How money moves, with Erik Torenberg

This week I'm joined by Erik Torenberg, who runs Turpentine (which produces this podcast). We talk about one of my long-time fascinations which ended up being pretty important career-wise: how money moves. This conversation covers, in a somewhat breezy fashion, most of the first hour of orientation I'd give a new colleague prior to diving into specifics of rails that were professionally interesting to them.

[Patrick notes: This podcast was recorded on Election Day. As always, transcripts for Complex System have my notes added throughout, set from the text in this fashion. I do not edit the transcripts with a goal of exactly reproducing what was said, but rather attempting to reproduce the soul of the conversation in written form.]

Sponsor

Check is the leading payroll infrastructure provider and pioneer of embedded payroll. Check makes it easy for any SaaS platform to build a payroll business, and already powers 60+ popular platforms. Head to checkhq.com/complex and tell them patio11 sent you.

Timestamps

(00:00) Intro
(00:49) What is money?
(02:39) How money moves
(06:00) Banking and correspondent accounts
(07:59) Clearing houses and payment systems
(12:09) The gold standard and fractional reserve banking
(18:15) Introduction to stablecoins
(19:38) Sponsor: Check
(21:31) Stablecoin models and issues
(27:30) Crypto skepticism and innovations
(30:53) Regulatory arbitrage and future of money
(36:36) Wrap

Transcript

Patrick: Hideho, everybody. My name is Patrick McKenzie, better known as patio11 on the Internet, and I'm here with my buddy Erik.

Erik: Patrick, great to be here. Excited to get into money, payments, and some banking.

Patrick: Yeah. We’re recording this early on Election Day, before anyone knows the major news story for tonight.

So, I want to discuss something that is thoroughly apolitical—like money. (That’s a joke, but not much of one.) Regardless of what happens, the Republic endures, and money will still be spendable in a couple of days. And we want to talk a bit about what guarantees that.

Erik: From your lips to God’s ears! I hope so… just kidding. Yes, it will be. Well, let’s dive in.

What is money?

Do you want to start with an overview of what money is? Then we can explore how it moves.

Patrick: Sure. Academics have written book-length treatises on what constitutes money from a philosophical perspective.

A quick disclaimer: I used to work at Stripe, and while I’m speaking in a personal capacity, Stripe does not necessarily endorse anything I say here. (Stripe is a large payments company.)

I tend to approach these topics more as a practitioner than as an academic.

To put it simply, money is anything that people broadly agree is money. But that definition isn’t exactly publishable, so academics usually describe money as having three classic properties.

First, it’s a store of value—meaning you can expect it to retain its ability to purchase goods and services over time. Now, the words "retain" and "over time" come with asterisks, because inflation is real thing in the world. It affects most economies to a degree and some quite severely. But broadly speaking, we expect money to keep working as a store of value.

Second, money is a unit of account. It lets you measure what you own or owe, and it’s a standardized way to keep those numbers consistent.

Third, money is a means of exchange. People around you will accept it in trade for goods and services.

When you find something that embodies these three characteristics, you’ve got money—regardless of whether a government has blessed it as official money, what crypto folks might call “fiat” money.

In fact, at various points in history, private money has even been the dominant form. There is an entire history about this.

But let’s assume everyone here knows what money is from daily experience. Let's shift to how it actually moves around.

Erik: Sounds great. Let’s get into it.

How money moves

Patrick: So, I want to walk people through the basics here and then layer in more complexity. Erik, let’s say we’re exchanging goods and services, and I owe you something for doing me a favor—coming on this podcast on short notice. I’d like to pay you for this.

This scenario illustrates two crucial concepts. First, when we exchange goods or services, there’s often a debt involved, and money is what we use to settle that debt. Second, debts themselves can function as a type of money under certain conditions. When those debts exhibit the properties we mentioned—retaining value, being widely accepted, and usable in exchange—they effectively act as money.

One specific example of debt we’re all familiar with is a debt owed by a bank to a person or business. That debt is called a deposit. 

Now, a lot of people might think of a deposit as something they own outright, like when you put physical currency into a bank, the bank just holds it until you come back for it, perhaps with a little interest for your trouble. But technically, a deposit is a debt that the bank owes to you. 

