Okay, I lied, this one’s for smart people…
For the uninitiated, MakerDAO is the organization behind the Dai stablecoin and its accompanying decentralized credit system. In my last post, Maker for Dummies, I explained the basic principles of this system in broad strokes. If you’re unfamiliar with Maker, I recommend you check out that post and our official blog / documentation (a system-wide FAQ is coming soon and I will link to it here when it’s done).
During my tenure as Maker’s Head of Business Development, I’ve explained Dai and CDPs to countless individuals, startups, corporations, and nonprofits — it never fails to inspire when presented in its entirety. Although I believe Maker for Dummies provides a satisfactory introduction to how the system works, it fails to articulate the bigger picture and vision of Maker. In short, this post is going to be about why we’re all so excited and why I think that you should be as well.
Note: I have geared this towards those with an understanding of the financial system as it currently exists. If you’re not familiar with finance jargon, you may find yourself doing a fair amount of Googling. Also, this post does not refer to the current “Single-Collateral” implementation of the Maker system, but to the next iteration which will be called “Multi-Collateral.”
Breaking Down the Smart Contract
First, let’s have a quick refresher of what Maker’s smart contract fundamentally does (simplified for this example):
- Accept user’s pledged collateral
- Check parameters for lending against this specific collateral
- Create a Dai facility against the collateral
- Given the loan is paid back in full with interest, release the collateral and send part of the interest payment to MKR holders by burning MKR tokens.
— — — — — — — — — — — — — — — —
- If the loan goes below a certain threshold, liquidate the collateral for Dai up to the value of the loan — destroy this Dai
- In the event of a liquidation, if there is not enough collateral to cover the value of the loan in Dai, dilute MKR token holders by printing MKR tokens and selling it for Dai until there is enough to offset the loan
User flow with real numbers (assume a minimum 150% collateralization ratio and a 1% stability fee):
- Deposit $1000 of “Token.”
- Borrow 500 Dai (50% LTV or 200% collateralized).
- Hold loan for one year and pay back 505 Dai (1% APY).
- 500 Dai destroyed.
- Liquidate 5 Dai for MKR tokens and burn them.
- Withdraw “Token.”
— — — — — — — — — — — — — — — -
- If the value of “Token” < $750 (<150% collateralization ratio), liquidate enough “Token” to raise 500 Dai.
- Return excess “Token” to the user.
- If “Token” liquidation results in < 500 Dai raised, print sufficient MKR and sell it until there is a total of 500 Dai available.
- Destroy the 500 Dai.
It may seem simple, but it is, in fact, a potent little program. In these few lines of code, we have effectively created (a) a better form of money, (b) a decentralized credit facility for secured lending, (c) a more efficient central bank (d) a decentralized insurance company, and (e) the infrastructure to create any kind of derivative in a trustless environment. I’m aware these are bold claims, but let me explain…
The Bitcoin Fallacy
Bitcoin, when announced, purported to be “a purely peer-to-peer version of electronic cash [that] would allow online payments to be sent directly from one party to another without going through a financial institution.” While it executed spectacularly on avoiding financial institutions (through Nakamoto Consensus and the idea of a “blockchain”), it has left most wanting when being used as a substitute for cash. Sure, we can debate on what was intended by the term “cash,” but I’m going to assume that Satoshi meant money. Most agree that money should always have the following characteristics:
“Good” money should be (inherently):
Note: “Generally Accepted” is usually added to this list. While this makes good money in practice, we’re looking at the prerequisites to this step.
Let’s put bitcoin to this test. Fungible? Check. Portable? Check. Tradable? Check. Durable? Check. Valuable? Check. Stable? Nope.
Bitcoin predicted that over time its price would stabilize with use. So how come, after nearly a decade of consistent user growth, has bitcoin not stabilized? I believe that it is because of a fundamental flaw in its reasoning and design, which I’ve termed the “bitcoin fallacy.” The bitcoin fallacy is that the basis of money is scarcity, or more accurately a fixed (known) supply. I will further argue that credit is the basis of “good” money, and any alternative falls into the trap of the “value/stability trade-off.”
