Rethinking Ownership, Stablecoins, and Tokenization from First Principles

Addison (@0xaddi) and I have recently been discussing the huge interest in cryptocurrencies between TradFi (traditional finance) and their actual core use cases. Below, we formally lay out our conversation around the US financial system and how cryptocurrencies fit into it from a first principles perspective:

The current narrative is that tokenization will solve many financial problems, which may or may not be true.

Stablecoins lead to new money issuance, just like banks do. The current trajectory of stablecoins raises significant questions about how they interact with the traditional “fractional reserve banking system” — in which banks hold only a fraction of deposits as reserves and lend the rest, effectively creating new money.

Tokenization has become a hot topic

The narrative now is that “tokenizing everything” — from public market stocks to private market stocks to short-term Treasury bills — is good for crypto and good for the world. To think about what’s happening in the market from a first principles perspective, it’s helpful to review the following:

Currently, in the US, large asset issuers (such as public companies) grant custody of their certificates to DTCC (Depository Trust Clearing Corporation). DTCC then tracks ownership of the approximately 6,000 accounts that interact with it, which in turn manage their respective ownership ledgers for their end users. For private companies, the model is slightly different: companies like Carta only manage the enterprise's ledger.

Both models involve highly centralized ledgers. The DTCC model has a “Russian doll” of ledgers where an individual may need to go through 1 to 4 different entities to get to the actual ledger entry at DTCC. These entities may include the brokerage firm or bank where the investor has an account, the brokerage firm’s custodian or clearing firm, and DTCC itself. While the average end user (retail investor) is not subject to this hierarchy, it creates a lot of due diligence work and legal risk for institutions. If DTCC itself tokenized its assets in a native way, the reliance on these entities would be reduced because it would become easier to interact directly with the clearing house - but this is not the model proposed in the current popular discussion.

The current tokenization model involves an entity holding an underlying asset as a line item in its primary ledger (e.g., as a subset of entries in DTCC or Carta), and then creating new, tokenized representations of its holdings for use on-chain. This model is inherently inefficient because it creates another entity that can extract value, incur counterparty risk, and cause settlement/clearing delays. Introducing another entity undermines composability because it results in an extra step to “wrap and unwrap” the security in order to interact with the rest of TradFi or DeFi, which can cause delays.

A better approach might be for DTCC or Carta to put the ledger on-chain, allowing all asset holders to gain the benefits of programmability, thereby achieving native “tokenization” of all assets.

One of the main arguments for enabling tokenized stocks is global market access and 24/7 trading and settlement. If tokenization is the mechanism for “delivering” stocks to people in emerging markets, then this would of course be a huge improvement on how the current system works and open up access to U.S. capital markets to billions of people. But it remains unclear whether tokenization via blockchain is necessary, as the task is primarily regulatory. It remains to be seen whether tokenized assets will become an effective form of regulatory arbitrage, as stablecoins have been over a long enough period of time. Similarly, a common bullish argument for on-chain stocks is perpetual contracts; however, the obstacles to perpetual contracts (including stocks) are entirely regulatory, not technical.

Stablecoins (tokenized dollars) are similar in structure to tokenized stocks, but stocks have a much more complex (and heavily regulated) market structure, including a range of clearing houses, exchanges, and brokers. Tokenized stocks are inherently different from “normal” crypto assets, which do not have any “backing” but are natively tokenized and composable (e.g. BTC).

To achieve an efficient on-chain market, the entire TradFi system would need to be replicated , which is an extremely complex and difficult task due to concentrated liquidity and existing network effects. Simply putting tokenized shares on-chain will not solve all problems, as it requires a lot of thought and infrastructure to ensure they are liquid and composable with the rest of TradFi. However, if Congress passed a law allowing companies to issue digital securities directly on-chain (rather than doing an IPO), this would completely eliminate the need for many TradFi entities (and this is likely to be outlined in the New Market Structure Act). Tokenized shares will also reduce the compliance costs of listing in the traditional way.

Currently, emerging market governments have no incentive to legalize access to U.S. capital markets because they prefer to keep capital within their own economies ; for the United States, opening access from the U.S. side would introduce anti-money laundering issues.

