Join Delphi Research today and immediately get access to our full Member Portal!
Join Delphi Research today and immediately get access to our full Member Portal!

Terra Autumn Series Ep. 4: Mirror Protocol Enables On-Chain Price Exposure to Real-World Assets With Synthetic Asset Tokens

Oct 21, 2021 · 45 min media

By Jose Maria Macedo

Jose Maria Macedo sits down with Stanford Liu (Head of Research) and Joe Lim (Product Manager) of Terraform Labs, the team behind Mirror, a protocol that enables the trading of fungible assets, “synthetics,” that track the price of real-world assets. The trio discuss Mirror’s synthetics architecture, implementing concentrated liquidity for mAssets (“mirror” versions of real-world assets), notable upgrades in Mirror v2, and much more!



Interview Transcript:

Jose (02:01):

Today I am thrilled to be hosting Stanford Liu and Joe Lim from Terraform labs, Stanford and Joe contributed to architecting Mirror protocol, Terra’s synthetic acid primitive, which has since grown to become a behemoth, with nearly 2 billion in TBL across both Ethereum and Terra. Most importantly, Mirror is a key part of Terra’s value prop of making UST useful by allowing users to invest in traditional assets using Terra’s stable coins. In this podcast, we explore what Mirror is, why it matters, as well as the v3 roadmap. Stanford and Joe, thanks very much for being here.

Joe (02:32):

Hey, thanks for having us.

Stanford (02:34):

Thanks for having us here. [inaudible 00:02:35] 

Jose (02:35):

Awesome. I mean, before we kick off, I’d love to get a quick description of kind of your backgrounds and how you ended up getting into crypto.

Joe (02:43):

So previously, I was working in a company that operated a validator node. And terra and a lot of the other projects in Cosmos ecosystem. Then I got to know Terra a little better. And I had an opportunity to join here as a PM. And I thought making products on blockchain ecosystem is something very interesting for me to do. And I [inaudible 00:03:06] missed a chance and came here and started working on Mirror. Yeah, that’s [inaudible 00:03:12].

Stanford (03:11):

As for me, my background’s slightly difference. So originally, I actually worked in quantitative finance, in trading and research in New York. And then I ended up moving back to Korea and working with Joe on the research side of things as well as working with Joe [inaudible 00:03:27]. And then currently at Terra, we’re working on a few other projects besides Mirror. So we’re really excited to see how this pans out after the release of Columbus five.

Jose (03:35):

That’s awesome. And how did you both come across crypto. Joe with the added you start working at the validator company.

Joe (03:42):

So before I started working at about the company, I was actually on a FinTech industry. And a lot of my co-workers were talking about crypto. It was in 2016 and ’17 when everything was going on, I started taking a look into it. And I saw that something called Cosmos was coming up. And it was a blockchain that enabled a lot of the other blockchains to be built on top of. And I think Cosmos is like the gateway of how I got to know blockchain in the first place. And after that, a new Terra and all the different products that are being built on Ethereum we’re moving on to Terra. So I just found that much interesting. 

Jose (04:25):

Very cool. What about you Stanford? How did you come across crypto? 

Stanford (04:29):

Yeah, so it’s a little bit weird story when I was really young, I played a game called Runescape and so that’s how I learned actually a lot of my English. A. But inside there, there was like, I guess like the in game currency and what I would do is I would find the in game currency and like kill other people in the game and get their stuff and then sell the currency for crypto. That was like my first exposure to BTC and Ethereum several years back, and then naturally as I was really interested in trading and things like that, I was a pretty big trader. I think slightly different background from Joe. I think Joe comes from like a product background. So it gives him like a lot of insight into there. And then I come from both like the financial side of things. 

Jose (05:10):

That’s really funny. That’s actually the same way that Piers, on our side, Piers[inaudible 00:05:15] that kind of runs the gaming side of our business, also going into crypto. He was like on Runescape, PKing and farming and that’s how he first got into Bitcoin as well. And how he came across crypto. So I think a lot of people actually have that story, which is interesting. 

Stanford (05:30):

Yeah, super cool. 

Jose (05:32):

So your official title is kind of head of research. And Joe, you’re more on the product side. But I’d be curious to know like, what is it day in the life look like for both of you at Terraform labs. I think people are super interested to know kind of what happens at Terraform labs. So yeah, I’d be curious to know what that looks like.

Joe (05:47):

So for I’m actually working on currently three different projects at the same time right now. So which means app to have a lot of meetings with core developers and front end side and designers as well. And like a lot of these decisions that are made are not done in a single meeting. So we have multiple meetings throughout the day. And at the end of the day, I try to like organize those things into documents so that it can be shared. So that we’re all going the same direction in terms of making the same product.