The value/stability trade-off is the concept that, in order for an asset to have even the pretense of value, it must be scarce. Scarcity can mean a lot of things, but mainly it implies that the supply of an asset is relatively predictable going forward. Value, on the other hand, deals with why an individual would purchase something. Often this is because the asset has utility or produces cash flows. Other times it’s because the asset is so scarce, like gold, that it becomes a schelling point for those in search of a safe haven of value, i.e. the refugees of “bad” money.
The idea that money has historically based on scarcity has been heavily criticized, perhaps most eloquently by David Graeber in his book Debt, in which he analyzes 5,000 years of monetary regimes. He concludes that unless one is living in tumultuous times where property rights are always at risk, money has arisen from credit. In-depth analysis and historical precedent aside, there is a simple logical progression one can follow to see why this is the case — the only reason that bitcoin is valuable is that it’s scarce. This proposition is what keeps investors coming to the table. Bitcoin is a great store of value. No serious investor is buying bitcoin because they think it will be an excellent way to buy coffee in a few years — they’re buying it in the hope to earn a return on their investment denominated in the currency (i.e., actual money) they’ve invested. This has been a dominant bootstrapping mechanism for bitcoin and has fueled an evangelist movement rarely seen in finance.
Bitcoin is also volatile because it’s scarce — you cannot have both stability and a fixed supply; they are like oil and water. Some recent entrants attempting to create a better digital cash, like Basis, believe that if you make the supply dynamic (can expand/contract as needed) but predictable you can escape the value/stability tradeoff. It should be fairly obvious why this will not work. Basis hopes to create stability by making its supply dynamic. The Basis system attempts to punish users for purchasing basecoins above $1 and reward them for doing so below $1 through incentivized supply manipulation — there are no assets backing basecoins besides, at most, a derivative of basecoins. It is backed by nothing. If the system functions properly, there’s no hope of making a return on your investment for buying basecoins. In doing so they’ve made a completely illogical tradeoff. Bitcoin is valuable even though nothing backs it because investors hope to make a profit and is, consequently, not stable. Basecoin will attempt stability but will have also no inherent value, and therefore even loses its ability to be stable because it’s worthless — unless you consider stability around 0 the intended consequence. Now let’s compare this to Dai, which uses credit to avoid the value/stability trade-off.
But first, why credit? It seems as arbitrary a thing with which to create money as any other scarce asset. But credit is not a commodity, nor does it automatically produce cash flow, nor is it useful in any particular application. Credit is much better than other assets that can be used to create money, because credit is money. Credit is an account of what we owe each other and owing each other is what makes money useful. At its core, regardless of what form it may take, money is an abstraction of credit. Let’s further employ this analysis to Dai.
Dai is valuable because it derives its value from pledged collateral, and its supply is dynamic because it’s created and destroyed based on loans made relative to that collateral. Dai supply does not expand as a reaction to price but in response to a demand for credit and a surplus of acceptable collateral. Its supply does not contract solely due to a lack of demand, but due to a scarcity of acceptable collateral. Which is nothing new; it’s how the global financial system has operated under fractional reserve banking for hundreds of years.
Dai: A Secured Loan with a Loan-To-Value Ratio < 1
Now that we’ve established how Dai is created/destroyed as part of a broader credit system let’s look into the specifics of how this functions. Dai is created by being borrowed against collateral. These loans will always have a Loan-To-Value (LTV) ratio of <1. In the Maker system, we quote this number inversely and called it a collateralization ratio (i.e., there must always be more than 100% of the loan in collateral). What this means is that for every one Dai that’s borrowed, a minimum of $1 in collateral must remain in the smart contract. The reason it is a minimum of $1 and not one Euro or one ounce of gold is simply because that’s the value to which we’ve set the price feed — this number can be anything. We chose the dollar because it’s the backbone of global finance and currently the best money we know of. The core concept is that users input a volatile asset and borrow a stable asset. The stable asset remains stable because other incentives in the system encourage arbitrageurs and debt holders (“Keepers” in Maker speak) to participate in moving the market price back to $1. These arbitrageurs have the confidence to participate because they know that Dai is intrinsically worth $1 because of the pledged collateral and that the Maker governors can force a redemption of Dai for the underlying collateral through a process called an emergency shutdown (formerly known as Global Settlement). An emergency shutdown is a mechanism that makes 1 Dai redeemable for $1 of the pool of underlying collateral at a given point in time, thus acting as a forced redemption which shifts volatility risk back to Dai holders as a measure of last resort.