Let’s say we both bank at the "Bank of Podcasters," and I want to pay you five dollars for coming on the podcast today. I can instruct the Bank of Podcasters to transfer five dollars from my account to yours. Essentially, I tell them, “Bank of Podcasters, you owe me five dollars less and owe Erik five dollars more.”

From the bank’s perspective, this transaction doesn’t change their overall assets. The assets—what the bank owns—stay the same. What changes is the liability side, which represents what the bank owes to whom. After the transfer, the total debt the bank owes doesn’t change; it’s just that they owe five dollars less to me and five dollars more to you.

By sending messages to the bank, about the status of a debt between us, and a desired future world state, I have moved money to you. [Patrick notes: I have previously described “payments”, in the sense of the payments industry, as just being messages about debts which, by convention and more formal mechanisms, we have varying credence levels about.]

Now, if you wanted to turn that five dollars into physical cash, you could go to the bank and withdraw it. But since this is a deposit, it doesn’t necessarily mean you have to take out cash. In fact, most money today isn’t physical; it’s digital entries on balance sheets. Physical cash is just a small fraction of all the money circulating in the world.

[Patrick notes: There are about $2.2 trillion in legitimate U.S. currency floating around physically, against an M1 money supply of about $18.2 trillion. To save you the boring specialistic definition, M1 is “physical currency plus things that are very strikingly like bank deposits” and its cousin, M2, adds retail money market funds and small CDs.] 

Patrick: Now you can easily convince anyone else who uses the Bank of Podcasters that $5 at the bank is worth $5 to them, by the same logic. But, of course, not everyone banks at the Bank of Podcasters. We’ve got a whole ecosystem of banks out there. So, let’s expand this scenario from a single bank to two banks before diving into the broader system.

Let’s make this more real. This week, I’m working with contractors—a word of advice: house construction projects are not for the faint of heart.

Anyway, imagine all these contractors bank at the "Bank of Home Improvement" while all podcasters, like us, bank at the "Bank of Podcasters." Now, I owe a contractor for work done—hammers were swung, and it’s time to settle up. But how do I move money to someone who banks elsewhere?

What most people do is instruct their bank to handle it. So, I go to the Bank of Podcasters and say, “Make this payment happen,” and the bank will provide some magical mechanism to handle this problem.

Banking and correspondent accounts

One historical method for handling this is through what’s known as a correspondent account. This idea is often underappreciated by non-specialist, but it’s essentially: banks, like any business, can hold accounts with other banks. In this case, the Bank of Podcasters could have an account at the Bank of Home Improvement.

So, when I say, “Pay my contractor a thousand dollars,” the Bank of Podcasters debits a thousand dollars from my account, reducing the amount they owe me.

[Patrick notes: The bookkeeping convention of which transactions are debits versus credits does not agree with the usual usage of these terms in day-to-day life or in banking, to the enduring frustration of small business owners

The bank debits my account (decreasing the value of their liability to me); I credit my account (decreasing the value of my asset, the deposit owed to me).] 

Now, they use their correspondent account at the Bank of Home Improvement to complete the transaction. They instruct the Bank of Home Improvement, “Okay, deduct a thousand dollars from our correspondent account and credit it to this contractor.” Through this series of offsetting transactions, we effect a payment: I agree that my bank owes me less money, my bank agrees that their bank owes it less money, their bank agrees that they owe my payee more money, and therefore in sum I have transferred money to my payee.

For many years, this approach worked well. But as the number of banks increased, the system became unwieldy. To move money between any two banks, each bank would need a direct, trusted relationship with every other bank. This becomes unmanageable quickly.

With just two banks, you only need one connection. Add a third, and now you have three connections in a triangle. With four, it’s six connections.

Life gets very complicated very fast, in what technologists would recognize as an N-squared relationship. [Patrick notes: Many readers have passed geometry at some point and know this  is the number of edges of an N-sided polygon plus the number of diagonals you could draw, which is (of course) N * (N - 1) / 2. In CS, we say “Eh, for sufficiently large N, that’s basically N^2.”] 

And as more banks join, the connections multiply—what technologists would recognize as an N-squared problem. By the time you get to a thousand banks, you’d need close to a million direct relationships. Well, an order of a million relationships. (You can do the exact count by drawing a diagram with 1,000 nodes and then a lot of lines on that diagram.)