Jose (06:20):

Makes sense. What about you Stanford, what does the head of research at Terraform labs do?

Stanford (06:25):

So I think, at Terra in general, I think also for Joe as well that the rules are a little bit blurred. So I do work on a little bit of product with Joe as well, as well as working with our external partners with the projects that they’re planning to release. So we do talk with you guys, for example, [inaudible 00:06:41]. 

And then as well as many other teams. So for example, we help support [inaudible 00:06:45], we help support pylon protocol, we have support a bunch of different projects, and give them an advice from both the technical standpoint as well, as just from the things that we’ve learned while working at Terra. 

Aside from that, so I do do research, as well as COO help, I have a trading team may assist in other projects, as well as like fundraising person projects, and things like that. So I think, in general, there’s like, no, what do I do daily. My daily schedule is pretty varied depending on like the type of thing I’m doing. So some days, I’ll work a little bit closer with Joe. And then some days, I’ll work a little bit closer with external community members.

Jose (07:22):

Super interesting. Okay, so pivoting to Mirror. In 2020, the sort of narrative around Terra was just stable coins. And the main narrative was around Chai. So using tariff stable coins to actually pay for stuff in the real world. Korea runs on Terra and all of that. I’m curious, at what point did you guys decide that you needed a Tefi, like Terra’s DeFi ecosystem? And why did you decide that synthetics was kind of the first primitive that you should launch?

Stanford (07:50):

Yeah, so I think, like a pretty easy and understandable use case for it like stable coins in general currency as payments. So I think that’s where the whole idea of we should make Chai, we should make a payment platform system to allow users to make really efficient transactions. And so as we did that, and it was wildly successful in Korea, and it’s still a very popular product in Korea. We decided to look more into what Ethereum was doing, what other ecosystems were doing. 

And in general Chai is targeted towards the average user. But whereas near and anchor and all these other products, they’re targeted, in general, towards crypto users. So I think we wanted to be a little bit more crypto native focus, a little bit more on our crypto community. And so that’s where we started coming up with a bunch of ideas. What type of product should we release? How can we make it so that Terra is one of the more widely used blockchains, aside from just payments and payment platform systems. And so back then, actually, when Chai was released, there was actually very little smart contract functionality. And so I think, doing that shift between China to Mirror and Anchor, smart contracts came out and became supported on terra. And so that’s where we started the full development process. 

Jose (09:04):

Makes sense? And what was the long term vision for Mirror. I know it’s a community owned project and kind of the community makes decisions on what to develop. A. But what did you all hope it would become. Why was that the first primitive.

Joe (09:18):

So in the end, we actually want Mirror’s core functions to become an exchange where people can trade any type of asset [inaudible 00:09:27] real world in a synthetic world. And to enable that we are trying to figure out a lot of ways to list more assets without having Mirror rewards diluted across like multiple assets. And trying to find ways where things could be more community driven and more decentralized. So that there is absolutely no interference from our side or purely just driven by community through governance. 

Jose (10:01):

Very cool. And how did you decide on… A. Because obviously I guess you decided you wanted to do synthetic assets because it was important for Terra users to be able to have exposure to traditional assets. But then how did you settle on the architecture because Mirror’s architecture is a little bit different. It’s an individually collateralized architecture, whereas maybe the most well known synthetic protocol on Ethereum, Synthetics, it uses more of a pooled collateral architecture. Whereas the next stakers are the counterparties to each trade. I’m curious what research did you guys do into the different models? And how did you land on Mirror’s architecture?

Stanford (10:36):

Yeah, so I think we actually looked at a lot of models, including synthetics, UNA, a bunch of different synthetic asset protocols. And I think one of the main… so we do think like Synthetics is a great model. And it was definitely a big inspiration for the team as we were developing our product. One of the big things that we wanted to avoid was actually to avoid pooling risk among users. So that was our biggest motivation. So we wanted to separate risk and silo these risks depending on your assets. 

So in synthetics, I need to admit SUSD, using my SNX as collateral. so that ends up pooling. So it’s like an aggregated book where all users are exposed to the same amount of risk. So all the minters are exposed to similar rates of risk. In Mirror, there is the disadvantage that there is slippage and things like that. But there is the advantage where an individual mentor is only exposed to the amount of risks that they decide to take on. So I think that was the main advantage point, or the main reason why we decided on this structure versus like a different structure. 

