"All the Dizzying Ways People are Farming Yield is Proof of DeFi Composability's Success:" Ampleforth CTO

Hello Defiers! In this week’s interview, I spoke with Brandon Iles, the CTO of Ampleforth. Brandon is a former Google and Uber software engineer, and built his first crypto app together with Coinbase CEO Brian Armstrong. Today, he’s trying to rebuild the way money works.

For its huge ambitions, Ampleforth is a relatively little-known project. It wants to create an entirely new category of monetary assets — one that doesn’t have Bitcoin’s gold-like fixed supply, it doesn’t have Ether’s oil-like utility, it’s not fiat-backed like USDC, and it’s not collateralized like Dai. So what is it? It’s a token — AMPL— whose value comes from having a price which moves in a predetermined band, thanks to a smart contract which automatically contracts or expands its supply. The end result is a token that trades at close to $1, the 2019 US dollar, to be precise, but Ampleforth doesn’t like to be called a stablecoin —read on to find out why :)

We also talked about the project’s recent liquidity mining program called Geyser, where traders can earn AMPL by supplying liquidity on Uniswap, and how for Brandon, the dizzying activity around these schemes is proof of DeFi’s success. Brandon also discussed the project’s roadmap, from being used as a way to diversify collateral in lending platforms like Maker and Compound, to denominating debt obligations, to very far out into the future, replacing central bank issued currencies.

You’re a paid subscriber, which means you get the full transcript below. Subscribers also get exclusive access to The Defiant’s Discord chat for the community, here’s a new link to join.

🎙Listen to the interview in this week’s podcast episode here:

💗 Consider supporting quality DeFi journalism by contributing to The Defiant’s Gitcoin Grant! You can make a big difference with not that much; 1 Dai is now being matched by 138 Dai. Thanks so* much to my contributors so far!

🙌 Together with Quantstamp, a leading blockchain security firm keeping your money legos safe, Kyber Network, the on-chain liquidity protocol for the tokenized world, and Keycard, the secure, contactless hardwallet & open source API.

Brandon Iles: My background is very much in technology as a pure science in school, worked in research labs after that. I happened on crypto almost by chance, so I first encountered the Bitcoin white paper through Google Search, completely out of context with no prompt or anything like that, it was in 2009. And so, I read it and back then, you know, I thought it was interesting. But with no context, I thought looking at the various aspects of it, like scalability and that sort of thing, it seemed like an interesting toy, more than anything else.

Camila Russo: In 2009, it must have been hard to see it as something else.

BI: Oh, for sure. Yeah, I mean, it was basically a PDF format, it looks sort of like an academic paper, but didn't have any university markings on there, had this funny name on top. And so, that’s pretty much all I had at the time. I sort of forgot about it from that point forward, you know, it’s around the time I was going to work at Google. Move on to California, started working there. And then a couple years later, late 2010 or 2011, a friend of mine from college sent me an email and said, hey, have you heard of this thing called Bitcoin? We should play around with it. And so, together, we built one of the first Android wallet applications to be released on the Android store. That got released in July 2011.

CR: And that was related with Bitcoin or no?

Brandon: Yeah, so that application was called Bitcoin Android. That was just a straight up Bitcoin wallet using the current Bitcoin J library. And so, the friend I build that with was Brian Armstrong, who has continued to go in crypto…

CR: With Coinbase.

BI: With Coinbase. Exactly. At that time, Google was a really exciting place, because that was very much the hotbed of machine learning and artificial intelligence. That was where deep learning was starting to explode. So, I was getting very deep into that area of technology, machine learning, worked in the machine intelligence group, in search ranking group. I got the opportunity to meet all kinds of incredibly smart people and do lots of interesting work.

I stayed in machine learning for a number of years, I went to Uber for a couple years. And then with the release of Ethereum was where I started to get excited with crypto again. So, I thought, okay, this is actually an interesting, a generalized platform we can work with and one that we could potentially solve some of the earlier problems that I didn't quite like so much about Bitcoin back in the day, the elasticity and fixed supply.

And so around that time, I started talking a lot with another friend of mine, Evan Kuo, the other founder of Ampleforth, and we were both trying to figure out what was going on in the space. Because this was when there was a lot going on in the blockchain world with Ethereum, and with the whole ICO craze that was coming and going. And we were just trying to figure it out, because none of it really seemed to make much sense to us.

Brandon Iles. Image source: Medium

CR: Was this 2017?

BI: This was, let me see, summer of 2017.

CR: And where did you meet with Evan?

BI: Him and I had already known each other. So, we met through mutual friends a couple years earlier and we found that we had a mutual interest in mathematics and solving different kinds of puzzles. And so, we often get together and just talk about different puzzles we encountered.

Basecoin

And together we try to collectively understand this great puzzle that was going on in the blockchain world. We happened on the Basecoin paper. And so, when I was reading that, that's when things really clicked, that okay, we can actually design new monetary assets here with goals that we create ourselves.

“We happened on the Basecoin paper. And so, when I was reading that, that's when things really clicked, that okay, we can actually design new monetary assets here with goals that we create ourselves.”

CR: Just to give them a little context or people who might not know about Basecoin, but it was a stable coin that was supposed to be algorithmically stabilized, right?

BI: Pretty much, yes. So, the intent there was to create a stablecoin. It had a three coin systems. There was the stable coin and there was bonds that you'd be issued to help the market bail out the system if it ever became underwater and there were shares. This was a big deal back in the day, because they raised a lot of money to make this happen from very smart organizations. Ultimately, they ran into a number of problems and decided to close the project and return the funds to investors.