[Patrick notes: This instant verbal correction was brought to you by my inner child, who was a member of the math team, and is extremely displeased that he grew old, boring, and unable to multiply 500 by 999 in his head when speaking. ] 

Clearing houses and payment systems

So, to solve this scaling issue, we came up with something called a clearinghouse. This concept has been reinvented repeatedly to handle various types of payments, from old-fashioned checks to today’s electronic transactions. The principle behind a clearinghouse is straightforward.

Instead of every bank needing a direct relationship with every other bank, we put an intermediary in the middle—the clearinghouse. Each bank only needs a relationship with this one intermediary, which then holds relationships with all the banks.

Here’s how it works: when a bank wants to move money to another bank, it doesn’t send funds directly. Instead, it passes a message to the clearinghouse, which handles the netting of these transactions. Netting means tallying up all the credits and debits each bank has with others over a period—say, a day—and then balancing them out. For example, if Bank A owes Bank B a certain amount and Bank B owes Bank A an offsetting amount, those obligations cancel out.

By doing this, instead of executing thousands of individual transactions, we reduce it to one or a small handful each day. The clearinghouse then sends each bank a single net settlement request, showing either the total they owe or the total they’re due.

[Patrick notes: I’m being a bit handwavy here. A clearinghouse doesn’t necessarily have to actually custody funds at any point. NACHA, for example, has banks settle through the Federal Reserve settlement system.]

This process makes everything much simpler and more efficient. A clearinghouse could work in a completely manual fashion, on paper, back when electronic communications didn’t even exist. [Patrick notes: The Clearing House was not the first clearinghouse, but dates to 1853. That’s a few short years after the invention of the telegraph.] You could do this entirely on dead tree.

At the end of the day, if Bank A owed a million dollars more than it had on deposit, the clearinghouse would request that Bank A “make good” on that difference within a given timeline. Alongside, it would provide a detailed list of the transfers each bank would need to adjust on their internal records to effect the transfers within your own books. These work exactly as they did in the case where it was internal book transfers at a bank.

So, instead of banks coordinating and exchanging checks individually, they had this streamlined, centralized process, which drastically reduced the complexity. Today, we take for granted that payments are messages bouncing around the internet, but back in the day, every check needed to physically move between banks. Some of them needed to be shipped across the country. The logistical challenges were immense.

And there's a fascinating history here, too. Wells Fargo, for instance, traces its roots back to a stagecoach company. Their primary business was physically transporting pieces of paper, representing other physical pieces of paper, across America. This was critical national infrastructure back then, moving physical checks and promissory notes over vast distances.

The risks involved were enormous. There were losses, fraud, delays—everything that could go wrong with physical transportation. And these weren’t just theoretical risks; they had real-world impacts, especially during times of disaster and financial crisis. In market crashes, these system weaknesses became painfully clear. Could a critical check or batch of checks make it to New York by the end of the day? If not, the consequences were severe—firms unable to make payroll, cascading defaults, and the desperation that led, literally, to people jumping out of windows.

So, in a nutshell, this is the role of a trusted intermediary.

The gold standard and fractional reserve banking

Erik: I understand there’s a school of thought—often associated with the Austrian school of economics—that is critical of both the departure from the gold standard and the practice of fractional reserve banking. They argue that these shifts introduced instability into the financial system, making it more prone to booms and busts.

But my impression is that most economists believe moving away from the gold standard was actually beneficial. It’s allowed for economic interventions and adjustments that help stabilize the economy. And with fractional reserve banking, it seems to have spurred substantial growth by increasing the availability of credit.

Feel free to refine my description if needed, but I’d love to hear your take on this.

Patrick: Sure, let’s break down what the gold standard was and how it evolved.

Today, we’re used to dollar bills issued by the government, but before that, paper currency was actually issued by individual banks. This was during the so-called "wildcat banking" era. Back then, a banknote was simply a physical, transferable piece of paper that represented a deposit at a particular bank. Let’s say it was a fictional “Bank of Middle America”—the note would promise that whoever held it could redeem it for a dollar’s worth of deposit at that bank.

