# 52 | Squaring (the) Circle
Circle Business, Money Transmission, and the Return of the Monetary Triangle
Staying away from chronicle served me well.
I have never had, along the years, an attraction for the incessant drumming of the news, the overflowing stream of smart comments, the hyper-detailed analysis, the evidence of great truths hidden behind the most mundane details of our daily lives. Chronicles, in order to thrive, must continuously inflate their own importance, and because of this they remain a continuously immature game for momentarily immature human beings. Telling stories wouldn’t stay a successful endeavour if it would go back being, simply, an extremely pedantic description of the insignificance of each tick along the history of compounding. In hindsight, some moments are definitely more impactful than others, but guessing them is a statistically insignificant effort. Extrapolating current events is our version of God wasting the seventh day binging on Netflix.
Also our industry—i.e. innovating value intermediation, has its own fix when it comes to overreacting to mundanity, and that fix is regulatory commentary. And Gary Gensler, seasoned professional and serving head of the US Securities and Exchange Commission, is of such a fix the once-innocuous-loner-turned-awkward-weirdo-turned-uncomfortably-large-character Tiger King.
Responsibility lies in the eye of the beholder.
Connecting Financial Systems
Although the theoretical predicate of decentralised finance remains one of resilience against the predatory control of centralised parties, truth is we all live and breathe in environments that remain inextricably exposed to regulatory capture of some kind. As you go back along the chain, you find more and more cranes and less and less skyhooks, as Dennett put it. The even more discomfortable truth is that, when it comes to monetary resilience, in all honesty nobody seems excessively to care about true decentralisation—whatever that means: we live in a fiat-facilitated world and we think in a fiat-denominated way, we know of the (at least superficial) incompatibility of the (heavily regulated) fiat ecosystem and the (extremely opaque) crypto one, yet we seamlessly switch between the two based on need and circumstance. What people care about is usability. Thank God, or Gary, we are all less maximalists we like to think ourselves to be.
Amongst the DeFi applications with a real impact in our daily lives, fiat-backed stablecoins remain the indisputable winners. They have also been the most resilient, currently floating around 17% below ATH—was c. $155b, when compared to DeFi’s 72% below a ATH of c. $170b. The reason sits with the use-case they solve for, and with what they leverage from other systems. While the core functionalities of blockchain-native DeFi have been i) facilitating value translation—read Uniswap, and ii) providing leverage—read everything else, fiat-backed stable have provided instead a phenomenal settlement layer for personal and professional users of any kind and geography. The need was there.
Circle’s Business Model #1: Delivering Unprofitable Services
In focusing on settlement, fiat-backed stablecoin issuers have decided to leave the legacy authorities alone in their ability to manipulate the monetary supply, engaging instead in intermediating money flows—something that sovereign issuers should be happy about, if done properly. Making money by providing transaction services is and was, however, a tough business at the beginning of the 2020s. Not much fat, a lot of technologically-enabled competitors, and significant regulatory overheads. This is even more valid when the making-money part is segregated to the minting (on-ramping) and redeeming (off-ramping) part, with not much to add in the middle. And even more so if that stablecoin is not the bridge into a captive liquidity system that offers a long list of other services—as in the case of Binance.
Looking at Circle’s Form S-4 Amendment from November 14th 2022, the latest available before they decided to withdraw their application, is useful. At the end of 2021, based on Circle’s audited financials, the firm was custodian of c. $42.3b deposited by customers holding its stablecoin, $USDC—a gigantic leap from the c. $4b of 2020. Total revenue for the year 2021 amounted to c. $85m, or 37 bps over average custodial assets. When excluding costs directly related to providing those same services, that number falls to 18 bps. Not great. Circle, as a consequence, closed 2021 with an operating loss of c. $86m.