It was an exciting project and people still talk about it today, because I think it's easy to understand. So reading that got both of us really excited about the possibilities in the space. We initially started with a project that we called Fragments. So, Fragments was fairly complicated, much more complicated than Ampleforth. And it also had multiple coins, had a bond market in there and we had a reserve. But over time, as we worked with the system more and more, we started to lose faith in a lot of the mechanisms that we were sort of piling on, probably it seems increasingly fragile and risky. It had a lot of dependencies on markets that we didn't like and that sort of thing. We iteratively started taking these pieces out one by one and eventually, we ended up with a system, Ampleforth, that is so simple and sort of straightforward. We decided this just deserves its own name and so we came up with the Ampleforth name instead of the other name.

CR: Okay, perfect. And by this time was the ICO 2017 bubble over? So, because I see like a lot of building happens in that kind of 2018 bear market.

BI: Exactly, yeah. So, by the time we officially started, so we first started exploring ideas in the summer of 2017. But it wasn't until the end of that year and then we officially started in January 2018. So, we were definitely in the territory at that point.

CR: Yeah. And was it just you and Evan or had you grown the team by this point?

BI: I mean, at that point, it was just me and Evan, and then Jessica was also helping with some of the business aspects of it. And so, we started with three of us, two of us being technical and then we hired a first engineer. Eventually, some other people I worked with at Google and Uber also came over to join us.

CR: Nice. And did you have fundraising from VCs or were you bootstrapping?

BI: We did. We went the traditional VC route. We raised a seed fund from some very high quality folks like True Ventures, Founders Collective, and Pantera. We did think an ICO, that was definitely an option to us. But given the choice between getting just funds or funds with the advice of very smart, respected people, I think we chose a second one, which is the traditional VC route, which thankfully we had access to just given location where we were. I think that was definitely the right move.

Monetary Asset

CR: Now, can you get into how Ampleforth works?

BI: Sure, yeah. Ampleforth is the name of the protocol and then the token is called AMPL, pronounced “ample.” AMPL is a monetary asset like Bitcoin, but it's meant to solve the problems that come along with a fixed supply asset. So, Bitcoin as you know has a maximum supply of 21 million. It’s somewhat analogous to gold. But the demand for holding currency fluctuates, it goes up and down naturally with trade cycles, with booms and busts, with technological advancements. And so, when you have a fixed supply asset like that, a change in demand gets expressed entirely in the change of the price of that currency. We've seen this play out historically with Bitcoin, the prices has been incredibly volatile, going as high as $20,000 to as low as $3,000 and then back up to somewhere in the middle. So, it's not that surprising that we've seen that behavior.

“Ampleforth is the name of the protocol and then the token is called AMPL, pronounced “ample.” AMPL is a monetary asset like Bitcoin, but it's meant to solve the problems that come along with a fixed supply asset.”

So, when the price of that currency changes, that also means that it gets manifested in the changes of prices and things denominated in that currency, so goods and services, but also, most importantly, contracts and debt. If I were to take a loan from my friend for one Bitcoin to be paid back a year from now, I don't know exactly how much in terms of raw purchasing power I'm actually going to have to pay back at that point. I maybe on the hook for potentially twice or more.

This can lead to defaults, which is not good, especially when you have various pieces stacked on top of each other, which we’ve absolutely seen today in DeFi. One of the greatest things about DeFi is the composability aspect and sort of the Lego blocks you can piece together and the stack seems to be getting precariously high in the last few weeks or so. A default at the bottom layer can cause cascading failures all up the stack and reverberate through many other areas, which is not a healthy thing to do.

Image source: ampleforth.org

Systemic Risk

So, there are some great properties of fixed supply, like Bitcoin has and that's one of the things that people like about it. It's non-dilutive. It can't be inflated away by someone else who's outside of your control. So far, it's nice. But it's not able to support a complex economy on top, financial instruments on top. There's too much exposure to risk of cascading failures in what you might call systemic risk.

“(Bitcoin) is not able to support a complex economy on top, financial instruments on top. There's too much exposure to risk of cascading failures in what you might call systemic risk.”

CR: What about like compared with Ether and which is, I mean, if we're talking about DeFi, I think maybe Ether is kind of a good comparison to make. So, Ether, it doesn't have a fixed supply, but it's still really volatile like Bitcoin. So, how would Ampleforth compare with ETH?

BI: So, ETH is a different sort of asset with a different sort of monetary policy. So, on Eth1, it has a sort of linear inflation schedule, so it's rules based, which is good. There's no person or group behind the curtain, pulling levers to make things go up or down, that sort of thing. So, it's predictable. But there's no feedback cycle between the needs of the economy and the issuance of the currency. And then there's also one other difference that I think is important between ETH and BTC. Whereas ETH has use value, whereas BTC does not.

CR: What do you mean that?

BI: So, ETH, you can use to pay for gas to run computations on the platform. BTC is used to be BTC and nothing else. BTC has no real use value, the same way that something like oil would. So, oil, you can use for industrial applications and running your car. When you're talking about purely monetary assets, a use value, in some ways is a liability, because an outside source of demand that can impact the price of the currency. And so, in one hand, it gives it some grounding of value; on the other hand, it's another source of volatility that maybe you want in a pure currency. So, the US dollar, for example, also has no use value, which is kind of a nice property.

“When you're talking about purely monetary assets, a use value (like paying for gas with ETH), in some ways is a liability, because an outside source of demand that can impact the price of the currency.”

