Welcome to another issue of Net Interest, where I distill 25 years of experience investing in the financial sector into a weekly email. If you’re reading this but haven’t yet signed up, you can join over 19,000 others and get Net Interest delivered to your inbox each Friday by subscribing here:
Reinventing the Financial System
Everyone thinks they understand money. People earn it, they count it, they save it, they spend it, they covet it, they talk about it, they don’t talk about it. It makes the world go round; everyone understands it.
Yet scratch away at the surface and what we all think of as something tangible dissolves into the abstract.
Economists traditionally thought that money was invented as a solution to a problem in the system of barter. They presumed that before money, people would simply swap goods and services. But barter presented a problem: the “coincidence of wants” – both sides wanting what the other has at the exact same time. My wife, a school teacher, once bartered an hour of tuition for a haircut. Her hairdresser’s son needed some extra help with his schoolwork at the exact same time my wife needed a haircut. Turns out, that’s a pretty fortuitous finding. In her case, skills perfectly aligned with preferences and time, and transaction costs were kept very low.
Fortunately, there’s an alternative: money. Money scales a lot better than barter and is simpler to process. Rather than having to price everything relative to everything else, services can be valued absolutely according to a distinct benchmark. Once money became widely accepted as a means of exchange, a system of credit could be layered on top.
The evolution from barter to money is a compelling one but it’s one David Graeber, in his book, Debt: The First 5,000 Years, regards as false. He calls it the founding myth of our system of economic relations. Looking back through history, he notices that money and debt appear on the scene at exactly the same time. Rather than money having evolved as a means of exchange, he argues it evolved as a means of accounting – specifically, accounting for credit. When my wife stepped out of the hairdresser’s, she emerged with a beautiful coiffure but also an obligation: she owed the hairdresser something. Keeping track of obligations is hard, hence money, which allows obligations to be precisely quantified. In Graeber’s reading, credit doesn’t come last, it comes first as the fundamental unit of commerce. Units of currency are abstract measures of debt and a coin is simply an IOU. In this sense, the value of a unit of currency is not the measure of the value of an object, but the measure of one’s trust in other human beings.
The link between credit and money lies at the heart of banking. When people think about how banks work, they often entertain a model of banks taking in deposits on one side, and lending out those deposits on the other. Although a useful representation, it’s not strictly accurate. As with the origin of money, credit comes first. Here’s how the Bank of England describes money creation (emphasis mine):
In the modern economy, most money takes the form of bank deposits. The principal way these deposits are created is through commercial banks making loans. Whenever a bank makes a loan, it simultaneously creates a matching deposit in the borrower’s bank account, thereby creating new money. For example, when a bank extends a mortgage to someone to buy a house, it does not typically do so by giving them thousands of pounds worth of banknotes. Instead, it credits their bank account with a bank deposit of the size of the mortgage. At that moment, new money is created.
Although commercial banks create money through lending, they cannot do so freely without limit. Banks are limited in how much they can lend if they are to remain profitable in a competitive banking system. Prudential regulation also acts as a constraint on banks’ activities in order to maintain the resilience of the financial system. And the households and companies who receive the money created by new lending may take actions that affect the stock of money – for instance, they could quickly ‘destroy’ money by using it to repay their existing debt.
I said it gets abstract! For most people, none of this matters any more than understanding how the device you’re reading this on really works. But if you want to reinvent the financial system, it does matter. And a lot of people do want to reinvent the financial system. Among them is Rune Christensen, the founder of Maker DAO, a new decentralised bank. This week, we take a deep look into Maker DAO to see how you reinvent banking from the perspective of Christensen’s creation.
Maker DAO, the Decentralised Bank
Rune Christensen became heavily involved in Bitcoin in 2011 from his residences in Denmark and China. Over time, he grew disillusioned with Bitcoin because of its volatility and turned his attention towards stablecoins instead. Unlike free floating currencies such as Bitcoin, stablecoins are pegged to other assets like the US Dollar. We touched on stablecoins in My Adventures in CryptoLand here, two weeks ago.
In 2015, Christensen spent six months spinning up a dollar-pegged stablecoin on the Ethereum platform. He called it the eDollar and posted his design on Reddit (“meager programming skills” notwithstanding).
Without realising it, he was reinventing banking.
“In the very beginning of the project, I remember we didn’t even realise, in the beginning of Maker, that we were essentially just building a protocol that did the same things as fractional reserve banking, did something very similar to how a banking balance sheet works and we were just implementing that as a blockchain protocol. We thought we were doing something completely, totally different from how money usually worked in the traditional sense.” [source]
Two years later, Christensen formalised his eDollar into a form of money called Dai. The bank he devised to create his money is called Maker DAO, and it works like this:
An investor comes into Maker DAO for a loan. He (yep, usually he) has some collateral he’s happy to keep locked in a vault. Right now, that collateral is usually a crypto asset like Ethereum. For every $100 worth of crypto assets, Maker is typically prepared to lend $66 – the gap adding a buffer of protection against a possible fall in the value of the collateral. Maker accepts the collateral and advances a loan, which it does by issuing its Dai money.