The largest component of Circle’s 2021 revenues were so-called Transaction and Treasury Services, followed by Reserve Interest Income—which is the money Circle makes (and keeps) by investing customers’ liquidity in approved forms. But what are Circle’s Transaction and Treasury Services? The S-4 defines them as:
Transaction Services: payment and custody services, mainly provided through Circle API Services, a unified payments and banking infrastructure product suite
Integration Services: fees earned by connecting Circle products like $USDC or $EUROC to third-party technology platform
Treasury Services: income in connection with the Circle Yield product, generated through lending $USDC to counterparties active in centralised blockchain-based lending markets. Circle Yield is not accepting new loans at the time of writing
Service fees are not directly linked to the amount of $USDC floating across the decentralised economy, but they are definitely a byproduct of it. Although all financial intermediation companies offer ancillary services to their customers, it has been very difficult historically for them to make a living through this kind of fees. The key profit centre of (most) financial companies1 has been the spread between the cost of assets (what they lend) and liabilities (what they borrow) based on their risk and maturity spectrum. When, as it is the case of Circle, the bandwidth of available counterparties to lend to is very very narrow—i.e. the US government only, and the interest rates paid by those counterparties basically zero, it is practically impossible to make an attractive profit.
Money transmitter reserve rules. More in detail, Circle manages stablecoin reserves according to state money transmitter laws, as well as guidance developed with the Centre Consortium. The Consortium is managed by a committee constituted of 3 individuals, a representative from Circle itself, another from Coinbase—$USDC main distributor, and a third shared nominee. In general, money transmitters are allowed to invest in a lot of very very safe stuff—cash, savings deposits, highly rated securities, US-backed securities, although Circle’s practices limited reserve deployments to cash and short-dated US government obligations, for which the firm keeps 100% of the yield.
All tokens issued and outstanding are backed by equivalent amounts of fiat currency denominated assets held in custody accounts. The Company earns interest on the fiat currency held in custody accounts and the Company’s investments in available-for-sale debt securities held for the exclusive benefit its stablecoin holders. The Company has entered into a revenue share agreement whereby the interest income earned on US Dollar assets held in custody accounts is shared with a digital asset exchange which holds USDC on its platform.
Risk and benefits. Interestingly enough, Circle segregated deposits are not immune to all types of risk. Yes, we can assume that credit risk is negligible here from a dollar perspective—with pretty much all segregated deposits invested in US government securities, but the same cannot be said for counterparty risk. Due to the massive volumes intermediated by Circle and the limit in scope of deposit insurance, out of the c. $10b that Circle was holding in cash in regulated financial institutions in 2021—24% of the c. $42b total segregated value as per their reports, only a mere $1.75m (millions—0.02%) in aggregate was covered by FDIC deposit insurance. This means that if a bank goes under, $USDC holders go in line with other creditors for the cash portion of the stablecoin backing, which is indeed 24%. It is quite a lot of counterparty risk for receiving zero (0.00%, 0 bps) income as a $USDC holder. Nobody cared when the risk-free proxy was also at zero, but that reality is far away in the past.
$USDC is not alone in being exposed to this type of counterparty risk; bank-currency, i.e. the stablecoin printed by commercial banks that fill your current accounts, are as well. It is difficult to rank objectively $USDC against a bank deposit from a risk perspective, with bank deposits having way more credit risk but a more controllable counterparty risk for small savers thanks to deposit insurance schemes. The reasons behind the success of stablecoins must therefore lie with their superior functionality: enter a bank branch on Sunday to complete an operation and you will see what I mean. Ah, sorry, you can’t, they are closed.
Circle’s Business Model #2: Profitable Low-Tech Lending
It was macro, rather than regulatory arbitrage or technology expansion, that came to Circle’s rescue during 2022. The US 3-Month Treasury Bill Rate, that at the end of 2021 was 6 bps—read 0.06%, had climbed to 4.40% by the end of 2022, a 73x increase. That is pretty good if you are in the business of lending money to that unique captive borrower. As a result, Circle’s (nominal) revenue engine started a dramatic shift as we can see below, even looking only at the first half of 2022—when the 3-Month rate had climbed to c. 1.70%.
I had fun with some gonzo-forecasting2 trying to guess what 2022-YE might have looked like by looking at the evolution of the 3-Month rate, and came with an estimate for the year of 65 bps gross revenue, and 42 bps net. That’s roughly $200m of operating revenue against the c. $42m of 2021. If we exclude non-recurring items, this could be enough to make Circle a profitable endeavour, although not by much.
It is definitely a great trajectory for the firm. Questions, however, remain:
How defensible are revenues vis-à-vis users?
How defensible are revenues vis-à-vis challengers?
What’s the regulatory posturing vis-à-vis those revenues?