Not Pegged, Not Collateralized

And so, Bitcoin, I would say is defined primarily by its monetary policy’s fixed supply. ETH, I think is defined primarily, and people might disagree with this, but I think ETH is defined primarily by its ability to be used to pay for computation. We could still change monetary policy at ETH, and it will be ETH. But if you change anything on that with BTC, it would no longer be BTC.

So, I think they are somewhat different classes of assets, but they do have some similarities. Ampleforth is really closer to BTC in the sense that there's no use value and so there's no outside source of volatility.

CR: Okay. And so how is it designed to be a stablecoin? I understand it's not supposed to be pegged to the US dollar, but does it have any sort of functionality that makes it not be as volatile as other cryptocurrencies?

BI: Yeah, well, it's more a question of how that volatility is expressed, I guess in Ampleforth. So, the AMPL token has what we call a long run price circuit. So, we say that it’s not pegged per se, because there's no underlying redeemability for the US dollar the way there is for Tether or TrueUSD or anything like that. AMPL is non-collateralized the same way that Bitcoin or Ethereum is non-collateralized.

“We say AMPL is not pegged per se, because there's no underlying redeemability for the US dollar the way there is for Tether or TrueUSD. AMPL is non-collateralized the same way that Bitcoin or Ethereum is non-collateralized.”

When the price of the AMPL token in the marketplace is higher than the price target, then it means we need to increase the overall supply of AMPL to help bring the price back down to the target. And then when the price of the AMPL token is below the target price, then it means we need to reduce the supply of AMPL token to bring the price back up.

So, those are the two core rules. If we're above the price, we expand; if we're below the price, we contract. Now, the way the protocol does this is rather unique; in the sense that when there's a supply adjustment, it happens universally to all parties in the system. It happens proportionally to what you own. So, what this means is that the number of tokens you own in your wallet can change day to day.

CR: The number or the value?

BI: Both, right. The protocol adjusts the supply tokens, but its actors in the marketplace incorporate the supply back into price. So, this is the feedback loop. In a perfect world, if we could control the price of something in the marketplace, that'd be great. But you know, history has shown when governors try to ignore market forces, bad things happen. So, all we can do is let the market play out, but make supply adjustments to help it reach back into the equilibrium point where one token has the price target.

“The protocol adjusts the supply tokens, but its actors in the marketplace incorporate the supply back into price.”

CR: Okay. And so, how is this supply manipulated? Is it a smart contract that automatically kind of detects when a level has been reached and issues more AMPL or buys more AMPL from the market? Like how is it technically done?

Built on Ethereum

BI: Exactly, yeah. So, right now, Ampleforth is put onto the Ethereum blockchain. We made a decision early in the project not to build our own blockchain. But instead of creating our own silo, we preferred from the very beginning to be in the ecosystem where the most number of people are making transactions, and storing value. So, instead of building our own blockchain, we decided to deploy onto another blockchain Ethereum. Potentially, we could be also on other blockchains in the future as a sort of cross chain asset.

But the rules, the protocol, as you said, are deployed and codified into a smart contract on chain. And so, the idea is that the system is rules-based completely, with no room for discretionary policy.

And so, there’re three main components to the system. There's the monetary policy that has the rules about how to react to the marketplace and adjust the supply. There's the token itself, which is an ERC20 token that allows you to make transactions. Then there are the oracles that take the market data and then also the CPI index, the Consumer Price Index, it feeds that into the monetary policy, so that it’s able to make these decisions.

CR: What oracle are you using?

BI: So, we have two oracles. The first one provides the 24-hour volume weighted average price of AMPL versus the US dollar. The second oracle provides the PCE, the Personal Consumption Expenditures. So, this is a measure of consumer price index that’s published by the Bureau of Economic Analysis. And what this does is it allows us to target the 2019 US dollar. So, AMPL doesn't even target the US dollar, it targets the US dollar at a particular point in time...

CR: Okay. So, that's why you mentioned that CPI was another component. Why did you decide to do that, target the US dollar only in 2019?

BI: We wanted the value of AMPL to remain constant throughout at all time, is the idea. So, the value of the dollar naturally depreciates through the tools available of the central banking system. As long as this inflation is moderate, then it's no problem at all. But it does change the value of the US dollar over the years. One Ample targets about $1.01, right? Five years from now, one Ample might be $1.07. So, the idea is that we have a consistent value and unit of account through time.

Not Fans of the Word Stablecoin

CR: That's an interesting difference with most stablecoins, which reflect the value of the dollar. But in crypto many people are in this space because they disagree with central banks inflating their currencies. So that's always been a little bit of a disconnect, I think between crypto and the huge popularity of stablecoins still tracking the US dollar which is what people are supposed to be getting away from in the first place.

BI: Yeah, so back around 2015 or so, stablecoin meant something very different than it seems to now. It sort of represented the holy grail of crypto, the cryptocurrency with a predictable purchasing power.

But then, over time, as the needs of traders became very important for the development ecosystem, these exchanges needed some representation of US dollar to trade against, so stable coins became to mean a representation of the dollar on the blockchain. We were not a huge fan of the word stablecoin. I think, it's actually kind of a misguided framing. I like to ask, is the US dollar a stablecoin? And people don't really know how to answer that precisely because it is what stablecoins target. But is it a stablecoin? It's not really a stablecoin. It gives it stability through denomination and usage through pricing, denominating wages and contracts and that sort of thing, which gives it some nominal rigidity. And so, it's those market dynamics that give the US dollar most of its stability. So, the circulating supply of US dollar can change, but it takes a while until it actually propagates through the prices.