So instead of writing checks or using electronic transfers or going to the branch, people would physically hand over these banknotes as a way of transferring value.

We moved away from privately issued physical currency, sometimes called specie. Why?

For one, some banks started to realize that if people trusted these pieces of paper as money, who is keeping a running tally of how many pieces of paper we have printed? Being in the money printing business sounds fantastically lucrative. This led to a temptation—essentially, “Why not print more and get rich quickly?” But it quickly became a risky, unsustainable practice.

This (and other issues in the banking system contemporaneously) result in banks having individualized credit risk. As a result, banknotes (and deposits) weren’t all seen as equal. Notes from some banks traded at a discount compared to others, depending on the issuing bank’s perceived stability. This variance even became a major function of early business newspapers, which would publish lists showing the market value of different banks’ notes, helping people navigate this patchwork system.

Eventually, we moved away from this model for several reasons, one of which was to create a more stable, unified currency without individual bank credit risks affecting the value of the money itself.

Now, try to imagine how complex this would be if it were happening today. It’s one of those historical issues we’re lucky to have left behind. Imagine if a “Bank of America dollar” and a “Chase dollar” weren’t worth the same amount. You’d have daily fluctuations: maybe Chase dollars are up a couple of basis points, but Bank of America dollars are down because they reported a rough quarter. Then there’s a run on the “Bank of Bigness” dollars. Absolute chaos.

To simplify things, we moved to government-issued banknotes. In the past, just like with private banknotes, the government notes promised you could redeem them for something valuable—in this case, gold. A dollar bill was essentially a claim to a fixed amount of gold, which reassured people that the paper itself had real value. [Patrick notes: I’m being handwavy here, again, and if you’re the sort of person that geeks out on this then does the government have a publication for you. For a brief gander, take a look at the Civil War era greenback.]

Then, of course, history happened. One major change was that we banned the private ownership of gold for a while. [Patrick notes: c.f. Executive Order 6102 from 1933.] Then the U.S. and most peer nations formally abandoned the gold standard. [Patrick notes: Quite a bit later; generally dated to 1971 with the abandonment of the Bretton Woods system. But I’d note that I’m not a historian of the philosophy of money.]

Patrick: When the U.S. went off the gold standard, we suspended the convertibility of dollars. Dollars are still useful: you can still pay your taxes with them. You can still convince people to give you valuable goods and services for them. But you can’t exchange them for gold anymore at a fixed rate with the government, either at retail scales or in size. We’re keeping the gold.

Various governments still own gold, mostly in their foreign reserves portfolios. They broadly don’t have retail-accessible windows to convert gold to currency or vice versa.

As for the Austrian school, they’re strong critics of this system, favoring gold-backed money and condemning practices like fractional reserve banking. They argue that these modern systems introduce instability. Mainstream economists generally disagree, seeing the gold standard as outdated and fractional reserve banking as essential for growth.

Personally, I’m indifferent. I have seen exactly this much gold in my entire life. *gestures to wedding band* 

Patrick: Exactly. Gold, for all its historical appeal, is heavy and hard to move. It’s prone to theft, and transporting it around the world introduces all sorts of logistical headaches. So, we’ve wrapped money in abstraction layers backed by the "full faith and credit" of the U.S. government, which come with their own pros and cons. It’s easier to track on ledgers, harder to steal, and doesn’t require a physical vault. But, unlike gold, it’s also less constrained in supply—meaning we can, to some extent, “manufacture” more of it.

This is a point of contention for those who favor the gold standard. They argue that back when money was tied to gold, inflation was generally lower because you couldn’t simply create more gold. To produce more gold, you need real-world resources: mining operations, labor, ingenuity. But the constraints on gold also mean that its supply doesn’t naturally scale with economic growth.

Economists often highlight this disconnect between gold’s supply and economic value. An ounce of gold will still weigh an ounce in a century, but that doesn’t mean it will represent the same proportion of economic value over time. If the economy is producing more goods, services, and innovation, and gold mining can’t keep up, then each ounce of gold effectively represents a smaller slice of total value. This rigidity is a limitation when trying to create a dynamic, growing economy.

Our current system, where money supply is influenced both by the government and by commercial banks, does a more flexible job of matching the dollar supply to economic needs. When demand for goods and services rises, the money supply can expand to meet it, and vice versa.

