On DeFi and OpFi
Rose is a rose is a rose is a rose.
~Gertrude Stein, Sacred Emily
The Code is the Code is the Code is the
I want to preface this by saying I personally appreciate all of the experiments being conducted in the space. This is in no way an attack on the projects that will be mentioned, and this is all just opinion. It’s simply a bit of observation in order to put things into perspective. As always, this space is a semantic wasteland — so let’s jump right in.
A History of DeFi
De(Decentralized) + Fi(Finance) = DeFi
Great — now that we got that out of the way:
Decentralization is quite simple, as it means taking power or authority from a centralized force and distributing it to localized forces, or even the establishment of a federated model of equal power. In the context of blockchains, this comes in through:
- Censorship Resistance and Immutability, or the likelihood of an adversary to censor transactions
- Verifiability, or whether or not an individual can fully audit and run the software associated with the system that fully verifies the rules and transactions
- Accessibility, or if there are gatekeepers to transacting in a particular way
- and Social Consensus, which evaluates whether or not participants are actively aware of their own choices or are driven by those with a disproportionate amount of power.
As for the finance section, this is self-explanatory and typically covers traditional financial services including but not limited to general transacting, banking, lending, and insurance. But when have these two words been used together in the past?
What I have presented above is a Google Ngram search for “Decentralized Finance” in order to lay out the origin of the term. It seems that usage of it originated in the early 1900s, before the time of blockchains, and was typically in reference to general funding distribution. Even the Corn Belt Meat Producers’ Association, one of the many farm organizations in the United States in the early 1900s, were quite ahead of their time in thinking about “decentralized finance” as a phrase. In an excerpt from their 7th annual meeting in 1921, the term came up when discussing funding sources in the United States.
But humor aside, it seems that the first major outlining of “decentralized finance” in the context of blockchains came from Max Bronstein’s Mapping the Decentralized Financial System in which he outlines the pitfalls of centralized financial services such as censorship, risk, and unequal access, and talks about the benefits of “blockchain-based financial services.” These benefits include being transparent, trustless, permissionless, censorship-resistant, and programmable. This eventually led to “DeFi” as we know it, with Set Protocol announcing that they’ve established an alliance with Dharma, 0x, Abacus, and Wallet Connect. With it came a subreddit and telegram group to bootstrap discussions around the topic.
Since then, hundreds of ‘DeFi’ projects have been propped up either through grants, venture capital, or just plain old ambition.
What I will argue in this article is the idea that most of what has been built on top of blockchains has been Open Finance and not Decentralized Finance. Open finance leverages some of the same components as decentralized finance, but might have additional centralized components to ease the friction for end-users — and that’s completely okay. What’s not okay is if the marketing calls it something that is “decentralized.” A utopian future may hold decentralized products that have currently undeveloped mechanisms that allow them to function as well as centralized products — but that future isn’t here yet.
~A note on sustainability~
Something I won’t cover here is the sustainability model of a DeFi “project.” Since DeFi is meant to have the aforementioned characteristics, projects don’t have much of a sustainability model to rely on other than burning any existing capital. Taking a transaction fee would require a centralized form of intermediation (which consequently is one of the few ways to monetize), and would then place the protocol outside of the DeFi category. What would be lucrative in a case such as this would be to maintain the protocol as an open rail and build an open finance service on top in which the creator could act as an intermediary for providing a service with specific centralized provisions to abstract the experience for the end-user — jumping on the supply side of services around said DeFi protocol.
Investors will begin to care about this because speculation on everyone inventing their own money dries up after a certain point. Eventually, these teams will need to figure out how to achieve profitability without relying on their treasuries or good-will to remain alive.
Centralized Finance to Decentralized Finance: A Spectrum
Op(Open) + Fi(Finance) = OpFi
Great — now that we got that out of the way:
To avoid telling the same joke twice, “Open Finance” being used in a modern bank-y sense probably originated in the 90s with the advent of ‘online’ banking, and the eventual transition to digital systems. I like to think of OpFi as something more in the realm of ease of access to both financial institutions and services (non-discriminatory). Although centralized companies provide these services, the goal for blockchain-people is to eventually move to permissionless models in which any individual can access and utilize services that look similar in ease and feel to the ones in production today. A good example of this came from Bitcoin — how can we create a permissionless basic financial system for value transfer that can’t discriminate against anyone?
Although OpFi isn’t a new concept, strong-enough public chains provide open rails to enable OpFi applications to be built quickly and with enough security to be consumer-facing rather quickly. They may not be as decentralized as the base layer (you need that degree of neutrality to eventually have a ‘decentralized’ product), they strive to provide a great user experience and a permissionless tool for anyone to use.
I’m going to take the hard-liner stance and make the claim that any product that has components that enable the various traits associated with decentralized finance with any centralized core component is an example of OpFi rather than DeFi. Yes, we all are aware that blockchains are logically centralized around a particular piece of software — this is more in reference to services that abstract particular aspects of an experience for an end-user.
The first example to point to can be controlled contracts. Although permissionless and usable by anyone, if a contract is able to be paused or halted by any single party, it isn’t DeFi. It’s better classified as a custodial or communal contract, so long as the source code is open source and independently verifiable — that goes for anything OpFi.