“We were not a huge fan of the word stablecoin. I think, it's actually kind of a misguided framing (…) back around 2015 or so, stablecoin represented the holy grail of crypto, the cryptocurrency with a predictable purchasing power. But then as the needs of traders became very important for the development ecosystem stablecoins became to mean a representation of the dollar on the blockchain.”

Necessary Volatility

CR: Right. And then I was looking at AMPL’s price chart on CoinGecko, I don't know if that's the best place to look at. But I was noticing that it more or less maintains the value of around $1, but it's not super stable and especially, I see some really odd spikes to up to $2 now and again. So, is that the mechanism working as it's supposed to or are you refining it? Or what are those huge spikes that you see there?

BI: Great question. So far the system is working as designed. There's no surprises there. It is important to keep in mind that we are a relatively early project with a relatively small network size and we don't have the luxury that the US dollar does in terms of denominating everything and so we should expect fairly high amount of volatility in the system just from ETH. And so, we look at the project in terms of life cycle, different periods of its lifetime. In the early days, we do expect a fair amount of volatility, which is not necessarily a bad thing. Bitcoin in its early days was a very sort of speculative asset. It's, I think, very similar for us right now.

Over time, as the market learns how to deal with this sort of unique asset, they'll be better at sort of handling in the sense that once we get out of the equilibrium band —so there's a band of the price target, which gets supply adjustments happen. So, we want supply adjustments to happen only when there's some sort of structural change in the underlying demand. When that happens, we go outside of the equilibrium zone, we go into expansion or contraction, the market is essentially in a race to guess what the next equilibrium supply should be. So, once we start expanding, we should continue to expand until we get to the next supply that the economy is able to support and same with contraction.

Volatility is a necessary part of the system, because that's what engages the supply change to happen to begin with. And then as we get larger, as the market gets more experienced, we expect that the windows of time when we are in the equilibrium band will extend and become longer and the windows where we are in the dynamic states will get shorter.

“We want supply adjustments to happen only when there's some sort of structural change in the underlying demand.” (…) “Volatility is a necessary part of the system, because that's what engages the supply change to happen to begin with.”

CR: Are those windows predefined? Like what's the equilibrium window? Is it like a specific price?

BI: Yeah, so great question. We have essentially two hyper parameters of the system. So, one is the size of the equilibrium band, so right now, that's plus or minus 5%. And then the second hyper parameter is a smoothing factor, we call this the Reaction Lag.

CR: And it's 5% plus or minus from a historical average of the price?

BI: From the price target. So today, the price target denominated in USD is $1.01, so it's plus or minus 5% of that.

So, going back to that same question, we did make one change. You know, a few months after lunch, we realized that the system wasn't responding as fast as it needed to reach equilibrium and so we have made a single change to a hyper parameter. We changed the lag from 30 to 10. So, it's a smaller factor. And so, given the nature of crypto marketplaces today where there's a lot of volatility, that makes sense. So, once we made that one change, we did see it go back to the equilibrium zones much, much faster.

So those are the two hyper parameters of the system. One thing that we strive to as much as possible is to reduce the complexity of the system and reduce the needs of governance as much as possible. So, because we see the system as being rules-based and not up to the discretion of individuals or groups of individuals, we want this to be as autonomous as possible, and so, it's great to have decentralized governance, but even better than having a decentralized governance is not having the need for it. And so that's sort of our governing philosophy right now is to reduce the surface area of that as much as we can.

Minimized & Centralized Governance

CR: Okay. And right now, what's kind of your governance system? Like are people voting to make any of these decisions?

BI: So, I think it's still too early in the project to move to any sort of decentralized governance. In the early days, that's when there's the most risk. So, right now, it’s under the control of the development team through multisig wallets on chain. So there's no one person with unilateral ability to make any changes.

“(The project) is under the control of the development team through multisig wallets on chain.”

And we think this is the right form of governments today, but it's not necessarily what we want to stick with. So, we've made public posts about how we see this potentially evolving in the future. But we also want to stay open minded as well, because there's a lot of great development going on in the space through other groups. And there's a great argument you can make that we should not enact any form of governance that hasn't already been proven out somewhere else. Because we don't really have a shallow end of the pool where we can make mistakes. Because our need to government are still small, every piece is critical for the functioning of the system and so, we want that to be as reliable as possible when we get there.

So, everything is completely transparent, because it exists on chain and we use all of our social channels when we make any changes, which have been very small up until now.

CR: So, far you've been live for over a year?

BI: We are coming up to a year, I think this week. We launched about a year ago last June.

CR: What has adoption been like? Volume?

Bitfinex IEO

BI: We launched on Bitfinex, we were one of the first projects to run an IEO. An IEO is an event with a lot going on. There's a lot of fanfare, a lot of exuberance And so, there's a lot of impact on projects just in the marketplace, purely in terms of price and demand over that period, which is not necessarily the best kind of environment to launch a project like this, which is supposed to be a roughly stable asset. Because a lot of people who participate in these types of events, they're almost like ticket scalpers in a lot of ways, flippers. They're maybe only looking a day or less into the future when they're participating. And so, there's a lot of market churn you have to overcome.

Then from there, it was just a matter of being in the marketplace and laying the rules. After all the fanfare, there was a prolonged period of contraction that you can see in the history of network. By the way, we have a dashboard on our website, Ampleforth.org/dashboard that shows all the stats. I definitely recommend people go visit because it shows the workings of the system.

“We launched on Bitfinex, we were one of the first projects to run an IEO. An IEO is an event with a lot going on (…) After all the fanfare, there was a prolonged period of contraction that you can see in the history of network”

And then after we made that one change the parameter, we did sort of level up. And then we've seen some slow and steady growth since then, as more people have heard about us. We're getting a little bit more attention. People are starting to be more interested in what we're doing now.