(1) Retain users. The 10 largest on-chain holders of $USDC control 15-20% of the total float—source Etherscan, with Maker by far the largest holder at 5.5% or c. $2.1b. That number has been going down a lot from its ATH of c. $5.5b. Maker is a bit of a canary in the coal mine for $USDC: while divesting $USDC was cause of immense pain for the decentralised workforce of the protocol, the burden of holding $5.5b of assets at zero yield, i.e. with an opportunity cost of give-or-take $200m/ year, was way too high to swallow. Circle, by choice or most probably by design, was unable to rebate even part of the yield to the protocol. Trust me, I know for a fact that Maker tried, and then started to diversify holdings through i) other stables—$GUSD and $USDP, and ii) direct lending programs through intermediaries—somehow affiliated Monetalis now manages c. $500m in short-term US Treasury instruments. Whether this will be the beginning of a trend remains to be proven, but several parties have approached Maker to engage the Protocol Stability Module—where most proprietary liquidity resides.
Take #1: it is very difficult to retain clients in the long run when you are giving them a 0% share of the money you make through them.
(2) Defeat challengers. Retaining its position is one of the aspects where Circle has historically excelled. Its two largest fiat-backed challengers, Tether’s $USDT and Binance’s $BUSD, have tended to play a different game: one focused on global liquidity rather than regulatory compliance. That is why (US on-shore) legit users have tended to use the Circle ramp, and maybe rotate into $USDT and $BUSD (or $DAI) for liquidity (or other) reasons. For Circle, it doesn’t really matter where secondary liquidity stays, as long as reserves stay in the firm’s balance sheet. 2022 has seen a long list of challengers, both fiat-backed and decentralised, disappearing; this seems to have reinforced the duopoly of $USDT and $USDC in being the off-shore/ on-shore duet of dollar-denominated liquidity on-chain.
Take #2: if you make money by doing one very specific thing, focusing on how to protect that very specific thing is a good strategy.
(3) What about the regulator. This brings us to the last topic, the elephant in the room. What about the regulator. How defendable is actually Circle’s ramp and how regulatory-friendly does it remain. Here we draw the line of pure speculation, and that is where I feel uncomfortable—line below.
It is enough to say that narrative seems shifting, with the SEC targeting stablecoin issuers that were previously considered friends. This week it was Paxos’ turn, with the SEC issuing a notice suggesting that $BUSD should be considered a security. Paxos, obviously, disagrees. Guessing why the Commission is expanding its mandate up to this point is unnecessary and unhelpful—even regulatory agencies are made of humans with their own worries, opinions, and agendas. Is the regulator really worried about consumer protection, or is it worried about control over international monetary flows. My gut feeling tilts towards the latter, and the same gut feeling tells me it is not a very smart idea. I am not alone in thinking this way.
Take #3: if you are counting on regulatory friendliness as the core competitive advantage of your business model, you are simply foregoing control.
There is an interesting precedent. In 2017 Ant Financial’s Yu’e Bao (the leftover fund) became the largest money-market fund in the world, reaching $268b in assets under management in 2018. The success of the fund was due to its superior functionality—i.e. the link to Alipay fintech services, and superior yield—compared to traditional banks. Savers were able to seamlessly access a pseudo checking account to pay for goods and services, limiting the idle time for their savings. In some way Yu’e Bao was an earlier version (yet at the time way more powerful for consumers) of a crypto dollar. The Chinese authorities didn’t like its explosive success, whether due to increasing risks trickling down to consumers—the securities argument, to the excessive influence over the domestic bond market—the market stability argument, or a protective interest for the banking sector—the control of the flows argument, it’s difficult to tell. It was, probably, a combination of the above, as it might be for the current SEC stance on stablecoin regulation.
18 months ago, almost at the beginning of my personal DR journey, we dedicated issue #3 to the monetary triangle: central bank <> commercial banks <> borrowers. It might be worth revisiting some of the arguments from those far away days. A civil war for money is on. Unfortunately, civil wars tend to be the cruelest of wars.
We should exclude from this pure investment banks that are a very strange and heterogeneous class difficult to pigeonhole.
Disclaimer: gonzo-forecasting is a back-of-the-envelope technique never to be trusted, it doesn’t follow AI-powered models but simply personal intuition, and shouldn’t be used by anyone else than the originator, yet it’s good enough for the specific purposes it is used for.