that lives on my porch - which continues to bloom but made zero fruit - last week. I wonder if a new delivery driver will be cool or if I’m on the Weirdo House List

DeFi is where: * Lenders get sub-risk-free rates for risky lending * Borrowers are apparently unable to run profitable trades like “borrow at less than tbills and buy a tbill” * Active, multi-strategy credit funds with struggle to provide yield competitive with an FDIC-insured savings account What are we even doing here? Are we all just trapped onchain with no way out? Even if the lenders are stuck, why are the borrowers not able to sustain very low borrow rates by real-world standards? And it’s not like we even built a DeFi that does what it was built for, which might justify the friction and low capacity for competent capital allocation. We were all on the censorship resistant, permissionless finance highway, and then the wannabe hedge fund guys grabbed the wheel and drove us into the ditch because “non-custodial software” didn’t earn them performance fees for underperforming Treasuries. It’s not even that centralized entities or replicating many TradFi structures onchain is sinful or shameful. They’re not. DeFi, CeFi, and TradFi can coexist. It’s that the onchain economy is apparently so unhealthy that the only way to remotely give lenders a reward in line with risk is via massive subsidies. All these gigantic Earn initiatives are money flowing the wrong way, swamping an already overcapitalized DeFi market where we apparently are incapable of scaling any product that’s not minute-by-minute margin or perps, the latter of which is zero-to-negative-sum and closed off from composability, so may as well be offchain from a macro perspective. RWAs were supposed to save us by letting yield flow from offchain markets to investors onchain. But all we got were “tokenized tbills” that were just nosebleed fees slapped onto a money market fund. I’m still waiting for these actual tbills so I can build a ladder of them without paying a middleman or three 60 bps of the 360 bps tbill yield. One gets the impression that onchain markets are only kept from draining into the real world by an invisible dam of CEXs’ and banks’ arbitrary freezing of funds keep people scared to off-ramp. DPRK can get the money out somehow but there’s not enough borrowers able to withstand a sub-5% borrow rate? Either there’s free money on the sidewalk or something is busted.

People invoke Gresham’s Law on here all the time, and I suspect ~0 of them (even economists) have actually gone back to read what Gresham said. Most invocations of Sir Thomas Gresham take the wrong lesson, and I see it a lot, so let’s sort it out. What did Gresham say in his correspondence to Queen Elizabeth I in 1558? He was explaining why the circulating coinage was primarily the heavily debased coins of her late brother and father. But people did not hold onto or hoard the older, more valuable coins. “all your fyne gold was conveyed out of this your realm” should be the dead giveaway to anyone parsing the Elizabethan English of the letter that he was explaining all the sound currency had been exported (to Flanders and Holland, primarily). He explains also that this is a phenomenon of the legal tender status of the coinage. Foreign partners were under no obligation to accept underweight coins, so the best coins had to be used instead, causing a net outflow of gold and silver to the Continent from England. This is NOT the story of “bad money drives out good” that gets told on Crypto Twitter, for a couple reasons. First, Gresham’s Law ONLY applies to legal tender money. Private currencies nearly always have to compete by being better than alternatives. It is only because the heavy hand of the law compelled acceptance of Henry VII’s and Edward VI’s crappy coins that could circulate at face value. Second, part of what was in play is the Alchian-Allen Theory, which explains why the best apples, best seafood, best whatever, are often exported and not available in their local markets. Transport of money was not trivial in terms of costs, and it is less expensive to ship a single chest of “good” coins than two chests of “bad” coins across the English Channel to trade partners who will discount the “bad” coins to a lower denomination. This is why large denominations existed in the first place, and why small denomination money has usually been undersupplied across all of history. Fixed transport and production costs make it relatively more expensive. So what lessons does Gresham’s Law have for crypto? Not a ton, unless some government or very powerful group enforces two stablecoins to have the same face value, despite differing actual values. Then you see those inside the ringfence of that government get left holding more and more “bad” money as the “good” money is used up trading with partners who aren’t compelled to accept the “bad” money.