Jose (11:38):

Yeah, that makes a lot of sense. And obviously, it ends up also being more capital efficient, because you’re using stable coins, a lot of the time, to mint versus SNX, which needs a lot more collateralization to kind of work. Are there any other alternative models that you considered?

Stanford (11:54):

So I think we did look at UNA. At a time, we figured that this model was actually the closest model to what we wanted to use. Obviously, the Mirror model is also not a perfect model. A. But basically took inspiration from Synthetics and [inaudible 00:12:10] actually. So the minting process is from like the MakerDAO, or the maker CDP process, and then for the liquidity provision and things like that this was taken from let’s say, Uniswap and [inaudible 00:12:22]. So you kind of combine two different mechanisms for two different protocols are made into a new system, what we call Mirror today. 

Jose (12:30):

Very cool. Yeah, that’s awesome. And I mean, the Mirror launched in December, like the growth has been pretty crazy, you’re 2 billion in TBL, huge trading volume as well. Were you surprised by how quickly it took off, not just in terms of kind of TBL and usage, but also in terms of the community, rallying around it and participating? What do you think was the big contributors to kind of Mirror succeeding?

Stanford (12:54):

Yeah, so actually, the funny thing was that before Mirror’s released, the only alpha on terra was basically [inaudible 00:13:02]. And that was based on TRT and at the time or marketer for USD was actually very low. So in the company we actually were thinking like when Mirror release, first month, like end of the year target would be around $10 million, but 10 million USD. 

And that was like clearly that broke expectations like buy several hundred fold. so I think in general what we’ve realized is that there’s a lot of demand for financial products both on Terra due to the native advantage that Terra has in terms of transaction speed, transaction fees and things like that. As well as in general the, I guess, popularity of financial products within the blockchain ecosystem itself. so it definitely broke our expectation so we’re very happy to be able to see that. Within the first few months it was like several 100 billion[inaudible 00:13:50] was already locked within Mirror.

Jose (13:52):

Yeah that’s interesting. Obviously Mirror’s doing a lot of volume but a lot of people or some people argue that the majority of this is still driven by kind of the farming rewards and people selling the farming rewards to compound versus people actually getting you know their stock exposure and commodities exposure directly from Mirror. What’s kind of your answer to this and how many people do you think are actually using Mirror to kind of get their stock and commodity exposure versus just kind of farming?

Stanford (14:20):

Yeah, so I think in general in DeFi just to bootstrap liquidity, as well as just the high yield chases, farming is definitely a pretty big thing. Now just in mirror but in DeFi in general. So I do agree that a lot of the actual volume is probably likely due to farming and compounding and things like that. So one of the reasons for this is due to the depth of liquidity. 

So due to the XYK curve that we currently use on [inaudible 00:14:46] slash Uniswap type model. It’s quite inefficient right to make large trades without a super deep, deep level of liquidity underneath. So one of the things that we aim to change in the upcoming versions is obviously to connect the actual points concentrated liquidity. Which I think will be able to offer much tighter spreads to users who want to actually make a relatively large trades. So that’s one of the main things that we’re waiting for in version three and version 3.5.

Jose (15:12):

Interesting. Yeah. I mean I actually personally use Mirror to get a lot of my stuff and commodity exposure. I know a few people who do that as well. But obviously, mainly holding versus kind of trading. Is there any way that… How do you think you incentivize that kind of trading volume to pick up or are there products that are being built on top of that you kind of feel have the potential to do that. 

Joe (15:33):

So like the Stanford mentioned on concentrated liquidity, I think one of the key things that exchange should have is like exposure to a lot of different assets. If you come into an exchange not expecting that the asset that you’re paying for isn’t there, then you can’t really use that as an exchange. But with concentrated liquidity, I think, we could increase the number of assets that are being traded on the protocol. And that would actually drive volume in the future, since people like people can easily find the assets they want to trade just on Mirror, and just by using USD and Terra [inaudible 00:16:16] transaction and cheap transaction fee.

Jose (16:19):

And you’re saying you would drive more trading volume, because you’d need kind of less TVL to guarantee good execution so people could spread their TVL across more pairs?

Joe (16:28):

It’s more like, right now, the liquidity rewards are diluted every time a new asset is listed on their protocol. But with concentrated liquidity, the rewards will be distributed to a specific range of prices that the liquidity is provided, which means that even with many assets listed on their protocol, rewards will be given to a specific range of prices for each asset. So I think a lot more liquidity can be provided for each asset class, and even with more emphasis. I think people for providing liquidity will get enough rewards so that trading environment can be more.