CR: I think Ampleforth is super unique compared with other stablecoins and tokens. I don't think I've seen a stablecoin that works in this way where the mechanism to achieve a peg is manipulating or changing the supply of the asset. And also, this minimized governance system.

I'm interested in your opinion of what role does this type of stablecoin have in DeFi? Because I think DeFi is very young, so, it's too early to say exactly how each asset will fit in the ecosystem, but so far it's been, Dai has been the dominant stablecoin for those who want the fully decentralized option and it's also used by people who want to gain leverage on their ETH and kind of put ETH as collateral to go long and issue Dai to buy other crypto.

So that's been a very a good product-market-fit in DeFi. And then there's USDT and the other more centralized stablecoins, which are more straightforward, and are used for arbitrage trading and that sort of thing, where you can be sure that these coins will maintain more or less their one-to-one peg with the dollar. So in this environment of stablecoins, where do you think AMPLE can play a role?

BI: Sure, yeah. So, there's a lot to unpack in that question, so I'll try to deconstruct it to the best I can. I think first of all in terms of DeFi, it's good to make a distinction between a primitive and non-primitive. Now, I think that's very illustrative for understanding the space and how things fit together.

So, a primitive in DeFi would be ETH, perhaps some trustless version of BTC, uncollateralized monetary assets like this. Dai is, I'd say, the most predominant, most successful stablecoin so far today, but it is itself a derivative, because it's based off of credit and debt with underlying collateral. So, there's a certain amount of risk in that system.

Under-Appreciated Risk

When it comes to monetary assets, I think there is some risk that's been under-appreciated in the space in terms of debt markets. So, with sovereign monies, like the USD, they can sort of compel usage and operate as a lender of last resort. When you're dealing with independent monies, you can't compel the bailouts the same way you can with sovereign money, so the best you can do is offer collateral or assets up for sale and then hope that people buy them. In times of stress, this isn't necessarily guaranteed. There's the common saying that in times of stress, all correlations go to one, and that’s downwards. So, we saw this on Black Thursday, across all markets, not just crypto, where people knew something dangerous is going on and so, their first action was to exit all the positions they could and get onto cash to give them time to figure out what they need to do next. So, people weren't necessarily looking to buy assets that were falling in value. They were trying to sell them.

“When it comes to monetary assets, I think there is some risk that's been under-appreciated in the space in terms of debt markets (…) When you're dealing with independent monies, you can't compel the bailouts the same way you can with sovereign money, so the best you can do is offer collateral or assets up for sale and then hope that people buy them. In times of stress, this isn't necessarily guaranteed.”

When you offer assets for sale, you can sort of cross your fingers and hope that the profit-motivated actors will bail you out, but that’s not guaranteed. So, there is some sort of risk in any system that's built on top of that external marketplace. So AMPL, similar to ETH and BTC is uncollateralized. And so, we view it more as a primitive.

Going back to what you said earlier in terms of similarities or differences to other assets and volatility and that sort of thing, the volatility system in the early days is, very important because this asset operates according to unique underlying rules, there are unique incentives in the marketplace to these two unique behaviors in the marketplace.

We published a paper a little over a year ago with our advisors at the Sanford Hoover Institute, where we outlined how we expected the movement of AMPL to operate in a marketplace where we call the sort of volatility fingerprint. And so, the short version is that we expect there to be a sort of step function like movement pattern in the marketplace with AMPL. There are periods of equilibrium and then periods of volatility, which give Ample the possibility of decoupling from other assets in the space.

Right now, if you pull up in a CoinGecko and look at the price history of any of the assets, they all pretty much go up and down together, especially if you look at the monetary assets, because they're essentially all equivalent. If we're not talking about special governance tokens or anything like that, just pure monetary assets, everything pretty much goes up and down with Bitcoin, even ETH.

Diversified Collateral

That creates some risk for you in the DeFi landscape as well, because if you have a system that's based on collateral, you want to diversify as much risk as you can. Everyone is trying to say, don't put all your eggs in one basket. So you do that by diversifying through a number of assets that have unique characteristics. Everyone who owns a 401k, for example, they don't hold everything into Tesla stock. They might call it an index fund to sort of diversify as much risk as they can.

“Right now, if you pull up in a CoinGecko and look at the price history of any of the assets, they all pretty much go up and down together (…) That creates some risk for you in the DeFi landscape as well, because if you have a system that's based on collateral, you want to diversify as much risk as you can.”

CR: So, in that sense, AMPL would be good as collateral for Dai rather than trying to compete with Dai directly, so that in a cases when there's a Black Thursday event and everything is dropping, AMPL has the way to decouple from that volatility?

BI: Potentially, yes. I think AMPL would be a great addition to the basket of collateral in platforms like Maker or even platforms like Compound or Aave. Even individuals can make use of this tool if they're adding collateral to these lending platforms, they can diversify the collateral they put in there so it makes them less likely to go underwater just on the individual level.

“I think AMPL would be a great addition to the basket of collateral in platforms like Maker or even platforms like Compound or Aave.”

CR: Makes sense. Would you also see it as a means of exchange? People using AMPL to price goods and buy and sell goods and services or not so much?

BI: It could become a medium exchange at some point in the future, but I don't see that happening anytime soon.

CR: It’s not as stable yet.