Recently, a great example of this came from a critical vulnerability found in 0x’s v2.0 exchange contract which would allow attackers to “fill certain orders with invalid signatures.” The team subsequently shut down the contract (and related AssetProxy contracts using the AssetProxyOwner contract to prevent any forms of exploitation. The team then patched the contract and redeployed it to the Ethereum mainnet in a swift and efficient fashion.
While it was down, relayers relying on the contract including Radar Relay and Tokenlon were shut out of service. Obviously anyone could’ve been using their own version of the contract by simply forking it, but reliance on the 0x-deployed contract was another logically centralizing piece. They would’ve also been faced with the same bug and would also be responsible for patching it if it remained undetected to maintain network effects. However, the fact that 0x was able to detect, fix, and redeploy the contract is fantastic — but is a clear indicator of 0x being an OpFi project rather than DeFi.
Also supporting the notion of OpFi rather than DeFi are relayers in the 0x ecosystem. They typically leverage an exchange protocol such as 0x while providing centralized tools in order to abstract certain pieces from users. These tools come in the form of order books, great UI/UX features, and support teams ready to help any users that have issues on their applications. Some even have been dealing with compliance requirements such as Paradex, where certain users are blocked from using relayers based on their jurisdiction’s rules. However, the key to these relayers is that they are non-custodial, rendering them a bit more on the OpFi side than pure centralization. Again, your comfort with abstraction comes at the expense of your own personal security.
Another great example of where “OpFi” should be applied over “DeFi” comes from the lending space. Compound currently holds over $107.2M in locked assets. Although many in the space refer to Compound as ‘DeFi,’ (excuse the weasel word, just search “Compound” and “DeFi” in the same tweets on Twitter), they’re quite centralized and pull most of the strings when it comes to the product in question.
From the Open Zeppelin audit:
The Compound team currently administers all aspects of the protocol to decide which assets can be loaned, the interest rate model for each asset, and how the system obtains price feeds. They also control various economic parameters including collateral requirements and the size of the incentive used to encourage third parties to liquidate under-collateralized loans.
If you want a great read breaking down all the things the Compound team can currently do with administrative functions, check out Ameen’s recent article:
They aren’t the only lending system in the ‘DeFi realm’ to have centralized components. Maker still relies on a semi-centralized oracle system, and Dharma just recently paused their contracts to pivot to a new product (while also subsidizing rates).
One of the key components to these services is being “user-first rather than developer-first,” which Nadav Hollander said earlier last month. It’s completely fine if these protocols and projects do this — so long as users are aware of the risks involved and what controls the companies may have. Most of these projects have been relatively open about their direction, with usual directives to ‘decentralize over time.’ For example, Compound has been very clear that their goal for the protocol is to replace the current administration with a DAO, and that they are the current controllers of the admin address. Let’s just hope that it’s not just ‘decentralization theater’ to pacify participants and followers.
My point with all of this is to help guide the discussion and present an end-goal rather than falsely claiming to be at said end-goal.
~A note on the Lego analogy~
Have you ever played with legos as a child? Then you understand the analogy being passed around like wildfire that attempts to envisage the various ways in which these protocols and projects work together. The reason for the analogy is simple: Ethereum offers a lot of composability. We can take a few products built on Ethereum, arrange them in a few different ways (as they easily interoperate) to reach a particular end goal. The issue I have with this analogy is as follows:
Imagine building a lego set that had a nice green and sturdy base. Now start building vertically in a very straight line going up, continually connecting pieces along the way. This isn’t usually an issue no matter how high you build considering that those lego pieces are quite sturdy. Now I want you to imagine taking the tower apart in the middle, removing one piece, throwing it away, and reconnecting the tower. Nothing really happened other than one component going away and the rest of the tower being able to seamlessly attach to the underlying base quite easily. The problem with this idea is that there is no assumed link and that these things can endlessly connect. With most of these applications being some form of open finance, there is a direct liability with each new piece added on top of the last — if the underlying piece is destroyed or attacked, the structural integrity is lost.
Using the same process from earlier — instead of gently lifting the top, removing a piece, and reconnecting, I want you to take a Louisville Slugger to the middle of the tower and see what happens. That’s a bit more accurate. The easier and more accurate analogy would be a Jenga set: remove enough pieces and the tower comes crumbling down.
So, are there any DeFi projects?
I believe when evaluating the DeFi landscape, Uniswap and the original dutchX contract get as close as possible in terms of being hard to stop, all components being open source, and no critical coordinating force in day-to-day operations of the protocols. I won’t be too picky and start to name the developing teams as centralized actors, but I’ll give them credit where credit is due, as they’ve strived to keep things absolutely open.
In the case of ‘true’ DeFi, the code is the code is the code is the code. There are no surprises, no hidden backdoors, no administrative control, and no intermediary. It is what it is and you use it as you wish. Although someone may have to be on the other side of your activity (as is the case with any financial transaction), you both came to the table willingly, and that table is in an absolutely neutral environment.
…until someone attacks the meeting place — but that would’ve meant this article would’ve been “On Blockchains.”