Jose (17:13):

Yep. Okay. So it’ll be more efficient, basically. You won’t be kind of… A. Because one thing people kind of don’t realize is the XYK is particularly inefficient for Mirror and DM assets. Because M assets are kind of several orders of magnitude less volatile than crypto assets, you have stuff like gold Bear. Which is like very, very non volatile. And so it ends up being like the XYK wastes even more liquidity than it does normally with the M assets.

Joe (17:39):

Yeah, exactly. 

Jose (17:41):

Cool. And so v1 was a pretty big success. But there were also some problems specifically with assets going off peg. Could you kind of explain what caused this problem and how you ended up fixing it in v2? Because I think that was a pretty successful and interesting solution.

Stanford (17:56):

Yeah, so I think due to the fact that Mirror was one of the first DeFi apps on Terra, as well as like the general large demand for users to either Farm Mirror, or to trade assets, or to hold assets for whatever reason, there was a pretty big demand. so we saw USD going above a dollar trading pegged to another stable coin. So there was a pretty big influx in demand into the Terra ecosystem. And I think that’s what pushed up the prices. So even if Apple was… The M Apple stock was trading a little bit above the actual stuff. Users were more than willing to pay this price because they would be able to farm beer and get a higher yield. And so that’s what we realized, that there was a really big demand. And so we want it to [inaudible 00:18:40] some, in some cases, to not lower the demand. 

But we wanted to make it so that we satisfied the amount of demand that’s being currently required by users, as well as being able to ensure that the price is sufficiently close to the actual underlying asset. So it’s pretty bad user experience for somebody to come into the platform. And say, I want to trade an Apple stock, but like the M Apple stock is trading at like a 20% premium to the actual underlying stock. 

So we want it to somehow balance these forces. And in doing so we thought of like the funding rate that’s often used for perpetuals. And so we implemented a system that rewarded users who were shorting. A little bit of a downside for that is the fact that a lot of these Delta hedge… Let’s say, delta hedge farms happen. And so now it’s like a very stagnant type of liquidity that’s being provided on Mirror. So we’re looking to fix this in the next iteration in the upcoming ones. 

Jose (19:41):

Very cool. And so you’ve also added a bunch of other features for v2, like pre-IPO assets, rewarding governance participants directly for voting, rather than just for staking and a bunch of other stuff. Can you talk about some of the most exciting features from your perspective, some of the stuff that you’re most excited about.

Joe (19:56):

So definitely like the incentive distributed to people who decide to shorter Mirror assets on terra swap is definitely the biggest feature we had. But also, the feature that we mentioned pre-IPO is also very interesting. Because once we have an expected price of an asset that is planned to go public on the market, like people can do governance, people can choose to list those assets before the asset is actually live in the real world. 

So once that happens, people can mint those assets at a price agreed on governance, and have some kind of price exposure and trading environment created by providing those minted assets into liquidity [inaudible 00:20:48]. I think it was pretty interesting. We did miss the time for this where we were actually thinking about doing pre-IPO for asset like Robin Hood. But I feel that there are more interesting assets in the future that we can try pre-IPO mechanisms with. And see how it actually goes. It hasn’t been really tested in the main net of Mirror yet. But I’m really looking forward for people to propose something interesting with pre-IPO.

Jose (21:21):

So how would the pre-IPO work? So you’d have governance decides on a price which is effectively like the Oracle price that people mint and burn at. A. But then trading can take the price around that or how would that work? 

Joe (21:35):

So do governance hold the price that will be used for minting will be decided. It’ll be at a fixed price until the actual IPO happens. And with the minted assets during the pre-IPO timeframe, people can provide liquidity into liquidity pools so that people can start trading. And once the asset goes live, the fixed price is swapped and changed into the actual price feed of the real world market.

Jose (22:06):

Okay, interesting. I guess once something like FTX lists pre-IPO assets, you could also potentially use something like that as an Oracle even sort of pre-IPO, for the minting. 

Stanford (22:18):

Yeah, so for FTX, we can actually utilize prices for FTX. And that’s one of the things that we actually considered. So this is obviously like a decision where users have to… The community has to think about how liquid these assets are. So pre-IPO assets on FTX in general, are also may also have problems with liquidity, and hence the price fees are very erratic. So that’s one risk that we’ve looked into and in general, it’s really dependent on the asset itself. But I do definitely agree that Mirror allows for the capability of price feeds from various different sources.