Maker doesn’t operate under the restrictions of a commercial bank; its operations are more consistent with the free banking model practiced in Scotland in the nineteenth century. Those banks were largely unregulated, allowed to issue money according to the market forces of supply and demand. The legacy of Scottish free banking survives in the form of private banknotes the three Scottish banks issue in their own name. Since 1845, the Bank of England has required Scottish issuing banks to set aside assets worth the value of notes in circulation, giving people confidence they can spend them in England even if they do get funny looks south of the Borders. (The Bank of England is responsible for checking compliance of the asset-backing rules, which it does by visiting the places where the assets are stored to count them and check they are held securely.)
Of course, if Maker locked up valuable crypto assets and simply issued lemons, nobody would be interested. Its money has to have some value, so it pegs it to the dollar, albeit differently from the way Scottish banks peg their money to the pound. In Maker’s case, borrowers pay a fee – which Maker calls its Stability Fee – over the duration of their loan. That fee can be flexed by Maker to influence the supply of Dai. At higher stability fee rates, fewer people want to borrow and so the supply of Dai is managed down; conversely, at lower stability fee rates, the supply of Dai can be managed up. Demand for Dai can similarly be managed via a Dai Savings Rate which Maker pays to holders of Dai – equivalent to interest on a bank deposit. By using these rates to influence both supply and demand, Maker can therefore maintain the value of its Dai close to its target level of US$1.00.
What makes Maker special from a bank analyst’s perspective is that it’s really quite profitable! The bank is set up as a DAO – a Decentralised Autonomous Organisation – which means it delegates governance decisions to its community. That makes it incredibly transparent. Its performance dashboard is live for all to see; discussions among its users are fully searchable online (imagine that for Board discussions of a bank); and its profits are filed monthly within days of the month end.
Maker makes money from three sources: its core lending business, a trading business (where it exchanges Dai for other US$ backed stablecoins) and from liquidating loan collateral (which can also generate losses).
In the first five months of this year, Maker earned $54.2 million of revenue. Most of its revenues came from lending, where it earned an average spread of 5.3% over the period. This represents the spread between the stability fees Maker earns from lending and the Dai savings rate it pays on Dai. At the end of May 2021, Maker had $2.4 billion loans on its balance sheet and $2.2 billion of liquidity for a total balance sheet size of $4.5 billion. This month, the Maker DAO “open market committee” proposed cutting stability fees, which will have an impact on profitability going forward, although higher volumes could provide an offset.
Costs are very low, only around $800k in the first five months of the year, so most of the revenue flows through into operating profit. Maker allocates some of its profits to a surplus buffer, which its “risk core unit” recommends is maintained at 3% of risk assets (around $70 million on the current loan book).
The bank runs a few risks.
First, it takes market risk in the event that Maker is unable to release the value of collateral quickly enough in a falling market. This happened in March 2020, when it took a hit of $5.8 million as the price of Ethereum fell by 45%. May was fine though; in the recent crash Maker made money on liquidations.
Second, there’s some counterparty risk in its liquidity holdings. Maker’s $2.2 billion of liquidity consists mainly of USDC stablecoins which are backed by US Dollars held by the Centre consortium, of which Coinbase is a member. It’s the closest thing you can get to a real dollar in the crypto world, as we discussed here two weeks ago. However, the assets aren’t checked as rigorously as the Bank of England checks the Scottish banks’ assets; they are ‘attested’ by accounting firm Grant Thornton. They’re close to US Dollars, but they’re not US Dollars. This presents another risk for Maker.
After setting aside some profits to cover these risks, the rest can be returned to Maker’s owners. These are people who own the MKR token. When Maker DAO was launched, it issued 1 million MKR tokens. Right now, there are around 82,000 holders. Andreessen Horowitz has a 2.8% share but otherwise ownership is broadly distributed. Profits are returned to holders via a buyback mechanism. Rather than being reinvested back in the business or returned to owners via dividend, surplus profits are used to buy MKR tokens in the market, which are then cancelled. (They prefer the terms destroy and burn in the crypto markets – so much more violent.) As at the end of May, there were around 908,000 MKR tokens left in circulation.