BI: We shouldn't expect a lot of stability of the tokens themselves. Now, unit of account is separate and I'll get to that in a second here. But I don't think it would be used to buy your morning coffee anytime soon. But also, I don't really see a demand for people right now to spend their crypto to buy morning coffee. So, I don't think that's really that big an issue and is not what people are necessarily asking for even. In the future, once we have reached prolonged periods of stability, then it could become that sort of alternative to central bank money, but it doesn't need to do that today to still be useful.

“In the future, once we have reached prolonged periods of stability, then it could become that sort of alternative to central bank money, but it doesn't need to do that today to still be useful.”

AMPL-Denominated Debt

Before we started denominating your goods like coffee in AMPL, I think it's much more likely that we'll start denominated contracts and that in AMPL. So, going all the way back to the beginning when we talked about the weaknesses of fixed supply assets, so if I were to borrow 1,000 Ample for a year, a year from now, I know exactly the purchasing power I will have to repay, because 1,000 Ample is targeted to the 2019 US dollar, irrespective of how the supply changes in the meantime. I know exactly how much I'm going to have to repay, right, so that drastically reduces the chance of defaulting on my debt, because of the predictability.

That's not the case with any primitive right now. Various stablecoins have risk from either centralization risk, like USDC, or risk from external debt markets, in the case of collateralized stablecoins So, this is a sort of unique offering, the ability to safely denominate debt with the smallest amount of risk possible.

CR: Interesting. But right now, we're not really seeing that use case play out in DeFi, right, not many people are actually issuing bonds in DeFi.

BI: Well, I guess, depending on how you look at it, a Maker type system is sort of based on a collateralized debt position. It is sort of is built into the name, Collateralized Debt Obligation. So, when your collateral gets underwater, it gets offered at auction and hopefully, people will buy that at auction.

So, I was talking to an engineer recently from another project in the space who was looking at this and trying to figure out, okay, how can we make this stronger? And so, they started working on a project and the whole purpose of this project was to get people to commit funds to buy MKR tokens at auction if you ever got there. But this is a lot to ask people because they essentially have to reserve capital and set it aside, not do anything useful to be able to buy something in the future that maybe they want to.

Just because you offer something for sale doesn't mean people will buy it. And so, there's some risk there in times of stress, which I think is important for us not to overlook. A lot of people did and they were upset with the outcome.

“Just because you offer something for sale doesn't mean people will buy it. And so, there's some risk there in times of stress, which I think is important for us not to overlook. A lot of people did and they were upset with the outcome.”

CR: Right. So, right now as everything else in DeFi, it would have to be still these collateralized debt positions. But yeah, I was thinking more of a case where it’s just like a traditional bond sale where you're getting money from investors and issuing a security where you promise to pay back in the future plus some interest. And in that case something like Ampleforth that doesn't have like this centralized stable coin risk and doesn't have the volatility of other crypto primitives like ETH and BTC would be useful. But we still need to get to the part where people can actually issue decentralized bonds like this, which I think you know some teams are working on, but not quite yet.

BI: But yeah, it's really interesting to look at the current behavior of these lending platforms in space right now. So, if you look at the markets on many platforms, almost all of the borrow activities is for stablecoins, mostly collateralized stable coins. Which is funny given the space is supposed to be decentralized as much as possible. But the reason why people are willing mostly to buy stablecoins is because it means you're not exposed to the volatility of another assets. Most of the behavior of people borrowing right now is to leverage long speculate, because they have ETH and they want to use that as collateral to buy more Ether or some other asset. So, they'll borrow stablecoins and buy that asset. They'll put that asset they borrow to work. Then when they repay, they had the predictability of how much they're going to have to repay.

Liquidity Mining

CR: I wanted to talk about this latest, hot thing in DeFi with liquidity mining and I know that Ampleforth has some news coming up in this area. So, if you can tell us about this new development and talk about this new trend?

BI: You have the space of on-chain liquidity is actually super interesting right now. So, there's a lot of activity in terms of projects you just mentioned. Liquidity mining is a term that seems to be all over the place now. But, if we look at what the project Ampleforth needs if it wants to have these sorts of use cases we talked about, then it needs a lot more liquidity available. So, people will need to be able to enter or exit AMPL positions whenever they want with low slippage.

And so, to help that goal, we developed a system we call it the “Token Geyser”, it’s sort of like a smart faucet. So, what this does is it provides rewards from our ecosystem funds to individuals who provide liquidity.

We are targeting this Uniswap V2 for the pilot program. And so, the basic structure, that is people can provide liquidity onto the ETH-AMPL pool on Uniswap V2 and then you’ll take those liquidity tokens and then you deposit it into the Geyser, which is also a smart contract on Ethereum. And then the more liquidity you provide for longer, the greater share you get of the drip of tokens from the system. So, we distribute tokens from ecosystem funds into the geyser and that unlocks over a period of time.

“The more liquidity you provide (to the Uniswap V2 ETH-AMPL pool) for longer, the greater share you get of the drip of tokens from the system”

So, there are two goals with this. One is we want to distribute ecosystem tokens in the most sensible way possible. And then we also want to reward people who help the overall health of the ecosystem. And so, I think, this says both of those.

CR: To recap, the idea is that you have a liquidity pool in Uniswap V2, it'll be an ETH-AMPL liquidity pool and you will reward liquidity providers of that pool. And so, these are people who are staking AMPL and ETH to this pool. And so, when they do that, they'll get a token in return. And with that token, they can go to this geyser and get more Ample.

BI: Exactly.

CR: So, how big is this fund? How many AMPL are you thinking of distributing? And is this a limited time exercise?