Jose (22:55):

Makes sense. super interesting mechanisms. And how do you think about like leverage? Because I remember when we first came across Mirror, we were saying the next step is that they’ll start minting sort of 2x and 3x assets, using the same mechanism, and then they can start enabling leverage. But you seem to remain pretty focused on the spot market. And there’s solutions like [inaudible 00:23:17] that are coming on board to kind of enable leveraged assets and then like perpetrating. Is the plan for Mirror to just be sort of spot like Synthetic exposure and then, like leverage other primitives like Mars or whatever it’s offered leverage, or do you plan on having like these 2x assets? And I guess the short assets are that in a way, but yeah, what are your plans there?

Stanford (23:37):

Yeah, so I think in general Mirror will stay focused on the spot market itself. So we do encourage, for example, other projects like Nirvana, Sigma, things like that, to provide users with the option to leverage their assets. And I think it’s a good thing to have different products or different teams working on these things, it spreads out the risk a little bit. I think, in general Mirror wants to serve as the basic foundation or basic infrastructure for other teams and other projects to use. 

So rather than us doing everything providing the full suite of options, we want us to provide the most basic stuff, and then all these other teams that build different mechanisms to work together across protocols, and then build a new product. Whether it’s leveraged, or whether it’s like a different type of financial product that uses M assets. Yeah, so in general, we prefer a method of decentralization, even amongst protocols, rather than one protocol, one team or one community doing everything. 

Jose (24:30):

Yeah, that makes a lot of sense. So you mentioned… I saw [inaudible 00:24:36] posted something on the forum about v3. I think one of the biggest features of v3 is concentrated liquidity which Astraport which we’re incubating is working on. We’ve kind of touched on why it’s such a big deal. Like what other v3 features are you kind of excited about, or if you want, we can just double tap on concentrated liquidity and kind of talk a bit more about that.

Stanford (24:56):

Yeah, so Astraport will provide concentrated liquidity and similar to in the sense of Uniswap v3. So one of the big things about Uniswap v3 is that it’s a little more complicated for a user to actually use in the sense that I’m not 100% sure, like where I should provide my range of liquidity. If the underlying assets currently trading at $100, in the most cases, it’s not an easy exercise to be able to say, I’m going to provide liquidity between 95 to 105. What’s the difference between 95 to 105, versus like 96, to 104, or something like that. 

So in general, Mirror will be utilizing Astroport’s constant trading liquidity, but on top of that, we will be building a vault that auto adjusts the liquidity ranges. so that users are able just to put their M asset there or put their [inaudible 00:25:44] under USD. And then Mirror’s contract itself will adjust that liquidity ranges based on the underlying prices. So that’s the big mechanism that we’re planning to work on while working with the Astraport in developing concentrated liquidity.

Jose (26:00):

That’s really interesting. So I guess you’re trying to solve that composability issue that you needed three kind of breaks. Where if you have these LP positions that are no longer fungible, they’re like NFT’s, then it’s kind of difficult to have that composability. And so what Mirror’s going to build is like this sort of strategy that’s going to be honored as a CW20. And it will be composable with other applications, is that kind of the good summary?

Stanford (26:26):

Right, exactly. And I think like one of the big focuses on top of that is not just providing rewards and things like that, making composability but also just improving the user experience. We want to make it… One of the big models of Terra and Terra ecosystem itself is to make products as easy to use as possible to the users. And so I think that kind of follows in the general theme of what we’re trying to at Terra and the [inaudible 00:26:50] platform.

Jose (26:49):

Yeah, that makes a lot of sense. And so will that be the only like strategy that’s under the CW20? I guess there’ll be other teams doing their own, but that will be kind of a selling point, at least for M assets, I guess.

Stanford (27:03):

Yeah. So I think the details are not 100% clear yet. But I think we also want to incentivize other teams, and basically incentivizes through governance as Terraform labs has no control over any of this stuff. So the other teams who build strategy similar in the sense of like, [inaudible 00:27:19], so the more strategies you build, like the vaults they built they can be provided some sort of, let’s say reward from both the Mirror ecosystem as well as from taking up maybe a percentage of the management fees.

Jose (27:30):

Make sense? And what does the strategy look like? Is it similar to kind of [inaudible 00:27:35] and Charm strategies on Ethereum? Are you doing something kind of different there? Could you walk us through it? 

Stanford (27:41):

Yeah. So the strategy is a little bit more complex than like Charm and Visors, although we do take inspiration from there. So I think in Visor, currently, they just use bollinger bands. So they basically look at the standard deviation and you put your liquidity somewhere within X amount of standard deviations.

