Fat Tailed Thoughts: Bad Money Rising, The Difference Between Starbucks and a Bank
There's a crisis brewing - bad money. Just one company holds $39B from 400M users. The protections? Little more than the company's word. We've turned bad money good before and we can do it again.
Hey friends -
This is a letter I've wanted to write for some time. On the back of a March 30th congressional hearing on the "Expanding Access to Affordable Bank Accounts Act," the time has come. I'm indebted to Daw Awrey and his brilliant paper Bad Money for much of the research.
We're talking Bad Money Rising, the Difference between Starbucks and a Bank.
In this week's letter:
Bad Money: why bank money is good money, the rise of bad money, and the path forwards
Stripe accepts crypto, Piketty's stuck in the 80s, and more cocktail talk
Last of the Oaxacans, a seemingly minor change to The Last Word creates an entirely new cocktail
Total read time: 12 minutes, 3 seconds.
Bank Money is Good Money
Thompson’s Bank Note Reporter. You probably haven't read it. There's not much of a reason to nowadays.
Circulation peaked in 1855 at 100,000 subscribers, the most popular periodical of its kind. Thirty years later it was no longer in circulation. Founder John Thompson could thank the National Banking Act of 1864 for that.
Pause for a moment and think about that year, 1864. It was the middle of the Civil War. Ulysses S. Grant's just taken over command of the entire Union army and is about to unleash total war on Virginia. General Sherman is just months away from his March to the Sea where he'll raze 20% of Georgia's farmland to the ground. President Lincoln's up for reelection.
And Congress passes a banking act. What could have possibly been so urgent?
The dollar and the national banking system
Before the National Banking Act, banks were chartered on a state-by-state basis and had license to print banknotes. A citizen could deposit their gold at a bank and receive that bank's notes in return.
Publications like Thompson’s Bank Note Reporter were critical information. They ranked banknotes based on the expected solvency of the bank and broadcasted counterfeits whenever they were found. At one point there were thousands of species issued by over 700 banks listed in Thompson’s alone.
Banknote reporters were quite literally how you knew if you held good money. It wasn't just the public perception of creditworthiness that mattered, distance to the bank and the strength of the domicile state's regulatory framework were key inputs too. The publication let you know if you should expect your dollar banknote to be worth a dollar or, as was often the case, something less.
That wasn't a great setup especially if you needed to fund a war. The National Banking Act fixed it.
The act created the Office of the Comptroller of the Currency (OCC) and gave it the authority to grant national bank charters. Nationally chartered banks were required to purchase US government bonds equal to one-third of their paid-in capital - a nice way to fund the war - and the bonds were to be deposited with the Treasury. There, they served as collateral against a special banknote that could only be issued by nationally chartered banks. These notes were identical in all respects other than the name of issuing bank and signatures of its officers. National banks were obligated to meet any redemption request in gold or silver coin. If a bank couldn't meet a request, the Treasury would liquidate the bank's bonds and pay the noteholders directly.
Said more plainly, the National Banking Act created the version one of the dollar. It's since gone through another major iteration in the early 1900s and is now issued by the Federal Reserve as a government obligation rather than a private bank obligation. But many of the essentials date to 1864.
I'd say it's worked reasonably well. We don't spend a lot of energy worrying if the grocery store or barber is going to accept our dollars. We don’t discount dollars issued by the Kansas City Federal Reserve because we live in New York. We simply pay. That's as meaningful a sign of success as any.
Yet we seem determined to push the boundaries of good money. We've pushed them so far that we accept billions of dollars of bad money without much of a second thought.
Privately issued good money
Full government backing is one way to create good money. Bank deposits and money market funds (MMFs) are two others.
The Starbucks reloadable loyalty rewards program is not. Money services businesses - like PayPal- are not. But there's still $41 billion of customer money stored as IOUs at the latter two companies. It's bad money.
Regulations turn the IOUs - private money liabilities - issued by banks and MMFs from bad money into good. The regulations do such a good job that when we analyze money supplies, we find that the stock of all good monies including privately issued is 3.5x the stock of just government-backed monies.
The regulations to make bank deposits and MMFs good money differ. Four mechanisms give soundness to bank deposits:
Lender of last resort facilities - in times of crisis, there must be a fall-back lender who will extend credit to illiquid (but not insolvent) lenders so they can weather the storm. The Federal Reserve serves this purpose today.
Special resolution regimes - in the case of bankruptcy or default, depositors get a Disney Fastpass+ equivalent - they get to skip straight to the front of the line. No waiting years for bankruptcy resolution, you get your money back fast.
Deposit guarantee schemes - FDIC insurance, sweet and simple. Deposits are insured for up to $250,000. This is quasi, but not direct, government backing. The FDIC is a government-owned corporation, but the insurance is funded by the banks themselves.
Prudential regulation and supervision - all this good-money backing creates moral hazard for the banks. Reserve ratios to ensure liquidity, minimum capital requirements to stave off bankruptcy, and mandated stress testing and reporting all help stave off potential catastrophe.