BI: We're starting with a pilot run of this program, and it’s going to last three months. And so, we've devoted 75,000 AMPL to be distributed to two or three quarters over the few months. Now, that amount can change, if the system expands, that number can go up; if the system contracts, that the number go down.

The pilot program will also target Uniswap V2, but it's not tied to any particular downstream platform, so, it's written in a general way. And we want the system to last a long time. This isn't like a one-off project. We compare it to Bitcoin mining more than anything else. AMPL is not a Layer 1 protocol and so it doesn't have miners. Bitcoin has miners and it issues new tokens to people who provide proof of work services. For us, providing liquidity is sort of the equivalent way of helping ecosystem perform. Replace proof-of-work with proof-of-liquidity and that's sort of where you end up with the Geyser.

“For us, providing liquidity is sort of the equivalent way of helping ecosystem perform. Replace proof-of-work with proof-of-liquidity and that's sort of where you end up with the Geyser.”

Impremanent Loss

CR: And the purpose of this is to make sure that there's enough AMPL there for people to buy and sell AMPL without having there be a big spread between those positions?

Brandon: Exactly. Yeah. So, liquidity on indexes is actually a tough puzzle to solve. Because if you look at these automated market maker platforms like Uniswap or Balancer, they were designed to address the liquidity problem of the long tail of tokens. So, if you have to rely on order books to provide liquidity, then it's just hard oftentimes to make trades happen when there aren't piles of traders in the marketplace the way there are for BTC and EET and the top coins today, right.

And so, these AMM systems, they work with two pools that sit next to each other. And so, when you buy or sell into Uniswap, you buy or sell into a pool and then the relative sizes of these two pools change as you swap out of them, which changes the price. So, the price is just a function of the relative size of these two pools.

Okay, so that's roughly how they work, which in theory sounds great, you know, you have the constant liquidity available provided by these pools on chain for traders that don't need a counterparty. However, there is a pernicious problem with these called an Impermanent Loss. So, Impermanent Loss is something that you can be exposed to as a liquidity provider on these AMMs when the relative market price of the two coins changes.

If trading with coin A and coin B and have the exact same price and they stay in lockstep together then there's no impermanent loss. But if the price of B goes up by 2x relative to A, then arbitrageurs can come and extract some value from the system. So, when the prices diverge, liquidity providers lose potentially value and then they have to make up that value through trading fees. Uniswap has a 0.3% trading fee that goes into the pot of collateral, that's claimed by liquidity providers. But the longtail tokens, they don't have necessarily a whole lot of volume to create these fees to overcome the impermanent loss.

These AMMs, they have a great intention of trying to help liquidity for longtail tokens, but they're not always able to do so in a healthy way for liquidity providers. So, the goal for the Geyser is to one, stimulate providing liquidity for AMPL token so we can get large enough to unlock the potential of the token. So that's the first one.

“These AMMs, they have a great intention of trying to help liquidity for longtail tokens, but they're not always able to do so in a healthy way for liquidity providers.”

And then there are also exciting developments going on in the space of AMMs too that I think could potentially alleviate this. So, I think V2 has been announced is going to be launching sometime this quarter and they claim to have a solution to the impermanent loss problem. They haven't published the exact mechanism how it works, but people have been speculating here and there, but we do know it relies on an Oracle. So, you're able to get this if you're okay relying on an external oracle system. And this might be getting kind of nerdy for your deep listeners here who might this mechanism like this.

But there's a really neat interaction, I think between the AMPL token and Balancer. So, Balancer is an AMM, sort of like Uniswap, but it's generalized, so you can have weights associated with each pool. So, you can have a 50-50 or you have a 60-40 pool, you can have more tokens if you want to, but for simplicity, let's just consider two tokens, so, AMPL and USDC. So, if the supply of AMPL went up 10X and USDC stays at $1, then normally, liquidity providers wouldn’t suffer impermanent loss.

But the cool thing about AMPL is that since the supply is adjusted programmatically on chain, we can also programmatically update the weights in the Balancer pools, which means that liquidity providers are able to supply liquidity for Ample in something like Balancer, trading as a stable coin and suffer no impermanent loss. So, that's the sort of special combination.

CR: I didn't quite get how that works. Okay, so I understand that Balancer automatically rebalances their pool so that the weight of both tokens stays the same. So, if it's 60 AMPL, 40 USDC, that will always be the case because of how these pools work. But in addition to that, you are saying Ample also adjusts for the supply of AMPL in the pool?

BI: Yeah. So, I'll go through an example to make it really simple. So, say we started with a 50-50 of USDC and Ample and then a rebase, you know, just the supply of AMPL across the entire the blockchain, so now, say we doubled the supply of AMPL. So normally the value of the AMPL pool would be twice the size of the USDC pool, which would create impermanent loss. Because it’s a 50-50 pool in Balancer. But since we know that we're doubling the supply of AMPL, we can at the same time as we adjust the supply, change the weights of the Balancer pool. So, instead of being a 50-50 pool, we turn it into a 66-33 pool.

CR: How do you do that? Is that something that someone has to go and program or is it done automatically?

BI: Yeah, so we have the ability within the protocol, we have what's called the orchestrator that notifies downstream platforms of supply adjustments. So, this happens automatically in the same operation that adjusts the supply.

Best of Bitcoin and Stablecoins

CR: Okay, so this orchestrator would notify Balancer’s liquidity providers that this is going on, that there's going to be an increase in the supply of AMPL and so, the weight in the pool needs to be adjusted?