What we try to do is basically just try to look at, like the historical distribution of the prices, and then look at certain characteristics… I’m trying to keep this as simple as possible, but look at certain characteristics in order to determine the range of liquidity, and specifically, not just the range of liquidity, but also like when we should readjust the prices. so we’re centering most of our liquidity currently, at the current spot price. How far does the spot price have to change before we rebalance our whole liquidity position? So I think like, the strategy is a twofold for Mirror where some of these other strategies is a little bit different. 

Jose (28:35):

Interesting, yeah. Because the kind of performance of the strategies on Ethereum has been a little bit underwhelming. So a lot of people say that’s because in order to really market make successfully on [inaudible 00:28:46] you also need to have like centralized exchange order flow as well. And to kind of be market making between the two, which is something that obviously, like a strategy alone will never be able to do. What do you think of that? And do you think there will be… Like these strategies do have the potential to kind of have pretty good performance? Or do you think they’ll always be limited if you don’t have access to, I guess, in Mirror’s case, like tapping into kind of volume from traditional exchanges?

Stanford (29:12):

Yeah, so I think in general, obviously, we hope for the Mirror strategies to do extremely well, at least in terms of helping users provide their liquidity within an efficient range. Obviously, having more market participants, whether it’s through a decentralized exchange or through a centralized exchange through market makers and things like that. It’s always good to have more participants that makes markets a little bit more efficient. But I think we’ll have to wait and see the actual performance of the strategy.

Jose (29:40):

For sure. And so, like with… switching gears a bit, we’ve seen a lot of increased regulatory scrutiny on DeFi which is prompting a lot of projects to kind of accelerate their timeline to decentralization. Mirror is probably one of the most successful in this regard. Like you have a very strong community. You did a fair launch. All of this. But I’m curious what are the main vectors on which you still think there’s kind of work to be done in terms of centralization around Mirror and that you’re actively working on.

Stanford (30:10):

So Mirror is completely decentralized in the sense that those specific entity controls any of these contracts. Everything has to be approved through governance. One of the few things that we do find as a potential attack vector on Mirror is that all these web apps or the front end interfaces of Mirror are actually not completely decentralized. So they’re either hosted through Amazon Web Services, or they’re hosted through somewhere else.

One thing that we want to do is basically incentivize community participants to basically host their own version of the web app, whether it be the same interface or like a different slightly modified interface, that we want to provide corresponding incentives to get these parties to be able to host it, and then also attracting new users or current users to their interface.

Jose (30:57):

Yeah, makes sense. What about oracles? Like, what’s the current oracle infrastructure for Mirror?

Joe (31:03):

Right now we are the Mirror Oracle is using band protocol. And the price feed, while band protocol is a decentralized Oracle. And we are using decentralized Oracle to provide these mirror protocol assets, we still want to give users the options to be able to choose from multiple types of Oracles to support specific assets. And think through our journey to be [inaudible 00:31:36] one of the things that we want to add is the capability to choose the type of Oracle the band or something else to whitelist an asset or maybe also change the Oracle for specific assets into another user’s if the governance users think that they’re unreliable. And I think this is something that we want to do pretty quickly. But even now, it’s still pretty decentralized. Because certain protocols such as centralized Oracle.

Jose (32:09):

Make sense. In terms of like UV3, and all this stuff, all the projects coming to Terra. At some point, there will be probably the need to like scale Terra. Like, how does Terraform labs think about the necessity to scale Terra? I spoke to Do about this earlier and you mentioned kind of IBC is actually like a scaling solution. But I’m curious if you’ve looked into that at all, because I imagine if Terra’s successful as we all think it’s going to be that’s going to be a necessity kind of fairly soon.

Stanford (32:39):

Yeah, so there’s a few ways to do this. Obviously, Terra has made several bridges, to different chains. And that was one of the major reasons why we’ve done this is to increase user flow into Terra ecosystem, and also export Terra native assets into other ecosystems. So in addition to IBC, [inaudible 00:32:59] also be in the future maybe in a month or so releasing a wormhole, which is a decentralized bridge between Terra and Savannah. 

So I think in general, we just want to open access to different chains and have users gain experience in Terra, and then also have Terra users gain experience in Savannah and other ecosystems like Ethereum. And in doing so, I think there’s like a mutual benefit to both ecosystems. Users are able to see oh on this ecosystem, they do something else really well. So we should perhaps bring up the Terra or like, Terra has the certain advantages, has these really innovative products that we can bring over to a different ecosystem. So in general, we want to facilitate like the communication across different chains.

Jose (33:39):

Yeah, that part makes sense, in terms of bridging, and adding a wormhole and IBC, are really cool there. But when do you think like actual transaction capacity is going to become a bottleneck for Terra? Do you have any plans to kind of scale whether it’s sort of separating some of the applications into their own app chains effectively with a different validator set. Or some form of layer two or something like that? Do you think this is something that you need to think about now? It’s kind of too early? And do you have any plans there? 