When a user borrows Dai, they have to lock their collateral into a smart contract called a collateralized debt position (CDP). After they borrow Dai, they cannot fully unlock this collateral until they pay the loan back in full, with interest. The interest rate in Dai is called the Stability Fee (more on this later). Once the user has borrowed Dai they must maintain a pre-set minimum collateralization ratio (the inverse of the LTV). Should the value of the user’s collateral not meet this minimum at any point in time, the loan is considered to be in default and is “foreclosed.” When this happens the smart contract automatically liquidates the collateral for Dai up to the amount of the outstanding loan — it does this via an open auction. When Dai raised from the sale of collateral is equal to the amount of the outstanding loan, the auction closes and whatever remains of the collateral is returned to the borrower. This Dai is burned to offset the supply created on the origination of the loan.
Once again, this is nothing new. Maker is a digital reflection of its analog competitors, banks. What is new is boiling down the fragmented, complex, rent-seeking global credit system into a compact and straightforward layer of automated digital infrastructure. If you’re wondering why this hasn’t happened before, you can thank Mr. Nakamoto for figuring out how to exclude financial institutions from being a necessary intermediary in financial transactions. But more importantly, you can look to the trend of tokenization that is taking the world by storm.
As I stated before, Maker lets users take a volatile asset and borrow a stable asset. On a macro level, this is a pool of volatile assets all backing one stable asset, which is very important. Without an adequately diversified collateral portfolio to support the single fungible stable asset, Maker would be routinely brought to its knees by tail-events in the collateral portfolio that cause outstanding loans to be significantly underwater. In plain English, Maker is betting that blockchain-based tokens slowly begin to replace traditional certificates in the capital structure of organizations. In this light, ICOs, as we’ve seen them executed, can be considered equity financing for the 21st century — a programmable share. Maker intends to provide debt financing for the 21st century — programmable credit.
Maker: The Decentralized Central Bank
The above description of the Maker credit system pertains mostly to the micro level, that is, loans as they relate to individual borrowers. Expanding our vantage point, we can see that on a macro level Maker has much in common with a central bank.
Currently, in most banking/monetary regimes, money is “printed” by central banks which then rely on commercial banks to increase the supply of money through loans and investments. This is the foundation of modern fractional reserve banking. The central bank can then influence the outstanding supply of money indirectly through interest rate manipulation, and directly through reserve requirements and open market operations. As a consequence, these intermediaries can extract large rent by borrowing cheaply from their depositors and lending more expensively to the money creators (generally the sovereign). Fun fact: If you’re rich enough, banks will let you share in this higher interest rate — i.e., not screw you.
Another consequence of the current intermediated banking system is a systemic misalignment of incentives. Banks inevitably extend too much credit (see: the great financial crisis), and they should, in theory, be held accountable for underwriting these bad loans. As we all know, what happens in practice is that these institutions are too systemically important to permit failure. Pushing the onus of covering this bad debt to the central bank, which in turn has no backstop except its credibility, which it has posted as collateral in the form of fiat currency. Hence when banks behave poorly, everyone pays. It’s a perfect variant of the idiom “If you owe the bank $100 that’s your problem. If you owe the bank $100 million, that’s the bank’s problem.” If the banks issue enough lousy debt, it is you who will pay.
So what do these lessons you’ve heard ad nauseam for the past decade have to do with Maker? The Maker system makes several vital improvements. For one, it eliminates the possibility of just creating money from nothing. As already discussed, money in the Maker system is created by borrowing against existing assets. On a macro level, this creates a diversified pool of collateral which backs the currency, not the credibility of a central bank. Dai could eventually be pegged to CPI or a similar metric should the Federal Reserve lose credibility.