Inflation, of course, is a whole can of worms—one economists love to discuss but few people enjoy. And since it’s Election Day, we can sidestep that political football for now.

Introduction to stablecoins

Erik: One thing we also wanted to discuss is stablecoins. Can you explain why they’re so impactful and what’s driving all the excitement around them?

Patrick: Absolutely. Before diving into stablecoins [Patrick notes: I’ve written about stablecoins previously for Bits about Money], let me quickly touch on the concept of a trusted intermediary.

This is a key idea we’ve built much of the financial system on. That intermediary could be a government entity—like the Federal Reserve in the U.S.—or a central bank in another country. It could also be a consortium of banks that sets up an entity to manage payments. These entities then create what the industry calls rails for transferring money.

Rails are essentially the technical, operational, legal, and financial infrastructure layers that facilitate payments. In the U.S., we have multiple sets of rails. For example, we have both the Federal Reserve’s systems and the Automated Clearing House (ACH) network, which are two different rails with distinct settlement methods and features. Think of it like how you can send a message via text, email, Twitter DM, or Slack; each channel has unique characteristics but serves the same general purpose of transferring information from one party to another.

Credit cards, interbank payments, and checks all follow variations on this trusted intermediary model. Then we add layers of complexity based on the specific needs of each type of transaction. Unfortunately, we don’t have time to explore all those details today, but that background is essential when discussing stablecoins, which aim to create a similar structure outside of traditional banking.

Stablecoins are digital currencies that attempt to offer the stability of traditional currency (like the U.S. dollar) while functioning on blockchain infrastructure. This stability—backed typically by reserves of fiat currency or other assets—aims to address the volatility associated with other cryptocurrencies, making stablecoins particularly intriguing for both payments and as a store of value in the digital economy.

Stablecoin models and issues

Patrick: Let’s dive into stablecoins, since they’re all over the news these days. As something of a crypto skeptic, I’ll say this: stablecoins are probably the most intriguing project the crypto space has developed so far. Some people are betting heavily that they’ll play a major role in how we handle physical and digital money movements in the future. [Patrick notes: My prior employer, Stripe, recently announced a plan to acquire Bridge, a company which creates stablecoin infrastructure. As always, Stripe does not necessarily endorse what I say in my personal spaces.] 

To set the stage, stablecoins are digital tokens that act as an evidence of indebtedness. That is a bit of a magic word because, unfortunately for crypto enthusiasts, it is part of the legal definition of a security.

Think back to old banknotes; they were essentially physical tokens that proved a bank owed value to whoever held the note. Stablecoins follow a similar principle but in digital form. Here’s how they work, broken down into three main models.

Money market fund stablecoins

In the first model, stablecoins are tokens that represent claims on a central organizer—think of a company or entity that guarantees each token has a one-to-one backing with actual dollars or assets. The central organizer essentially “pinky swears” that for every token, there’s a matching asset in reserve. Behind the scenes, there’s blockchain, encryption, and public-private key management, but from a user’s perspective, none of that matters much. The appeal is that these digital claims are easy to move across computer systems, making them a practical tool for transactions.

The process mirrors the way people interact with bank deposits today. Most of us don’t withdraw physical cash every time we receive money. Similarly, with stablecoins, you could theoretically redeem them for cash in the traditional banking system, but in practice, most people don’t. Enthusiasts are betting that the sheer convenience of keeping the money digital, within the crypto ecosystem, will result in most people choosing never to cash it out, just like most deposits at banks are not exchanged for physical currency on the typical day, but rather either held or used for routine payments.

So, in the first model of stablecoins, the issuer effectively operates a money market fund. They create digital tokens that, like shares in a money market, are meant to stay close to a dollar in value.

To maintain this stability, the issuer accepts deposits, issues tokens, and invests those funds in very low-risk assets—like short-term U.S. Treasury notes or high-quality commercial paper. When someone redeems their tokens, the issuer uses these liquid, short-term assets to buy them back at close to a dollar, on demand and at scale. This setup creates the “stable” in stablecoin.

