Developing Standards for Universal Market Access — Featuring Allison Lu of UMA Protocol🎙️
Kerman Kohli: [00:00:00]
Hey, all. Welcome to another episode of the DeFi Podcast. Today we have Allison Lu, the co-founder of UMA Protocol, here to tell us about financial primitives and why we need them. Alison, could you explain briefly what financial primitives are and how we can benefit from them?
Alison Lu: [00:00:19]
Thanks for having me on the podcast. So I think of financial primitives as kind of being like “money Legos.” They’re the individual components that can be put together to create a larger, more complex system.
So in theory, in blockchain, you can really almost think of the theory itself as being the first financial primitive that defines how you create custody and transfer a tokenized medium of exchange and then there are all these other projects that are now built on top of the theory and that define other primitives that can cover the whole range of financial services.
Everything from borrowing, lending to investments and insurance makes sense.
Kerman Kohli: [00:01:00]
And with these primitives, like what’s an example of how they might be used and why they’d be useful support?
Alison Lu: [00:01:09]
Yeah. So from a developer’s perspective, you don’t want to have to reinvent the wheel and build everything yourself. What’s really cool about DFI and watching primitives in particular is that you can just put these little Lego pieces together to create the final end user-facing product that you want from an end-user’s perspective.
Once somebody actually owns some crypto money, right now, there’s not a whole lot that you can do with it besides just hold onto it. But as we build more and more of these financial primitives, you can start doing more with your crypto assets again, using it as collateral to borrow so that take out a loan, for example, or to lend it out in order to earn yield or to instead of just owning, you know, the native cryptocurrency asset, being able to leverage the value of that cryptocurrency asset to get exposure to other things.
So it can really cover the whole range of financial services that a person might need.
Kerman Kohli: [00:02:15]
Sure. See, you have some like lending, borrowing, maybe synthetic assets. Is there anything else you’d include as part of like those primitives which could be useful for from an end-user perspective?
Alison Lu: [00:02:28]
I really believe that’s actually a really tough one to answer because one of the things that gets me most excited about what we’re building here in blockchain is the ability to make the cost of financial innovation basically zero. So the analogy that I like to think about here is when YouTube first invented nobody. The first thing that people did is they just took stuff from the regular world and then they put it onto YouTube.
Alison Lu: [00:02:53]
So they like took their Pirates of the Caribbean movie and uploaded it into not a great use case. Not super interesting.
Alison Lu: [00:03:01]
But then over time it evolved. And suddenly now some of the most popular people on YouTube do things like unpacking presents, you know, or destroying their video game players like completely unheard of, unthought of forms of content got generated because suddenly the cost of generating and distributing content was basically costless. So the same thing I think can happen with financial innovation.
Alison Lu: [00:03:25]
And so I don’t know what the equivalent of hacking Jeff is going to be in financial services, but I’m excited.
Kerman Kohli: [00:03:35]
Perfect. And how did you may or you may help with creating these primitives, like on a more granular level?
Alison Lu: [00:03:45]
So you might have heard of the phrase programmable money. Basically, we’re creating a specific type of program, one that makes it really easy for you to take the money you have. Like if you have either a theorem and you can use that money to create long and short risk on something else in the form of a token.
So just going through an example, let’s say Alice wants long exposure to gold, meaning she wants to make money if the price of gold goes up. And Bob wants the opposite. So he wants to make money off. The price of gold goes down. He’ll lose money if the price of gold goes up.
What our protocol can do is it can make it so that both of them can have this need met. And neither of them have to touch physical gold. We simply define a protocol so that as the price of gold moves. Alice and Bob each have claim to be more or less money from a common pot.
So there’s really two primitives actually here that are required for this to work. The first is a primitive that defines how much each participant’s claim is worth relative to some reference asset like gold. And the second is an oracle, which is the protocol for getting that reference asset price on chain securely.
And all of this stuff that you know, that the thing that’s really cool about Defi is we’re building on top of the work of others. And others can build on top of our work or extend it as well. So, for example, you could use a stable coin like DI as the digital money in this common part between us and Bob.
You can tokenize Alice and Bob’s positions and you can trade it out on a deck or you can lend it out on compound. You can reference any price index really that you can think of off-chain assets like gold cross-chain assets like bitcoin or stuff that’s completely non-tradable, like the amount of rainfall in London.
Allison Lu: [00:05:35]
So it’s it’s really a primitive because the illusion of financial products can really explode out of this.
Kerman Kohli: [00:05:48]
Perfect. And what are some of approaches to creating financial print, as you see in the industry that differ to how you guys approach the problem?
Allison Lu: [00:05:56]
Sure. So as I mentioned before, we’re really kind of operating in two primitive spaces. One is defining synthetic markets and the other is oracles. So our whole system design is the only one that we seen which combines both a synthetic contract that allows top-ups and an Oracle design that’s based on a selling point game just in the interest of time.
Maybe I’ll discuss the Oracle first. And I think that that that design is a bit more unique basically in order for you to get information securely on-chain.
Well, getting information securely on-chain is, first of all, very difficult and then using it inside of a smart contract is actually really dangerous because your smart contract is simply a bunch of logic. So if you tell the smart contracts that the sky is purple, the price of bitcoin is zero.