Aside: To some extent, the variable interest entity (VIE) structure used by Alibaba ($BABA) on US exchanges already represents a form of “tokenization” where US investors do not directly own native BABA shares, but rather a Cayman Islands company that has contractual rights to Alibaba’s economic interests. This does open up the market, but it also creates a new entity and new shares, greatly increasing the complexity of these assets.

The Real Dollar and the Federal Reserve

A real dollar is an entry in the Fed’s ledger. Currently, there are about 4,500 entities (banks, credit unions, certain government entities, etc.) that have access to these “real dollars” through the Fed’s master account. None of these entities are crypto-native, unless you count Lead Bank and Column Bank that serve specific cryptocurrency clients like Bridge. Through the master account, these entities have access to Fedwire, a super cheap and near-instant payment network where wire transfers can be sent 23 hours a day and settle almost instantly. Real dollars are in M0: the sum of all balances on the Fed’s main ledger. “Fake” dollars (“created” through lending by private banks) are M1, which is about 6 times larger than M0.

Interacting with real dollars is actually pretty convenient: transfers only cost about 50 cents, and you get instant settlement. Any time you send money from your bank account, your bank interacts with Fedwire, which has nearly perfect uptime, instant settlement, and cheap transfer latency — but regulatory tail risks, AML requirements, and fraud detection cause banks to put up a lot of guardrails around large payments (which is a source of friction for end users).

Under this structure, a bearish argument for stablecoins is to expand access to these “real dollars” through an instant system that does not require an intermediary who 1) captures the underlying yield (as is the case with the two largest stablecoins), and 2) restricts redemption privileges. Currently, stablecoin issuers work with banks that have master accounts at the Fed (Circle with JPMorgan/BONY Mellon) or with financial institutions that have significant access to the U.S. banking system (Tether with Cantor Fitzgerald).

So if having a master account at the Fed is basically a cheat code for stablecoin issuers where they can get 100% of the risk-free Treasury yield with 1) no liquidity issues and 2) faster settlement times, then why wouldn’t they want it?

A stablecoin issuer’s case for a Fed master account could be rejected in the same way that The Narrow Bank’s application was rejected (and crypto banks like Custodia have also been denied master accounts). However, Circle’s relationships with its partner banks are likely close enough that a master account would not significantly improve the flow of funds.

The reason the Fed does not approve master account applications from stablecoin issuers is that the U.S. dollar model is only compatible with a fractional reserve banking system: the entire economy is built on banks holding just a few percentage points of reserves.

This is essentially how new money is created through debt and loans - but if anyone can get a 100% or 90% risk-free rate (no money for mortgages, business loans, etc.), then why would anyone use a regular bank? If they don't use a regular bank, there are no deposits to create loans and more money, and the economy grinds to a halt.

The two core principles cited by the Fed regarding master account eligibility include: 1) granting master accounts to institutions must not introduce undue cyber risks; and 2) must not interfere with the Fed’s implementation of monetary policy. For these reasons, at least for current stablecoin issuers, it is unlikely that they will be granted master accounts.

The only scenario where a stablecoin issuer might actually gain master account access is if they “became” a bank (which is probably not what they want). The GENIUS Act would establish bank-like regulation for issuers with a market cap of more than $10 billion — essentially, the argument here is that since they’d be regulated like a bank anyway, they could operate more like a bank for a long enough period of time. However, stablecoin issuers still cannot engage in fractional reserve banking-like activities under the GENIUS Act due to the 1:1 reserve requirement.

Stablecoins have not been outlawed to date because most of them exist overseas through Tether. The Fed would love to see the dollar expand globally in this way — even if not through a fractional reserve banking model — because it reinforces the dollar’s ​​status as a reserve currency. However, if an entity like Circle (or even a narrow bank) were to be used on a large scale in deposit-based accounts in the U.S. that is orders of magnitude larger, then the Fed and Treasury would likely be concerned (because it would drain money from banks running fractional reserve models, through which the Fed can implement its monetary policy).

