Why it is Difficult to Decentralize Finance
Author Byrne Hobart | February 6th, 2023
The dream of cryptocurrency is to remove all overhead from the financial system—expensive buildings full of highly paid people making ad hoc decisions with all-too-human fallibility—and replace it with zero-marginal-cost software that anyone can interact with and that works securely and transparently without the need for any third party to trust.
Until now it has not been so.
In a sense, financial systems are already naturally decentralized. The general structure in capitalist countries is that there are households and companies, which interact with the banking system and other financial intermediaries, and these intermediaries, especially banks, are backed by a central bank. But there are many layers within that system; For example, a single transaction may involve a payment processing software provider, the issuing bank, the merchant's bank, and the credit card network. Each of these roles has its established actors, but in principle it is possible for a new actor to enter at any time. A lending relationship like a mortgage is a mess of similar complexity: there's someone who gets paid to deliver a potential customer to a lender, then the lender typically sells that mortgage to some long-term holder, while each mortgage payment generates revenue for the mortgage servicer. And the holder of that mortgage can pay more fees or overhead to cover their mortgage portfolio against interest rate fluctuations.
Unfortunately, all these intermediaries have the function of a) interposing between the person who spends the money and the person who receives it; and, b) charge a commission for the privilege of doing so. Obviously, in a specific case, it would be more profitable for the client to transfer the funds directly and save a lot of commissions.
On the other hand, the fact that these systems with a large number of intermediaries are gaining market share against their privacy-first competition indicates that the intermediaries provide some value.
A quick look at a fairly common form of decentralized finance ("DeFi") can show why this is so. Let's take a DeFi protocol that, like many DeFi protocols, exists to lend money and pay interest to investors. It's a bank, without all the high paid bankers who never seem to get into legal trouble even though they cause financial trouble!
The first problem a decentralized protocol will run into is that it needs some kind of "oracle" to measure the value of the collateral. If a decentralized protocol makes loans to small businesses, or to people buying homes, the likely borrower is someone who was turned down by a more conventional lender. Underwriting a business means trying to understand how it works, a process for which there are few standardized rules. Lending against real estate means knowing something about the value of that property, and because real estate borrowers are often fairly leveraged, the margin for error is small. But there is one category of lending where the oracle problem is easily solved: lending on margin against some commodity with a frequently quoted price and reasonable trading volume. In the case of margined loans, the value of the collateral can be measured based on the current market price, and the risk of that collateral deteriorating can be estimated by looking at the volatility of the asset.
Providing margin is economically equivalent to writing an out-of-the-money put option with a strike price at which settlement begins. Normally, it provides a steady stream of income, but when prices move violently enough, the lender loses. And the faster prices move, the faster the lender's exposure increases: If you lend someone $5,000 in exchange for a $10,000 asset, it's easy to ignore when the value of the collateral falls to $9,999, but each drop in $1 below $5,000 means a loss of $1 of the lender's principal.
Worse yet, automatic sources of liquidity change the way assets move. If there is an incremental provider of leverage, that provider will a) continue to increase assets, and b) create an ever-increasing air pocket between the price at which selloffs become a prominent concern and where prices end up when they are over.
Let's say you identify this problem and choose a solution: You'll create a DeFi liquidity provider that accepts lower returns but only supports the safest loans. Letting clients borrow 95% of the purchase price is extremely risky, but letting them borrow 20%, even on volatile assets, is quite safe. There are two problems here: first, such a DeFi product will not be able to raise much money, since the fringe lender is not doing very sophisticated analysis; At best, you're trying to filter out obvious Ponzi schemes, and at worst, you're just sorting by current performance.
Meanwhile, as these products grow, the classic Minsky dynamic takes over: readily available credit raises the price of borrowable assets, and lowers their volatility because someone can always get more leverage to buy the dip.
At any given time, the most irresponsible borrower and the most irresponsible lender are the most likely to win, the borrower because his assets continue to grow with available credit, and the lender because he can always deploy more money at a rate he considers satisfying. But if the market is increasingly defined by the dumbest player, then market prices will eventually become dumb. And then a small disturbance will be enough to start a cascade of liquidations.