Kerman Kohli: [00:06:52]
Know anything like the smart contract doesn’t actually know, right?
Allison Lu: [00:06:54]
It does what it’s told. And so what we’ve designed is actually an oracle that comes with an economic guarantee, a guarantee that if you have a smart contract with a bunch of money inside of it, that’s depending on this price, this oracle price and the costs to corrupt our oracle is actually going to be greater than any possible money that you can make or lose by using our oracle.
So that’s really a unique factor that we think makes or cling using external data a lot safer for smart contracts on four perfect days.
Kerman Kohli: [00:07:30]
That’s a really good answer. And I guess all we talk about financial parameters and going want to talk another. The obvious question, it really comes up, is it what is the risk with these primitives and how they make the compound in future or even just now?
Allison Lu: [00:07:49]
I mean, there’s just basic smart contract risk. And then, of course, compelling smart contract risk upon smart contract risk is another thing.
Allison Lu: [00:07:55]
I think the risk of that comes from maybe a lack of knowledge and time tested and time tested methods. So the first one is addressable or both are really addressable with time. So when it comes to like a lack of knowledge.
Allison Lu: [00:08:17]
The reason why I think it’s really addressable is because all the stuff that we’re doing is generally out in the open. It’s generally open-source.
So sure, as a brand new user, you may not know if this product or that product is safe or not, but you can look at the reviews of the people before you.
People who may actually know how to read and audit code and get their take on what’s happening in this code, how it’s. Recruiting and how safe it is.
Allison Lu: [00:08:45]
On the other side, in terms of like time-tested libraries and things like that, I mean, a lot of this stuff is kind of fresh cryptography right out of the research labs.
And it’ll take only take time to get going to be able to tell how robust some of these methods are at maybe a more basic level just to be more practical for maybe the listener out there today. Finding one of the things that I find risky about using primitives is just the risk that people either create things or parameters, things that are really silly.
So for example, if you’re familiar with Maker, MakerDAO, it’s a system for borrowing and lending. You can take your Ether as collateral and borrow a stable coin called Dai.
There’s a bunch of parameters that got decided when they watch the system and they can be updated. But you know, these parameters define like how much you’re allowed to borrow, how much each individual is allowed to borrow along with how much the entire system is allowed to borrow and what the penalties are if you don’t obey by the terms of the system.
And those parameters, you know, they actually chose very reasonable parameters, I think, but had they chosen a different set of parameters than you might see a lot more users who are underwater today or people who lost money in some fashion or the whole system might not have worked out really well had they chosen or chosen core parameters.
And I think that’s a risk as we watch some of these systems that the first set of parameters that we launch with that any individual system watch launches with might not it might not work well empirically.
Kerman Kohli: [00:10:31]
Sure. And maybe like just on a quick side note, like what are the regulatory issues that different local jurisdictions have, like using these parameters which may or may not be legal based on where the end user is at?
Kerman Kohli: [00:10:44]
Is there a risk over there which people might need to be mindful of, or is it still too early to say?
Allison Lu: [00:10:52]
Yeah, I mean, there is no international regulatory framework for any of this stuff. So it’s it’s challenging and I just generally always say like kind of fire. Be aware of what’s happening in your local jurisdiction.
But I guess since we tend to operate kind of more at the protocol level, I would say that the activity that we’re trying to do right now is similar, is similar to the activities that you know of like building batteries. And so we’re like BitTorrent, we’re not like Pirate Bay. And you as the end-user would be wise to use the core technology in a way that local jurisdiction.
Kerman Kohli: [00:11:39]
Sure. Okay, cool. That’s fine. I guess possible. Well, why did you choose to solve this problem in particular?
Allison Lu: [00:11:49]
The first thing that I co-founder and I talked about when we got together to start working on this is what the mission of it was and it really is our mission is to bring universal and fair market access to the world. It’s right in our name. UMA stands for Universal Market Access.
But beyond that, I think that both my co-founder and I had Bakra a very strong background, temperament and interest in what we’re doing. For me in particular, I started my career in institutional finance. I used to manage a $30 million book at Goldman as a trader, and after that I ran the credit desk at a digital lender that operates in third world countries.
Allison Lu: [00:12:29]
And so whether I was dealing with central banks or micro-entrepreneurs, my old life was always in the financial system. And in my work there, I saw how deeply unaligned incentives were between different stakeholders in the financial system.
I just got obsessed with the idea of being able to embed economic incentives into a technology to hopefully better coordinate and align the incentives of everybody using a system. So that value actually ends up accruing to the users rather than just being extracted for them from them.
Kerman Kohli: [00:13:09]
Perfect. It’s great. No character has super-smart people on our side from the old system. And I guess, is there anything you’d like to tell or mention out there, which maybe I haven’t covered before we kind of wrap up.
Allison Lu: [00:13:25]
I guess for anybody listening who’s kind of interested in the work that we’re doing and either wants to learn more or follow along, you can follow us on Twitter at UMAProtocol. We’re going to have some exciting news towards the end the year with our main that launch of our Oracle and some really interesting concepts coming out on the net shortly after that.
Kerman Kohli: [00:13:45]
Perfect. Thanks a lot for your time today, Allison. Look forward to seeing you out on me.
Allison Lu: [00:13:52]