[Patrick notes: In practice, many stablecoin issuers work closely with a small handful of what the cryptocurrency community calls “market makers” and who traditional finance would say are similar to the Authorized Participants in a mutual fund. Rather than having a broadly accessible window for conversion between dollars and stablecoins, the issuer interacts mostly with a few participants known very well to it who transact routinely in very large sizes. They buy the coin when it is below $1 and sell the coin when it is below $1, occasionally rebalancing by buying the coin directly from the issuer or redeeming the coin for actual dollars.

This is operationally easier for the issuers than having millions of relationships with end-users, which is a different business entirely, and tends to be delegated to exchanges in the crypto universe. Also it allows stablecoin issuers to pretend they are not running large money transmission networks without anything like the level of AML or KYC controls required to do that licitly. “Look, we only have five customers, and we know them very well indeed! What happens with the coins after they buy them is out of scope for us!”]  

Patrick: Now, they’re careful to avoid the language of “shares” in the U.S., as that would trigger certain securities regulations, so instead of calling them shares, they refer to them as “coins.” It’s a tokenized claim, backed by assets, that ideally holds its value through transparent investment and redemption practices. But this model requires trusting the issuer—a team of bankers managing those reserves, who, like any financial institution, can be susceptible to various risks or compliance mandates from regulators.

“Algorithmic” stablecoins

The crypto community’s natural skepticism of centralized entities leads us to the second model, where stablecoins try to minimize reliance on a trusted sponsor. In model one, a well-regulated issuer might have to enforce anti-money laundering (AML) and know-your-customer (KYC) rules, which introduces the influence of regulators on stablecoin operations. For those wary of centralized control, that’s a drawback. They’d rather have a system where the only thing you need to trust is the code and the math behind it, with no human intervention or regulatory compliance.

[Patrick notes: This is more a pitch than it is a description of engineering reality. In fact, humans are quite involved in the governance of most algorithmic stablecoins with any adoption. They are frequently described as being decentralized. Personally, when I see a software startup with an engineering team, investors, a cap table, and a community marketing function, “decentralized” is not the word I reach for, but crypto advocates like to posture that if you call it a DAO and not an LLC, pay the members in utility tokens and not shares/units, and do the work on Discord rather than Slack, then maybe, just maybe, securities laws don’t apply to you.]

This brings us to the so-called algorithmic stablecoin. Here, instead of traditional reserves, the value of the stablecoin is backed by a pool of assets and algorithmic controls. The idea is to create a dollar-equivalent token from a mix of collateralized assets and mathematical rules. [Patrick notes: We have a word for this in traditional finance. It is “banking.” And if you thought the crypto people hate when you mention that they are issuing securities they really, really, really incandescently hate when you say they are banking.]

Now, there are better and worse ways to do this—many of the worst have ended in collapse. [Patrick notes: Terra/Luna is the most prominent example, and eventually dominoed into the collapse of much of the industry, but the graveyard for failed algorithmic stablecoins is full and the tombstones are monuments to hubris.]

But the lure of an algorithmic system that magically generates a dollar-like claim from assets that don’t include actual dollars is strong. It’s akin to a perpetual motion machine for finance.

Several smart teams have taken a crack at this approach, and others likely will in the future. Some models involve a collateral pool with two tokens: one to carry risk, the other designed to stay stable at about a dollar. Using a mix of collateral ratios, incentives, algorithms, [Patrick notes: and frequently human-powered market stabilization mechanisms which you’re not supposed to ask about], they aim to balance this token’s value across various market conditions. The claim is that these dollar-like tokens can remain stable in nearly all conceivable future scenarios, though history suggests there are some sharp limits to that confidence.

In short, algorithmic stablecoins attempt to replicate the stability of a dollar without physical or regulated backing—a high-stakes game that can work well… until it doesn’t.

But, who knows, maybe someone will create one that works. It might not be physically impossible.

Fraud-backed stablecoins

Now, onto stablecoin model number three: this is where a stablecoin issuer claims they’re running a money market-style stablecoin but is, in reality, doing whatever they please with the reserves. When people deposit funds, those funds could go into any investment or asset the issuer chooses, regardless of its actual stability. And as for audits? Well, they might worry about those only if the federal government knocks on the door a decade or so down the road.