Stanford (34:09):

That makes sense? Yeah. So definitely, I think this is something we are heavily focusing on, internally within Terra and mainly specifically the engineering team. So we’ve recently seen that Savannah that was down for like a day due to like infrastructure issues. We’ve also seen like in launches, for example, within Terra of different projects the blockchain be slowed down and have transactions be piling up. 

So I think if 2021 was the year of us pushing out a lot of different projects, 2022 one of the main things that we want to focus on is basically improving the scalability of the blockchain. I think this is definitely a problem that we’re facing right now. It’s definitely something that’s very critical to us in improving. so that as users enter into the ecosystem, we’re able to actively support these ecosystems, support these teams. And we don’t want to break the user experience by saying like you submit a transaction, but [inaudible 00:35:07] confirm. Even a minute later, we want it almost instantly. And we’re taking a lot of efforts to be able to improve this in the future.

Jose (35:15):

Can you talk at all about any of the stuff that you’re doing there? I’m really curious to kind of understand what your approach to that is?

Joe (35:24):

[inaudible 00:35:24] very perspective. some of the instances are run only on a few different servers. For example [inaudible 00:35:36] or even for [inaudible 00:35:38] protocol, I’m sure it’s not run by multiple people in the community. But just like what we said about Mirror web application, we want to get the community more involved in spinning up their own instances, so that we don’t have to rely on one or two servers that are in order to maintain applications. 

And I think, similar to infrastructural side, I’m not 100% sure how to go about this. But I think it’ll be similar to our structural side where more of the instances are provided by the community could actually help with some of these transaction scalability as well. 

Jose (36:23):

Interesting. Yeah, I’m looking forward to kind of hearing more about that. I know you recently put in some big funding to kind of help scale Cosmos infrastructure as well. Could you maybe talk a little bit about that? 

Stanford (36:37):

Yeah, so I think, as I mentioned a little bit earlier, Terra wants to now support the development and building of new infrastructural services. And so we’ve allocated a lot of funds for teams to be able to ask for funding. So let’s say a team wants to improve Cosmos [inaudible 00:36:53], or they want to do something related to the infrastructure on Terra. We’re more than happy to fund this as well. 

And so as Joe mentioned a little bit earlier, we’re trying to decentralize in the sense that a lot of projects currently are using Terra provided public infrastructure, and we want to encourage them to set up their own infrastructure. So there are certain services that make a lot of requests to our infrastructure, and we want them to set up their own nodes. So that not only does it lower the burden on Terraform labs, as well as just public users who are using public infrastructure. A. But also provides a little bit of security in the sense that in the event that there are a lot of requests they don’t have problems as they’re not reliant on Terra providing infrastructure. 

Jose (37:36):

Yeah, this makes a lot of sense. But is this more for the front end and those kinds of goals versus the actual blockchain stuff? 

Stanford (37:43):

Yeah. So I think this would be more off the blockchain specifically. So front end services that make a lot of requests to our servers and things like that. With regards to the actual core itself as I mentioned, we are funding support for… We’re funding development for anything related to improving the blockchain or improving services on top of the blockchain.

Jose (38:05):

Yeah, that makes sense. Yeah, we would love to hear about those ones, once there’s more information on them. So as head of research for Terraform labs and as product lead for Terraform Labs, you must get to see a lot of the kind of coolest stuff and the bleeding edge of what’s happening on Terra. I’m curious, what are the projects that you’re most excited about other than kind of Astroport you’ve already mentioned, like what projects or initiatives that you’re seeing right now kind of get you really pumped.

Joe (38:30):

One of the things that are planning to come out in Q4 is called Nebula. Not that specifically because I’m also working on it. But I find this interesting that, so Nebula works in a way so that you bring a lot of different assets into one basket and create a ETF like asset. And while users are doing that, like they can provide a specific number of assets to balance the inventory. It will be called a basket. And they get incentivized for bringing those inventories to a specified target of values within the inventory. So I think it’s pretty interesting that you could create another form of asset by providing multiple assets such as, let’s say, [inaudible 00:39:21] Mirror or Anchor. And I think this will also be something the community is really important, and will be interested in using.

Jose (39:30):

Yeah, for sure. Yeah, the rebalancing mechanism for that is also super interesting. Rather than having these weekly or monthly rebalances, like traditional ETFs, and like most crypto indexes as well, it’s more of like a curve. That’s encouraging bottom up users to come in and actually rebalance it themselves. Can you talk a little bit about that?

