DeFi often looks like a venue for online gambling, with little innovation besides ambiguity of responsibility. In this piece, I argue that DeFi is actually the first step in a complete rebuilding of the global financial system.
Introduction
Software has eaten the world. Industries that previously ran on people and paper now run on computers. Many successful software companies have been built by bringing industries stuck in the 20th century to the internet age.
But some large industries have resisted this change–notably finance, education, and healthcare. It is not for lack of trying by imaginative entrepreneurs. Large changes to these industries are hard to come by, as they have significant direct government involvement, are heavily regulated, and have many large stakeholders with veto power over systemic changes.
Modern Finance
The fundamentals of modern finance were built over 100 years ago, when bank deposits and stock notes were literal pieces of paper. Since then, the system has evolved iteratively through patchwork, picking up institutional and technical debt along the way. The most successful fintech companies—PayPal, Stripe, Robinhood, Plaid, Square, Shopify—are clean interfaces to the gnarly depths of the traditional financial system.
Today we have T+2 settlement, large transaction fees and transaction times, limited trading hours, poor yields for consumers, and unsatisfactory asset custody and transfer solutions. These just don’t make sense in the internet age. If one were to build a financial system from the ground up today, it would look nothing like the current system.
How to Disrupt the Traditional Financial System
If the current system is so bad, how can we fix it?
In software engineering, systems often need to be changed: the spec has changed, flaws in the original design have been uncovered, the existing system can’t handle increased throughput, or new technologies unlock improvements. These systems then undergo refactoring or rewriting.
Refactoring involves taking the existing system and making incremental improvements to it. This makes sense when there is a lot worth preserving in the existing system, the changes to be made are minor, and the amount of complexity in the system is appropriate.
Rewriting involves throwing away the old code and rebuilding from scratch. This makes sense when fundamental organizational changes need to be made, or unnecessary complexity (technical debt) is overwhelming.
The traditional approach to disrupting the financial system looks roughly like starting a neo-bank as a wedge, then reforming the system from the inside. This is akin to refactoring the backend of the financial system, piece by piece.
Change created this way is unlikely to be significant. Each incremental advance runs up against powerful stakeholders resistant to change and a hundred years of institutional debt.
DeFi’s approach is to rewrite the backend of the financial system.
Why DeFi
DeFi is building internet-native financial primitives (asset custody, transfer, trade, borrow/lend) on decentralized infrastructure.
There are multiple difficulties in rebuilding the backend of the financial system:
Regulation makes it expensive and difficult to build anything at all.
The existing system is trusted; institutions interact with each other under well-defined rules and the rules are credibly enforced by regulators. Nobody will migrate assets to a system they don’t trust.
Even if a better backend is built, it’s hard to coordinate stakeholders on the old system to migrate.
The DeFi approach addresses all of these problems.
The Regulatory Problem
Because DeFi is built on decentralized infrastructure, it’s very difficult for regulators to stop on-chain financial innovation and experimentation. There is no entity that can take down an application, and builders themselves may be anonymous.
Nevertheless, regulators still have levers to pull if they want to slow the growth of DeFi:
Shut down fiat/crypto onramps via banks (attacks stablecoins and centralized exchanges).
Go after doxxed protocol developers.
There is an obvious consumer protection angle they can take here. Consumer protection that exists in traditional finance doesn’t exist in DeFi. And most retail users lose money in crypto, even though crypto is the best performing asset class of the last decade.
The Trust Problem
DeFi begins with no social credibility, so the immutability, verifiability, and uncensorability properties offered by blockchain infrastructure allow users to be convinced the system is secure. No inside or state actors can turn it off, and anybody can read the code and verify that it is correctly executed.
The Coordination Problem
Once a better system is built, how do we migrate off of the old system? Staying on the old system is a stable equilibrium, as the new system is worthless without users and liquidity.
DeFi is executing a decentralized vampire attack on traditional finance. High yields drive institutional demand. These institutions–corporations, banks, and sovereign entities–then dip their toes into the crypto waters, as allowed by their compliance and risk departments. Finally, interfaces that tap into traditional finance will have to support DeFi due to demand.
The Path Forward
DeFi is a viable path to rebuilding the backend of the financial system, but the product experience leaves a lot to be desired. Self-custody and immutability are required for rebuilding this backend, but are very undesirable properties in mass market products. What is built on top of and around DeFi does not need to be decentralized.
Most users–both retail and institutional–do not want to self-custody digital assets guarded by a private key. Instead, they will interact with DeFi through centralized, trusted services. Most usage will go through centralized performant alternate backends compatible with DeFi, similar to rollups on an L1.
Crucially, the existence of decentralized alternatives means centralized financial services companies will have to offer competitive retail products, such as market rate yields and overcollateralized borrowing.
Retail interfaces are very natural for consumer finance companies like Robinhood, Block, and Coinbase to build. These interfaces will provide custodial services, insurance, fraud protection, etc. They will look and feel similar to existing retail financial interfaces.
Institutional services will be compliant, likely involving some level of KYC and mandatory reporting to authorities.
Of course, the fully decentralized and uncensorable version of DeFi will always exist. But its position will be that of a hedge: if the services in the middle fail, users can fall back on decentralization.
Summary
The existing financial system is slow-moving and constantly stuck in local optima.
DeFi enables rebuilding the backend of the financial system.
This will bring more efficiency, access, and transparency to finance.
Centralized entities, both tradfi and crypto-native, will continue to dominate.
Tell me why I’m wrong on Twitter.
Thanks to @GaussianProcess, @hasufl, @inf0theory, and others for their helpful comments.