A recurring issue in DeFi is the persistence of at-or-below risk-free rates. This is largely due to a failure to originate debt, and it indicates several areas of dysfunction within DeFi lending markets. First, it’s helpful to examine why risk-free rates exist anywhere. Under normal circumstances, you primarily expect to see below-risk-free rates on cash (naturally), money market funds, and bank accounts. Cash is no mystery. Money market funds come with some administrative drag - you have to pay someone to run the thing after all. And bank accounts bundle a variety of transaction, custody, and ancillary services, so the standard bank account is not an investment. High-yield savings accounts do exist, of course, but they often pay at or slightly above the risk-free rate - there is no shortage of US banks that offer >4% on savings accounts or short-term CDs. And of course the funds are insured up to $250,000. But DeFi lending deposits aren’t cash, they aren’t the risk profile of a money market fund, and they don’t come with a bundle of transaction, custody, and ancillary services like a bank account. So they’re not a good risk-reward at below-risk-free rates (and I say this as someone with money sitting in them). Now let’s turn to the cause of the low rates in DeFi. Lending is cyclical, and DeFi goes through regular periods where loan demand is quite low. As I write this, the 1-month weighted average stablecoin supply APY across DeFi sits at 3.1%. And that’s including a very wide selection of lending opportunities up and down the risk curve - this is no money market fund with impeccable collateral or a bank account with insurance. It’s instructive to think about what a bank would do if it took in more deposits than it could profitably lend out. It would go out and buy treasuries, commercial paper, or loan portfolios from other lenders. In short, it would simply buy debt. DeFi protocols generally can’t and won’t do this. It requires credit expertise - which DeFi protocols completely lack in most cases, for reasons that will become clear below - and it requires access to purchasable assets. How can I say there is generally very little credit expertise in DeFi lending? Mostly because DeFi lending is structured to avoid the need for it. Debt in DeFi is mainly underwritten by the secondary market liquidity available to service instant liquidation. This requires liquidity modeling, but allows a risk consultant to be pretty agnostic to the actual asset. While some curators and protocols have attempted to do actual credit underwriting, they have mostly shown themselves to be bad at their job. The few exceptions don’t have enough years in the market to distinguish good luck from good underwriting. There’s a huge opportunity in this space, even if you’re a 4th-rate underwriter, by the way. You’d be delivering relative outperformance. Another trait of DeFi lending is just how primitive and unsophisticated it is. Almost all lending is pure margin lending. There are no tenors on the vast majority of DeFi debt, so rates of utilization are highly unstable, sometimes even on an intraday basis. The rates are also priced by utilization, rather than a model of expected loss and recovery. This is downstream of most underwriting being done on the basis of liquidity, and to the extent that lending within a vault or protocol continues to stay underwritten based on liquidity, that’s workable. Sell-into-available-liquidity functions similar to securitization in TradFi in that it standardizes the debt product from the investor’s perspective. But as soon as you begin to let actual credit underwriting (intentionally or otherwise) into the portfolio, utilization becomes a poor way to price debt, since it becomes heterogeneous. You will over time tend to lend more against the worst collateral, simply because other ares of the market are more sensitive to risk-reward and your mispriced risk create carry trades for the borrower. 1/2

A major unforced error in crypto is treating technical dashboards as financial dashboards. Nowhere is this as obvious as with TVL of lending protocols. TVL is NOT a substitute for accounting! Let’s look at TVL defined as “Value of all coins held in smart contracts of the protocol”, and how it would treat a bank with the following balance sheet: Deposits (a liability): $100m Loans (an asset): $80m Reserves (an asset): $20m Equity: $10m The TVL of this simplified balance sheet would show up as: $100m deposits - $80m loans + $10m equity = $30m TVL Does that feel accurate to you? It should not, because it structurally undercounts economic activity. In fact, TVL - a technical metric - is treating the bank’s largest asset (its loan book) as a liability and largest liability (its deposits) as an asset! The problem is one of using the wrong tool for the job. TVL counts how many tokens are in a smart contract or group of affiliated smart contracts. That’s it. In its most simple form, TVL is mostly just counting the reserve ratio of the bank (or lending protocol). TVL is not a substitute for actual accounting, and people need to understand this. A deposit on Aave/Morpho/SparkLend/Compound/Euler/Curvance is a liability to that protocol or pool. You could put $1 trillion in deposits onto one of those platforms and TVL would become $1 trillion. But that’s not an indication of economic activity! Now imagine $999.999 billion of that got lent out. TVL has crashed from $1 trillion to $1 million. Looks bad on a chart, right? But now we’re seeing economic activity! There is a reason why TVL is not used outside of crypto - it is a technical metric, not a financial one, and any overlap is coincidental and concentrated in very basic protocols like DEXes.

I spoke with the head of a community bank yesterday - initially about stablecoin policy, but the conversation quickly expanded to ways community banks and DeFi can work together. Years ago, under the Biden administration, I was a major proponent of a deal between MakerDAO and Huntington Valley Bank (since purchased by Citizens Financial), where a $100m loan participation agreement was finalized in 2022. Those funds went primarily to business loans and construction loans, carrying interest of 5-9% (2022-23 vintage fixed and floating rate loans). But that was DeFi financing real business formation and expansion in New Jersey, Pennsylvania, Delaware, New Hampshire, and Connecticut. Under the more clear rules today, I think this could be done again - or something similar - but at a larger scale. Many people around the world would eagerly get exposure to a diversified loan portfolio underwritten by a US bank with local knowledge, financing their local community. If you’re a US bank looking to explore crypto as a funding source or to manage concentration limits within your portfolio, please reach out. My inbox is open.