Maker also eliminates the intermediaries; loans come directly from the “central bank.” This removes the ability of intermediary rent-seeking and permits Maker to charge an interest rate that is far more in line with market demand and unmolested by government influence. While this is not too different from the current eurodollar market, Dai is not stuck in the purgatory of foreign banks because it is free from banks altogether. This direct lending also makes it very easy for the Maker governors to control the supply of Dai via interest rates and the collateralization ratio. While it would be naive to think that the Maker system is impervious to politics, as a popular movement could introduce a collateral type that’s intrinsically worthless and cause massive inflation, the incentives in the system are at the very least adequately aligned. In the Maker system, there is an additional class of participants called MKR holders (MKR is a token on the Ethereum blockchain) whose role it is to underwrite these loans. They benefit when they do this job well and suffer when they do not.
MKR: The Decentralized Insurance Company
The MKR token holders function as the governors of the Maker system. They decide on the specific collateral assets for the system, set the amount of Dai that can be borrowed against that collateral, determine the collateralization ratios of these assets, and ultimately charge an interest rate called the “stability fee.” The stability fee exists because the MKR holders, in addition to their governance responsibilities, are effectively the underwriters of every loan and the collateral buyer of last resort. In the steady state of the system, when loans are being paid back and not being liquidated, the MKR holders collect the stability fee through supply contraction (Dai collected as an interest rate is auctioned off for MKR and then burned). Even after liquidating a loan due to not meeting the collateralization ratio, there is usually excess collateral in the CDP to provide a cushion for the 1–1 peg. However, if the price movement in an asset is sharp enough to cause a shortfall in collateral relative to the amount of Dai borrowed against it, the MKR token supply is diluted and sold onto the open market to purchase Dai and offset this imbalance. In turn, it encourages the MKR holders to govern the system well and charge an appropriate interest rate for the loans they are insuring. If the Maker governors issue bad debt, it is the Maker governors who pay. Yes, if all value in the MKR token is also wiped out, then Dai holders would ultimately suffer, but there is no situation where MKR holders are bailed out at the expense of Dai holders.
A Global On-Chain Derivative Market
The Maker system inputs a variety of volatile assets and outputs a single stable asset, which is possible due to the smart contracts Maker has designed, called CDPs. There is a fascinating way to use CDPs outside of the core system that I believe will completely change the landscape of derivative markets; Maker currently has multiple partners working on systems like this.
It’s important to note that this additional system cannot exist without a stablecoin like Dai and cannot be commingled with the system outputting Dai, so it should be no mystery why this has not been done yet. This new system would function as follows: Invert the CDP mechanism, so the input is a single stable asset (Dai) and set the output (price feed) to a variety of volatile assets.
Let’s go through an example. If a user were to deposit 100 Dai into a CDP and wished to borrow “Apple Stock Tokens,” the operators of this system would simply set the price feed on this CDP to “Apple Stock.” Now, rather than being liquidated when the collateral in the CDP fluctuates too far relative to the loan outstanding (which is unstable), the user is liquidated when the loan outstanding fluctuates too far relative to the collateral in the CDP (which is stable). To draw on our example, let’s say Apple Stock is trading at $200 and you deposit 400 Dai into a CDP. If you borrow 1 Apple Stock Token, your collateralization ratio is 200%. Let’s say that we impose a minimum of a 150% collateralization ratio and Apple Stock rallies to $265. Dai in the CDP is now liquidated for Apple Stock Token to offset the supply. It is a simple but potentially impactful system. It allows people to create derivatives on virtually anything without intermediaries, clearinghouses, exchanges, and especially without geographic borders (cough, China).
The problem with money has never been the process by which it’s created. The problem has been the trust we place in the institutions that create money. Maker offers a solution to this problem without “reinventing the wheel.” Maker’s dreams are ambitious, but they are not without thought and preparation. We hope to level the playing field between financial juggernauts and the poor and marginalized. Our team is the best in the world and is hell-bent on making this dream a reality.
Note: The views expressed in this post are mine alone and in no way representative of those of the MakerDAO organization.