MMF regulations by contrast focus almost entirely on the assets that back the fund:
The fund can only invest in high-quality, low credit risk assets that have less than 397-day maturities
At least 95% of the fund has to be in investments that can be sold at-or-near full value within one week
At least 30% of the fund has to be in cash, zero-coupon government securities with less than 60 days maturity, or other payables due within five business days
At least 10% of the fund has to be in cash, government securities, or payables due within one business day
There are additional more nuanced benefits granted to MMFs to keep their value-per-share close to $1. One benefit is the so-called penny rounding rule where MMFs can round the value-per-share to the nearest penny. Another is being permitted to value investments at cost rather than at market value, with interest payable accruing over the remaining term of the investment. New rules published in 2008 continue to improve on the existing model.
The net result of these regulations is that we treat bank deposits and money market funds as near-dollar equivalents - good money. This is money that's been tested during the Savings & Loans Crisis of the late 80s and in the depths of the Great Financial Crisis of 2008. It's held up through it all.
Bad money has not.
Bad Money Rising
Starbucks is not a bank.
Starbucks has almost $2 billion of customer money on its balance sheet. That's more than the total assets of 80% of the FDIC-insured banks in the US, a full 3,919 banks. Deposits at the banks are guaranteed by FDIC insurance. The banks themselves are subject to a whole slew of regulations. The money at Starbucks is backed by... the full faith and credit of Starbucks? Right.
In times of crisis, you don't want to rely on the money stored in your Starbucks loyalty account. Yet there's $39 billion in "funds payable and amounts due to customers" on PayPal's balance sheet. There's not a lot more backing that money than the money stored at Starbucks. Yet despite the lack of protections, money held at PayPal and other money services businesses is often thought of by consumers as no different than a checking account at a bank. Entire businesses are run bank-less through PayPal accounts. Yours truly often has meaningful sums of money stored at PayPal and its subsidiary Venmo despite knowing better.
The proposed congressional act mentioned in the intro to "Expand Access" misses the point entirely. “Affordable bank accounts” already exist. They simply do so outside of the banks and without appropriate protections.
To be clear - this isn't an attack on Starbucks or PayPal. To the best of my knowledge, neither has ever had problems meeting customer redemption demands and both seem to be upstanding corporate citizens that meet their regulatory obligations. I'm focused on the regulations themselves - they're woefully inadequate.
Money services businesses become demand deposits
Many money services businesses (MSBs) are old, like Western Union which was founded in 1851. They come in a lot of shapes and sizes because of the breadth of the definition - businesses that affect interstate or foreign commerce and are involved in transferring funds on behalf of the public by any and all means. In a software-enabled world, that's plenty of room to morph into something new and unexpected including the likes of PayPal and Coinbase.
MSBs register federally with the Treasury, but its involvement is light-touch. Most of the regulatory burden sits with the states, and each state gets to decide its own framework.
States generally take a "three-legged-stool" approach to regulating MSBs:
Minimum net worth - MSBs must have a minimum amount of retained earnings and equity. Often the amount is calculated based on the number of physical locations.
Surety bond, letter of credit, or related security requirements - MSBs must put aside a minimum amount of money to meet customer obligations in the event of bankruptcy or other disruptions.
Restrictions on permissible investments - MSBs can only invest customer funds in certain types of instruments.
Notably missing from the list is a "special resolution regime" - how to treat customers' funds in bankruptcy. Nonetheless, it seems to be a good start but the reality is less kind. Dan Awrey's analysis of PayPal, summarized below, lays bear the issues.
Net worth? PayPal could meet all 50 state requirements with just $3 million in retained earnings and equity held in Washington and Oklahoma. Four states don't even have net worth requirements and majority allow MSBs to meet them at an aggregate level, not state-specific.
Surety bonds or similar? Most states do have state-specific requirements but they’re far from sufficient. PayPal could meet its obligations with $42 million set aside against their billions in customer funds. Two states don't have requirements at all.
Permissible investments? Twelve states have no restrictions. Thirty-one states allow MSBs to invest in their own affiliates' accounts receivables, quite literally lending to themselves.
Only twenty-eight states have considered what happens in the event of bankruptcy. All twenty-eight require customer funds to be moved into a segregated trust with the customers as beneficial owners. The other twenty-two states have no requirements whatsoever.
All-in-all, not great. We saw the inevitable outcome of this in 2008 when MoneyGram, a large MSB, had to be bailed out by a private consortium to the tune of $1.5 billion. PayPal is significantly bigger and it's just one company. The world of MSBs has exploded to include cryptocurrency exchange Coinbase, USDC stablecoin issuer Circle, and many others.
But this isn't a new problem. This looks a lot like the world before the OCC, before the National Banking Act, before we put rules on the books to create good money. We know how to solve this, we've done it before.
Turning Bad Money Good
There are three possible solutions. Two are rather messy even though they seem to be the obvious choices. It's the third that's the most viable.
Option one is to get rid of this entire "private money" business and only have government money, likely including a new digital money that replaces stablecoins. Even if we put aside that this would meaningfully undermine the entire banking sector, the reality of a government-owned payments monopoly is unlikely to be enticing. Turning our financial sector into the US Post Office won't end well.