This model has dominated the stablecoin landscape for years, and yes, I’m subtweeting here—I’m talking about Tether. Tether, which I’ve publicly described as the largest financial fraud since Madoff, has an opaque reserve structure at best. It’s now grown much larger than Madoff ever did. While they claim backing, the actual nature and quality of their assets are murky. They’ve been caught in documented lies about those reserves, and while they might be lying less these days, let’s say “trust” in Tether exists on a spectrum.

We know they now have at least some assets within regulated institutions in New York City. Beyond that, there’s a lot of ambiguity in what backs Tether. They have frequently lied about this in the past.

So, to recap, we have three main approaches to backing stablecoins, each with its own mechanics. These tokens then move freely on blockchains, making transactions easier but creating varied levels of underlying risk.

Crypto skepticism and innovations

Now, since I’m a crypto skeptic, I’ll try to give the enthusiasts a bit of balance and say something nice.

The financial industry has, let’s say, a spotty track record when it comes to technological modernization. It was an enthusiastic early adopter of mainframe computers and went all-in on that tech. But for decades, the industry largely stayed there, relying on those systems for core operations.

In the meantime, tech advanced: personal computers, mobile, APIs—all of which opened new possibilities for financial systems. And while many institutions saw these developments, they often treated them as passing trends rather than strategic priorities. You can conceive of crypto less as technical innovation and more as a distributed social movement which created and incentivized a way to make the financial industry take infrastructure modernization seriously.

Historically, banks viewed tech spending and technologists’ salaries purely as business costs. The average bank CEO thought, “Let’s keep that number low—geek projects aren’t as pressing as our balance sheet, credit committee meetings, or landing the next big client.”

One could argue that blockchain mania got crypto folks in the door with top banking leadership. They managed to get meetings where they essentially said, “Let’s talk about how your back-office operations work.” Suddenly, discussions about technological updates, operations improvements, and better standards for moving money around were happening in rooms where traditional IT teams hadn’t been able to make those issues stick. And they might have achieved that by saying the word blockchain a lot, in ways that didn’t mean that much.

The truth is, there are pain points in how money moves, how securities are accounted for, and more. But with thousands of banks globally, each needing to coordinate, there’s an inherent problem: mutually agreed-upon standards are essential, but historically, they’ve been hard to come by. Often, when the industry tries to create a standard, and the resulting dozen-plus standards fall short, someone inevitably designs a new one that “finally” fits the need—leaving us with yet another standard to wrangle.

So for years, banking tech has relied on duct-taped solutions, layered over systems dating back to semaphore wires. Until recently, no significant, organized group was consistently pushing for a real overhaul. Crypto has, perhaps, provided that push and sparked cross-functional, cross-institutional conversations that were long overdue. And some people think that this momentum will continue, and that crypto is finally the long-awaited chance to reimagine financial infrastructure from first principles rather than merely layering kludge after kludge on legacy systems.

If one believes this, the actual artifact that is blockchains, the Nakamoto consensus algorithm, etc, probably don’t matter all that much. The thing that does matter is an effective distributed advocacy campaign for software stack modernization. If you blockchain the blockchain to the blockchain, costs go down and revenue goes up? Wow, suddenly a geek is speaking my language, as no geek has in the 50 years this financial institution has employed them. [Patrick notes: This is all quite tongue in cheek.]

Partially, I'm joking. And partially, this is basically every conversation that e.g. McKinsey had with e.g. State Street over a 10 year period. [Patrick notes: One has to be a Kremlinologist of the financial industry to read “We have made significant progress on a number of tokenization proofs-of-concept, including a pilot using distributed ledger technology (DLT) to automate the calculation of margins and the pledging of related collateral for foreign exchange forward transactions.” correctly. My gloss of it is “State Street doesn’t use blockchain for anything interesting, but some of us got promoted for it anyway.”]

Perhaps some of these projects will eventually come to fruition. Perhaps someday this will someday be an important rail on which messages about debts get passed. I’m not holding my breath, but it is fair to say that many smart people seem to like them.

Regulatory arbitrage and future of money

Erik: How much of this is real technological innovation versus just regulatory arbitrage? Or is even that a bit too generous in describing what's going on here?

Patrick: There’s definitely a lot of regulatory arbitrage—or let’s be direct, evasion—happening. But I’ll start with the positive side.