BI: Pretty much. Yeah. So, there's the concept of a smart pool in Balancer. So, it has this special functionality. And so, we haven't implemented this yet, but we'd like to very soon. This could become a future platform incentivized by the Geyser so that you would no longer be open to impermanent loss. So, a lot of people want to be able to trade AMPL versus USDC, because it gives you easy price discoverability. So, I know that if there's a price of $1.05 or like 1.05 USDC, it's roughly $1.05 for a token, right?

If I trade against ETH, if it's like 0.0045 ETH versus Ample, it's hard to know exactly what price I’m in without going through some extra steps. So, as a user, you want to be able to trade against these stable coins, but because impermanent loss, every project trades against ETH, sort of by default, because it's another floating price token and they're relatively correlated, so that's the idea. So, unless you're a stablecoin, every token pretty much trades against ETH on these AMMs, even though users would probably prefer trading against stable coins.

So, AMPL is kind of unique. We often say that Ample has the best of Bitcoin, and the best of stablecoins. So, this is a situation where, you sort of get the best of a stablecoin because you can trade a stablecoin with good price discovery and without any impermanent loss. But you also have the variable supply that gives you the same sort of non-dilutive properties that you would get from a Bitcoin. So, in this, smart pool, we can have hop from stable coin to a floating price token and then other floating price tokens could potentially trade against AMPL the same way they trade against ETH because they're both floating price.

And so, you have this sort of two hop system for liquidity providers where they're not exposed to impermanent loss. So, it's really interesting to consider Ample as a sort of trading pair on these systems, because it has unique qualities that kind of can draw from the best of both worlds.

CR: Interesting. And you said, this Geyser program will also be used in Balancer not just in in Uniswap, so people who are providing liquidity on Ample will be able to get these additional AMPL tokens?

BI: So, the current pilot program is starting solely in Uniswap V2. So, we chose Uniswap V2 because it's already launched and deployed and sort of battle tested in adversarial environments. It has a strong ecosystem of developers, large base of users. And the toolset around Uniswap V2 was end to end with our sort of unique elastic supply. And so, where I talked about before with these other AMM platforms, it's very exciting, but it's still a little bit speculative, because they haven’t been built yet. But it's interesting things that could potentially come down the line.

Multi-Chain Future

CR: This is all you know, ways like for AMPL to really achieve its long-term potential. So, you needed to gain more liquidity, to gain more adoption for people to start trading it against other tokens, using it as collateral. So, in this kind of long-term vision, where do you see AMPL going and more broadly, where do you see DeFi?

BI: Great question. In terms of where AMPL goes, I think because we're a monetary asset, I put the peers of AMPL to be the other monetary crypto assets today, right, so Bitcoin, Ethereum, Zcash and things like that. AMPL is today deployed on to Ethereum. We also have a vision where we list on multiple chains simultaneously. We could potentially deploy onto another blockchain, like Polkadot. And so, as long as there's a bridge from Ethereum to Polkadot, where I can trade from my Ethereum wallet to a Polkadot wallet and back. And as long as the monetary policy can adjust both chains at the same time, there's essentially the same currency, even though it's on different platforms. I think that's the sort of the medium-term roadmap for the project.

And then the long-term roadmap of the project is a sort of monetary asset with stable properties, but still has the nondilutive property of Bitcoin. And then the very long-term, it could even become something you buy your morning coffee with, an alternative central bank money. That's something that’s much, much further in the future. It's not something we're paying attention at for anytime soon. That's how we view Ampleforth developing.

I think DeFi in general, DeFi right now is both exciting and worrying, in the sense that it's exciting because we're starting to see emergent behavior that's arising from the combination of these Lego blocks that weren't designed for when this space was originally been created. All the dizzying variety of ways that people are farming yield today, were not necessarily imagined when we were first creating these systems. I think that's really exciting. It's a proof of success for the composability thesis of DeFi. It's really cool to see start to happen today.

“All the dizzying variety of ways that people are farming yield today, were not necessarily imagined when we were first creating these systems. I think that's really exciting. It's a proof of success for the composability thesis of DeFi. It's really cool to see start to happen today.”

But at the same time, I worry that we're not appreciating enough the risks that we're taking on these systems. Particularly when these systems rely on external markets, because external markets, they don't always work the way you'd like them to. Sometimes markets freeze, markets are not 100% efficient. Just because you offer something for sale doesn't mean that there's going to be counterparty. I think, as a collective group, we need to keep that in mind and be very honest with our users about where these risks can potentially happen and find ways to address those risks upfront really as much as possible.

CR: Totally agree. I also see DeFi as really promising right now, really exciting. But super early stage and an experimental and so we still need to find the right ways to protect against this incredible volatility and smart contract risk and bots. But I think the space progressively gets stronger as we suffer kind of the consequences. Right?

BI: Exactly. That's the cost of experimentation. You know, economics so far has mostly been about observing and reporting. But now with these tools, we can start building and experimenting, so I think that’s really cool.

Camila: Yeah. It’s exciting. Hopefully, Ampleforth will be a key piece in making this space safer with an uncorrelated asset and a stable decentralized asset. So, we'll see that play out.

BI: We hope so.

Hope you’re enjoying The Defiant. If you are, spread the word!

Share The Defiant

The Defiant is a daily newsletter focusing on decentralized finance, a new financial system that’s being built on top of open blockchains. The space is evolving at breakneck speed and revolutionizing tech and money.

About the editor: Camila Russo, is a financial journalist writing a book on Ethereum with Harper Collins. (Pre-order The Infinite Machine here). She was previously at Bloomberg News in New York, Madrid and Buenos Aires covering markets. She has extensively covered crypto and finance, and is now diving into DeFi, the intersection of the two.