Option two is require a banking license to issue money-like demand deposits. MSBs would either have to get bank charters or exit the business. The problem is that most MSBs don't have much in common with banks. They move money from one place to another and aggregate funds to make investments - they don't lend the way banks do to create new money. At best, the burden of bank regulations would stymie money-movement innovation. At worst, the regulators responsible would spend their limited resources trying to understand MSBs and fail to attend to their existing bank supervision responsibilities. The solution also invites potential catastrophe when it’s first implemented - how to prevent a run on the MSBs that will no longer be allowed to issue demand deposits.
Option three is to treat MSBs like something akin to money market funds. It's elegant in its simplicity and a reasonably good fit. Restrict the investments MSBs can make with customer funds so that customers are reassured that their money is safe. Prohibit MSBs from taking on other financial debts so there's no competition over customer funds in the case of bankruptcy. Require regular reporting to ensure MSBs can meet redemption requests even in times of stress. All of this is standard in the money market world.
An MMF-like regulatory regime is flexible enough that it could be implemented in relatively short order for most MSBs without materially disrupting their businesses. In the past year, Circle voluntarily updated the mix of funds backing its stablecoin to just cash and cash equivalents. If the world's largest stablecoin issuer can make that transition for $52 billion in customer funds in under a year, other MSBs can certainly follow suit.
Conveniently, there's even a regulator federal who could take it on - the OCC. It's already responsible for the safety and soundness of banks. Extending that purview to a new form of institution is not a stretch. It also brings visibility into the flow of funds among money-like instruments under a single regulator. Such visibility can uncover systemic risks that might otherwise be obscured.
Time will tell if we move to such a regime. The endless congressional hearings on stablecoins may well become the proverbial straw that breaks the camel's back. Stablecoins are higher profile and growing quickly, but MSBs are the larger exposure today. My sincere hope is that it happens proactively before customer funds are lost rather than in the aftermath of a disaster.
Stripe now supports crypto businesses. Consumers can purchase crypto, NFTs, and the like using credit and debit cards no differently than they would any other good. For me, it's a surprising development. Stripe isn't the master of its own destiny here - it connects to credit card networks who in turn connect to banks. The card networks and banks set the rules of the road - what types of payments will be processed - and they're only just starting to develop the risk practices and operations necessary to support crypto. (The Block)
"Hands on the Wheel Money." Not a concept I had thought about previously. Fintech Takes combines the apparent risk-taking predilection of Gen Z I covered in a previous letter with the unbundling of banks and predicts a rather unexpected future. Many analysts have predicted the future of fintech to be like a smart assistant that's monitoring your finances and always looking out for your best interests. Fintech Takes proposes a different evolution - millennials and Gen Z have been burned by traditional finance, they want better tools so they can be right in the driver's seat. Thanks to Refind for sharing. (Fintech Takes)
Thomas Piketty is a giant among living economists. Agree or disagree with his conclusions, you're obligated to engage with the tremendous volume of economist history he's uncovered. I greatly respect Piketty and vehemently disagree with his conclusions. But it came as a surprise to me to see him deliberately misuse income measurements to further his own point. He focuses his recent NY Times interview on the declining growth of national income per capita in the post-Regan era. "National income" is a poor way to measure income as Piketty well knows. It excludes capital gains and employer-funded health care. In a world of software-based innovation, tax-advantaged retirement accounts, and ballooning healthcare costs, Piketty chose a metric that conveniently ignores all of it. It's as much surprising as it is disappointing. I hope he rectifies it in subsequent discussions. (NY Times)
We often treat media as many distinct industries - books, film, video games, and so on - each with its own economics and nuances. Digital Native did a great roundup of disparate business models and how they're innovating. Keep in mind - yes they're separate industries, but they all compete over the same limited resource: free time. A publisher's biggest competitor probably isn't other publishers, it's Netflix & chill or the newest Call of Duty video game. Thanks to Colossus for sharing. (Digital Native)
Your Weekly Cocktail
The first of our The Last Word variants and one of my favorites.
Last of the Oaxacans
0.75oz Green Chartreuse
0.75oz Maraschino liqueur
0.75oz Lime Juice
Pour everything into a mixing glass. Add ice until it comes up over the top of the liquid. Stir for 20 seconds (~50 stirs) until the outside of the glass is frosted. Strain into a coupe glass (a rocks glass will do if you don’t have one) and enjoy.
This is the first of our The Last Word variants, just swapping mezcal for gin. It's remarkable how much different the drink becomes. Mezcal, especially Del Maguey Vida Mezcal, has both a lot more body and flavor than gin. Where gin is light on the tongue, this has a presence. Where gin is floral, this is earthy and smokey. In many ways it's a more balanced drink - mezcal stands up to Chartreuse in a way gin simply can't. I liked the cocktail so much when I first had it that I updated my standard mezcal margarita to incorporate Luxardo. That's about as high of praise as I can give.