A marginal lending algorithm can be built based on historical evidence, but what it cannot test is the change in market structure caused by its very existence. Of course, this is not unique to decentralized finance. That's a good description of what happened in 1987. Some smart academics discovered that an investor could replicate an options position using futures—as the market goes down, selling more puts replicates the position a market maker would have. of options to hedge a put, but in this case the market maker does not have to receive any premium for writing the option in the first place. This strategy became so popular that when the market faced a big crash, it triggered a flurry of mostly automated sales. The stock market plunged that day, and the Standard & Poor's 500 Index fell 20.5%. Futures tumbled further, trading at a 15% discount to the underlying shares. [3] The portfolio insurance backtest did not cover a period in which portfolio insurance existed, so it underestimated both the probabilities of a stock market crash and the probabilities of futures crashing further, ruining the picture. coverage.
Automated market making, as DeFi proposes, has the same problem. It is easy to create an automated market making strategy, but this strategy is actually a bet against volatility. [4] In normal times, a decentralized market maker will lurch around, producing constant profits from the bid-ask spread. And from time to time, there will be a big sell-off or short-covering burst, and the market maker, by design, will automatically be in the wrong position.
Automated payments can work—right now Bitcoin can be used to send arbitrary values to arbitrary people. But it works for a specific payment niche, with irreversible transactions and no built-in compliance.
This is a significant niche, both because international payments can be unusually cumbersome and because there is a population that wants assets that are technically difficult to seize: some of these individuals are criminals; others are paranoid, libertarian, or both; others are nation states that do not always expect to be on good terms with the United States. The possibility of getting out of the dollar system, and then back in if you behave well or change the rules, has a non-zero value. And with the dollar being the most user-friendly aspect of sanctions enforcement, the accessible market for decentralized payments and decentralized stores of value is expanding over time. (No, bitcoin is not a stable store of value, in the sense that its price fluctuates all the time. On the other hand, Russian dollar-denominated assets experienced more volatility this year than bitcoin, in the sense that their value for Russia suddenly went straight to zero).
Even if you were to create a functional decentralized system, you would face another hurdle—decentralized systems are not easy to discover. One of the first worthwhile businesses to be built on the Internet was Yahoo, aptly nicknamed Another Hierarchical Unofficial Oracle. If you start with a decentralized, permissionless system that anyone can use to create whatever content they want, the first thing the market demands is some centralized authority to make sense of the mess. Bittorrent is a decentralized protocol, but it has numerous centralized nodes that make it work, including clients and indexing sites like The Pirate Bay. Cryptocurrencies, likewise, have centralized on-ramps and off-ramps that necessarily link them to the legacy financial systems they interact with, such as Coinbase.
Subsets of finance, such as transactions, can be decentralized quite well. But the closer we get to providing credit—and, as it turns out, most forms of trade involve some extension of credit—the more we're playing a game of adverse selection. Playing these games is very difficult for players who don't realize what they are doing.
So generally speaking, DeFi will remain a pipe dream unless that underwriting issue is properly resolved. But finding the specific layers in which decentralization can occur is still a worthwhile project. We may be looking at a more centaur-like financial system, where there is high interoperability between automated components without a single owner, but also numerous chokepoints controlled by individual people who can make decisions, update models and heuristics. , and make the system readable to law enforcement and courts.
Perhaps the best way to look at DeFi is that it is a way to reach a new local high. The financial system has been evolving for a long time, from totally manual processes to mostly automated ones that keep starting with writing a computer program to do something that a human being has been doing all along. This describes high-frequency trading, of course, but it also describes all the financial innovations dating back to the development of double-entry bookkeeping. The DeFi approach starts with full automation, but that means that if it evolves, it will evolve in the direction of finding the minimum necessary level of human intervention, rather than starting with maximum human intervention and finding the minimum acceptable level of automation.