In some ways, this buyback mechanism is the logical conclusion of the way corporate finance has been developing. First there was the company. Then there were shares, which are meant to represent ownership stakes in a company, at least according to Warren Buffett, but in many cases they trade completely independently (AMC, GameStop, Clover). Along the way, dividends gave way to buybacks as a means to distribute capital. They have proven more controversial largely because they are more abstract (“nothing but paper manipulation” according to Elizabeth Warren). The token is the next stage: it’s a share without a company (Maker DAO isn’t technically a legal entity) and it’s all in on the abstraction of buybacks.
Over the last twelve months, Maker has earned $63 per MKR token. At its current price of $3,110, it trades on a trailing P/E multiple of 49x. Annualising current month earnings, the multiple is 10x. Adjusted for the stability fee revisions, the multiple goes up to between 13x and 26x depending on what level of liquidation income is sustainable. Either way, for a bank growing as quickly as this one, these look like interesting multiples.
There are a number of risks of course. One is that Maker DAO remains stuck in the world of crypto and is unable to expand into real world assets. The bank did put some real world assets on its balance sheet in April, in partnership with an entity called Centrifuge. Its strategy is to onboard $300 million of real world assets over time. But as a decentralised autonomous organisation that’s difficult because most real world counterparties want to deal with a traditional legal entity.
The second is regulation. There is a component of the Maker DAO community who push back on the idea that Maker is a bank. No wonder, given the regulatory scrutiny it attracts!
It’s going to be a long time before new banking models like Maker DAO enter the mainstream. Rune Christensen himself says about decentralised finance: “I don’t think that it will necessarily replace everything… The traditional financial system will actually largely remain the way it is. It will just replace certain parts of it that right now are really bad and really old… those things will be replaced with DeFi and blockchain, but the actual bank itself probably will remain.”
We spend a lot of time at Net Interest thinking about fintech, which is largely the front end of financial services. Entities like Maker DAO innovate at the back end. When they meet in the middle, interesting things will happen.
More Net Interest
Siobhain McDonagh, MP: Mr Greensill, are you a fraudster?
Lex Greensill: No, Ms McDonagh, I am not.
Steve Clapham and I published two joint pieces on Greensill. One, in July 2020, was referenced at the parliamentary hearing Lex Greensill was required to attend last month. The other, in March 2021, was a first blush of what went wrong in the aftermath of the firm’s collapse. Steve has written an additional piece over on his blog. He digs into the ten largest holdings of the Credit Suisse Supply Chain Fund that provided funding for Greensill-sourced assets and finds that not just one or two, but the vast majority throw up red flags. Companies using Greensill were either affiliated to Softbank (a Greensill shareholder), used to employ Greensill directors, or had no real business borrowing supply chain finance. The point we made in our earlier pieces remains clear: some businesses have limited natural scope; when that scope expands excessively, it’s time to look closer.
US neobank Dave is coming to the market via a merger with the SPAC, VPC Impact Acquisition Holdings III. Dave started out as a financial advice app helping consumers avoid overdrafts (and related fees; it reckons it has saved its customers $1 billion in overdraft fees); it only launched banking services last year.
Most of Dave’s revenue currently comes from its ‘ExtraCash’ product which provides cash advances to customers so that they can avoid going into overdraft. Going forward, growth is expected to come from interchange fees as debit card usage accelerates among its customer base.
Beyond that, the pattern follows the one we laid out in our piece on Paytm of India last week – whatever the starting point in consumer fintech, the endpoint is the superapp. Dave plans to launch savings products this quarter or next, social products in 4Q and investing and protection products over the longer term. Marketing expenses will grow but the company bets that it can absorb those through higher customer profitability. With the absolute numbers so low ($95 ARPU across ‘ExtraCash’ and interchange) it’s a tighter path than traditional banks have to tread.
A few months ago, we wrote about Pipe, a platform that allows companies to monetise their recurring revenue streams. The company went on to attain a valuation of $2 billion via a couple of back-to-back fundraisings. (We wrote about Klarna in October last year and that one has gone on to increase its valuation 4x to over $40 billion).
Pipe is one of several fintechs attacking the problem of revenue financing. These fintechs acknowledge that SaaS revenues are sufficiently recurring to lend sensibly against, allowing the borrowing company to avoid the dilution they would face if they raised equity. There’s a cross-over too into e-commerce. Although e-commerce revenues are not as stable, real-time data can render them more creditworthy than in the past. In both cases, companies sell future revenues for cash today, which they can reinvest in growth.
Another one of the fintechs in the crop is Capchase, which raised funding this week. One of its investors wrote a Twitter thread comparing its balance sheet model to Pipe’s marketplace model. He points out that “marketplace models have historically failed to dominate lending ecosystems” and references P2P lenders like Lending Club. He cites speed, certainty, privacy, flexibility, diversification and prospective financing as advantages of the direct balance sheet approach. It remains to be seen whether customers themselves have a preference but in the meantime, they’ll be happy that so much capital is being raised across the market to fund them.