Stablecoins do introduce some genuinely innovative elements. For instance, stablecoins enable a public specification that’s accessible to anyone on the internet. From there, people can create wallets, user interfaces, or even set up automated payments in response to invoices—all without needing specific permission from a bank. In the traditional system, doing something similar is far from straightforward.

People in the crypto world often tout stablecoins as the future because they represent “programmable money” in a more accessible form. It is very important to understand: actual money is also programmable money. You can get an API key from Chase.

But the traditional system is not “permissionless programmable.” If you want API access to a bank like Chase, it requires negotiating with Chase and meeting their legal and compliance standards. That means engaging with Chase’s legal and compliance teams who will walk you through anti-money laundering (AML) and know-your-customer (KYC) compliance, plus thousands of pages of other policies. And if you do not pass muster, they will shut you down.

In the early days, stablecoin issuers like Tether were quite transparent about their goal: transferring U.S. dollars without AML or KYC compliance. Over time, they adapted their messaging, now claiming to be “active partners” with federal authorities, freezing suspected illicit funds, and so on.

Tether’s claim to be an enthusiastic partner to law enforcement is a self-serving lie.

Some other issuers are more rigorous and perhaps closer to the standards of a traditional financial institution. (And, even among regulated financial institutions, there is a spectrum of how buttoned up they are.) 

Patrick: So, the “permissionless” aspect of stablecoins is actually fairly innovative.

The public key cryptography bit? Ehhhh. 

Bank websites have used SSL certificates for decades—it’s the foundation of secure online transactions. To the extent we actually care about the crypto in crypto, the novel bit is building payment rails where possession of the key is seen to represent ownership/authority, versus ownership/authority being something that is based on the legal system and merely represented in a computer.

In a way, it’s like creating a CPU in Minecraft. It’s possible to build a fully functional, programmable processor there—cool in theory, but I would not run production workloads on a Minecraft CPU. Still, it’s an interesting model to have out there, showing what’s technically achievable.

Now, on the topic of KYC and AML requirements, I find myself in agreement with some crypto advocates. These regulations, while intended to curb fraud and money laundering, disproportionately affect people on the socioeconomic margins and often fail to catch well-resourced money launderers like e.g. Sam Bankman-Fried. Ironically, the same “blockchain, blockchain, blockchain” magic words that got crypto players meetings with bank executives are now helping bring regulatory topics like KYC and AML into the conversation in Washington. This shift could open the door to potential reforms or carve-outs in these regulations.

For many people worldwide, stablecoins represent a practical way to access the U.S. dollar. Americans enjoy high-quality access to the dollar, but outside the U.S., reliable access to dollars is far more limited and usually requires dealing with intermediaries. This is intentional. It’s a stated policy goal of the U.S. to restrict dollar access globally to serve broader geopolitical objectives, which is why you see restrictions not just for Russian oligarchs but also for ordinary people in less developed countries.

Stablecoins could, with the right political momentum, open the U.S. financial system to deserving people globally—while still limiting access for sanctioned individuals. It’s a delicate balance, but with stablecoins and a bit of policy reform, we might actually create a model that offers controlled dollar access to the world without compromising U.S. policy priorities.

And that potential for political advocacy could indeed lead to some real changes. Again, today being Election Day, I’m staying true neutral on the whole “which party is better for crypto” debate. There are plenty of narratives floating around in the crypto space about which side supports their interests more, but I think they’re putting the cart before the horse.

To be clear, I’m also neutral on what specific sequence of events might eventually enable large-scale stablecoin transactions for underserved economies. Advocates for crypto have argued that we are already seeing substantial successful-Internet-product scales of adoption for stablecoins in routine payments in less developed countries. To the extent that these advocates describe stablecoins not as the future of payments but as the present of payments, I think they are exaggerating, in a fashion which is not exactly unknown in the startup community. 

Erik: Well, with that, we’re at time. This has been a fantastic discussion, and by the next time we chat, hopefully we’ll know the outcome of the election.

Patrick: We’ll have less uncertainty about some things, and the usual amount of uncertainty on others.

Erik: Yes, indeed.

Patrick: Great. Thanks very much, Erik, and thanks to all of you for tuning in. We’ll see you next week!