This is fundamentally the same problem that a stablecoin bank would face: to make loans, a banking license is needed - but if the stablecoin is not backed by real dollars, then it is no longer a real stablecoin and loses its entire meaning. This is where the fractional reserve model "breaks down". However, in theory, a stablecoin could be created and issued by a chartered bank (with a master account) that operates on a fractional reserve model.

Banks vs. Private Lending vs. Stablecoins

The only benefit of being a bank is access to a master account at the Federal Reserve and FDIC insurance. These two features allow the bank to tell its depositors that their deposits are safe "real dollars" (backed by the US government), even though all deposits are loaned out.

To make a loan, you don't need to be a bank (private credit companies do this all the time). However, the difference between a bank and private credit is that at a bank you receive a "receipt" that is considered to be actual dollars. It is therefore interchangeable with all other receipts at other banks. The backing of a bank's receipt is completely illiquid; however, the receipt itself is completely liquid. This conversion from deposits to illiquid assets (loans), while maintaining the idea that deposits retain their value, is the key to money production.

In the private credit world, your receipt is tokenized by the value of the underlying loan. Therefore, no new money is created; you can’t actually spend your private credit receipt.

Let’s use Aave to explain the analogy with banks and private credit in crypto. Private credit: In the existing world, you deposit USDC into Aave and receive aUSDC. aUSDC is not always fully backed by USDC, as part of the deposit is loaned out to users as collateralized loans. Just like merchants don’t accept private credit ownership, you can’t spend aUSDC.

However, if economic actors are willing to accept aUSDC in exactly the same way as USDC, then Aave is functionally equivalent to a bank, where aUSDC is the USD it tells its depositors they have, while at the same time all the backing (USDC) is loaned out.

As a simple example to go off topic: Addison gives Bridget Credit Fund $1,000 of tokenized private credit, which can be spent like dollars. Bridget then lends this $1,000 to someone else through a loan, and now there is $2,000 of value in the system ($1,000 loaned + $1,000 from Bridget Fund). In this case, the $1,000 loaned is just debt, which works like a bond: a claim on the $1,000 that Bridget loaned to someone else.

Stablecoins: The new money, or not?

If the above argument is applied to stablecoins, then stablecoins do functionally create “new money.” To further illustrate this point:

Let's say you buy $100 of a Treasury bill from the U.S. government. You now own a Treasury bill that you can't really spend as money, but you can sell it at a fluctuating market price. On the back end, the U.S. government is spending the money (because it's essentially a loan).

Let’s say you send $100 to Circle, and they use that money to buy short-term Treasury bills. The government is spending that $100 — but so are you. You receive 100 USDC, which you can spend anywhere.

In the first case, you have a Treasury bond that you can’t use. In the second case, Circle creates a representation of the Treasury bond that can be used in the same way as the U.S. dollar.

Stablecoins have a negligible amount of “money issuance” per dollar on deposit, because most stablecoins are backed by short-term Treasury bills, which are less susceptible to interest rate fluctuations. Banks have much higher money issuance per dollar, because their liabilities are longer-term and their loans are riskier. When you redeem your Treasury bill, you get money from the government selling another Treasury bill — and the cycle continues.

Ironically, in the cypherpunk values ​​of cryptocurrency, every time a stablecoin is issued, it makes it cheaper for governments to borrow and increases inflation (increases demand for treasuries, which are really just government spending).

If stablecoins get big enough (for example, if Circle owns about 30% of M2 — currently stablecoins are 1% of M2), they could become a threat to the U.S. economy. This is because every dollar that flows from banks to stablecoins is a net reduction in the money supply (because banks “create” more money than stablecoin issuance creates), which was previously only done by the Fed. Stablecoins also weaken the Fed’s power to implement monetary policy through the fractional reserve banking system. That said, the benefits of stablecoins on a global scale are indisputable: they extend the dominance of the U.S. dollar, strengthen the dollar’s ​​position as a reserve currency, make cross-border payments more efficient, and greatly help people outside the United States who need access to stablecoins.

When stablecoin supply reaches trillions of dollars, stablecoin issuers like Circle will likely be incorporated into the U.S. economy, and regulators will figure out how to intertwine monetary policy and the need for programmable money (which gets into the CBDC space, which we’ll discuss later).