Joe (39:53):

So if you think about it rebalancing isn’t something really easy to understand in terms of regular users, but we are planning to provide features that makes those rebalancing transactions easy, without having to calculate everything on their own. And with like some of these features [inaudible 00:40:18]. For example, if you enter a certain number of assets that you want to burn and redeem, the application will calculate the specific amount which will maximize the incentive you will get from this transaction. I think with these features that end users could easily use the application without having to think about all the difficult calculations.

Jose (40:40):

Yeah, that makes a lot of sense. So in the last T Phi Alpha [inaudible 00:40:46] talked about like this kill switch right in Terraform labs and the idea of Terraform Labs kind of started with nothing and kind of going back to nothing once you’ve helped kind of build out this ecosystem. I’m curious what do you see? How do you see that? And how do you see kind of your roles in the ecosystem going forward? Is it that at some point, you’ll kind of just be working on projects within the ecosystem? Or keep doing what you’re doing but without kind of this Terraform labs entity? How do you see that?

Stanford (41:14):

Yeah. So I think what [inaudible 00:41:17] was referring to is specifically in our final goal is to have this in a very decentralized fashion, similar to the sense of like how Ethereum currently is. So all these projects run independently of each other, they have different teams, they’re led by different people and things like that. And I think like in the long term future of what we want to do, which I don’t think is too far away, actually.

Me and Joe, we probably will still be in the Terra ecosystem. But we won’t be working as Terraform Labs members will be working as like a team member of a different project, much like how there are teams that work on Astroport, teams that are working on Star Dust and things like that. So in the end, we’ll still be in the community, we’ll still be in the ecosystem, but we will probably be running our own projects as well. 

Jose (41:59):

Very cool. Yeah, that’s awesome. Yeah, I mean, I’ve gone through everything I wanted to go through. I’m curious if there anything else that you want to touch on, or that you want to mention before we kind of head off?

Stanford (42:17):

Oh, no, I don’t think so. I think we covered a lot of interesting things today.

Jose (42:17):

Cool. And is there anywhere… Where should people go if they want to find out more about Mirror? If they want to kind of keep up with what’s going on? Where does most of the discussion happen?

Stanford (42:24):

Yes, I think there’s various community channels for Mirror. So there’s telegram, discord, I guess, on discord also. All the community channels are actually linked on the Mirror homepage, which is And then you can access any of these things out there. So there’s a lot of community discussion going on within Twitter as well. So we’re always welcome to a lot of feedback that users have. But we’ve been receiving continuous feedback from users. And it’s really impacted the development of Mirror. And we think it’s like a very big positive for the Terra ecosystem in general, to have such a strong community who’s always willing to give feedback, whether it’s a good feedback, or whether it’s something that they think is bad. 

Jose (43:03):

Yeah, absolutely. That’s awesome. Yeah, you’ve done an incredible job kind of building mirror. It’s amazing that we can get like exposure to so many traditional assets just using a completely decentralized stable coin. And I’m really excited to see how stuff like [inaudible 00:43:19] and then Nebula enable even more, sort of composability on top of that with stuff like leverage and indexes and having kind of a full fledged financial system on Terra. So really appreciate you coming on. Really appreciate all the work you do. And yeah, I look forward to doing this again soon.

Stanford (43:33):

Yeah, thanks for having us.

Joe (43:35):

Thank you. 

Show Notes: 

(00:00:00) – Introduction.

(00:02:36) – Guests’ backgrounds.  

(00:05:32) – A day in the life at Terraform Labs. 

(00:07:24) – Introduction to Mirror.

(00:09:06) – The long term vision for Mirror.

(00:10:03) – The decision process behind Mirror’s synthetics architecture. 

(00:12:32) – Biggest contributors to Mirror’s success.

(00:13:53) – Users getting exposure to Mirror assets vs. yield farming. 

(00:15:24) – Incentivising trading volume.

(00:17:41) – Mirror v1 vs. Mirror v2. 

(00:19:40) – Most exciting features of Mirror v2. 

(00:24:33) – Using concentrated liquidity on Astroport.

(00:26:53) – Strategies for inter-protocol composability.

(00:29:45) – Regulation and decentralizing Mirror.

(00:32:10) – Thoughts on scaling Terra. 

(00:36:28) – Funding for scaling Cosmos.

(00:38:10) – Most exciting upcoming projects.

(00:40:43) – Removing Terraform Labs from the equation.

(00:42:18) – How to participate in the Mirror community.