zkLend
Published in

zkLend

zkLend X Stork Oracle AMA Recap 07/07/2022

On the third episode in the ZEND&FRIENDS Series: Infrastructure, we had the pleasure of having Vlad, Co-Founder of Stork Oracle and Abe, Co-Founder of Dexterity Capital join Jane on zkLend’s Twitter Space for a new AMA. Price feeds, 1st/3rd party data, node operators, and all things oracles were covered.

Full Recording

Listen to the full recording on our Spotify now.

AMA TLDR

The ‘Why’ of Stork [02:27]

The first real interest came from the Dexterity research team, and over the last year, it became very clear that StarkNet was the most promising L2 solution out there. Both because of validity rollups and in general, they have a very technically strong team. That’s what attracted us, and oracles are a public infrastructure that everybody needs. So Dexterity as a trading firm immediately understood the value of accurate prices, and that’s how we landed on StarkNet building an oracle.

Stork — Dexterity Relationship [05:15]

Dexterity is and first and foremost an investor in Stork. You can think about it as somebody who is incubating Stork. I’m in their office a lot of the time, I work very closely with their engineering team, they’re actually helping us with some of the deeper pricing data like liquidity, they have a ton of experience.

Overview: Oracles [07:05]

Let’s take the example of a money market protocol in the DeFi space, you want to take an asset as collateral and if the value of the collateral falls below a certain value you need to be able to sell it to cover your losses. How do you know what is the market value of your collateral? An oracle is somebody that provides off-chain information on-chain, doesn’t have to be just pricing or crypto-assets, can be the weather, for example, basically anything that happens off-chain that we want on-chain.

Intuitively the easiest thing to do is to write a script and scrape some data and add it to your smart contract. What if your machine goes down? To solve this you will need multiple machines. Another step is to get your data from several APIs, make sure your data is robust and don’t rely on one centralised feed, multiple exchanges for example. With all of this, you arrive at what is a modern oracle implementation. Decentralised from both a hardware and data feed perspective.

You can also choose what type of data you want on your oracle, first-party or third party. First-party data can be, for example, an exchange that has trading that happens on their platform. Third-party data comes from someone that scrapes data directly from an exchange (first party).

Now you need to aggregate all this data you’ve collected. You can do something simple like a median, or something more sophisticated like a median and discard the extreme values. You have to consider things like a single participant in your data feeds can skew the prices too much for example.

We believe that first-party data is preferred because it is generally more robust data and more decentralised when compared to third-party data.

Common Risks [12:28]

If you have an implementation of an oracle that relies on a single source, and users of the oracle have more assets that are at risk than what it takes to move the price of an oracle. It sets up this situation where it is much cheaper to move the price of the oracle and then to act on the cascade of the price change. A way to help prevent this is by using a time-weighted average of prices since you don’t want to be too susceptible to a price movement.

At the same time, you want to respond quickly to a price change. If you’re using a time-weighted average you might not pick up on smaller changes which can affect if you’re a perpetual exchange, for example, you might miss on some liquidations and lose money because of this. So it’s very important to find the balance between these two.

We believe that we are approaching as much from a performance point of view as what people expect from traditional finance. So one of the main reasons for building on StarkNet is it drastically reduces the cost of storing more data. This allows for much more price feeds and the other part, is you can simply just store more metadata such as liquidity, for example, which can be very useful to know.

Benefits of Building an Oracle on StarkNet [19:17]

Ethereum provides this incredible level of security that StarkNet can leverage with much faster and cheaper data transfer. This enables a lot of very interesting ideas while maintaining stability. With Ethereum it is a little bit unfair, I imagine that any medium to high frequency DeFi project will inevitably migrate from Ethereum since we are discovering as a community there is currently no use case for high frequency. In this sense, StarkNet will enable many use cases for mid to high frequency.

Over the last year, it became very clear that StarkNet was the most promising L2 solution out there. Both because of validity rollups and in general, they have a very technically strong team. — Vlad

Comparing StarkNet to Other L2s [21:13]

I think it might be worth asking someone that’s a firm believer in the other L2s the same question. We have done a ton of our research, and a lot of these optimistic rollups might be great for some use cases, but I don’t think that the community will converge on those solutions. I am more interested in zero-knowledge, validity rollups and how they work. I think StarkNet is ahead by far technically.

Node Operators [24:03]

The two extremes that you can think about are: i) anyone can join, you run a node, you’re a participant and then there is some algorithm that weighs your contributions. A fully decentralised model, and maybe you get some sort of reward. ii) The other extreme is you go through some folks you know who have first-party pricing information and you use their feeds to participate as an oracle.

The risk with the fully decentralised model comes before you get the scale and you have a lot of distinct parties is that the asymmetry is such that it may be cheaper to stir your price. The only meaningful way to start is with a group of reputable oracles and ecosystem trust, and ideally with 1st party data.

That tends to be the profile of the folks that will be in our network. They’re brand names, and they have access to first-party data. We tend to find Oracles/Nodes rather than the other way around, and we want to make sure when you see those names on our website you’ll say ‘for now I trust those guys.’ And maybe in a year, we’ll add more names and transition into a purely decentralised model.

Use Cases [27:23]

Who needs an oracle? Anyone that uses exogenous data on-chain. Where we think Stork is the most helpful, and part of that is our connection with Dexterity is that we understand DeFi better than a lot of people out there, both as users and oracle providers. Where Stork is having a lot of traction today, where frequent price updates on a wide range of assets are very valuable. Where knowing if a price moves a minute earlier makes a massive difference, let alone a second earlier, for example, petrol swap exchanges. Given how levered the products offered are, being ahead of price movements or knowing exactly when they happen makes a difference whether you are liquid or non-liquid, profit or unprofitable, and that’s where we see a lot of traction for Stork.

Another use case for oracles is that we don’t talk as much, if you need to roll a fair dice, generating random numbers is something that has to happen off-chain, you cannot have randomness on-chain. There are a ton of use cases for oracles, and today we are only focusing on a small subset of them.

About Stork Oracle

Stork is designed from scratch to be native to StarkNet, StarkEx, and the upcoming world of StarkNet L3s.

Source: https://www.stork.network/

Touting regular price updates, on-chain validation, and deeper DeFi metadata such as asset liquidity, Stork leverages the benefits of validity-rollups to build critical pricing infrastructure for the ecosystem.

Their target audience includes money-markets, perpetuals and options protocols, and other DeFi applications that require external pricing data on both crypto and non-crypto assets.

As of June, Stork Oracle is live on StarkNet Goerli and supports multiple assets across various exchanges, as well as relying on its publishers proprietary first-party data to provide more diversified data feeds.

See demo: https://bit.ly/3yWYE0V

StarkEx and mainnet launches are planned for Q3 and Q4, respectively, along with an announcement of participating price publishers.

Stork Roadmap

The team was founded by Vlad Shulman in close partnership with Dexterity Capital. Vlad was the Co-Founder & CTO of Retain.ai, the first Customer Insights Engine (CIE) ($28M raised), while Abraham is the Co-Founder of Dexterity Capital, an algo high-frequency trading firm focused on digital assets. Dexterity is an investor in Stork and one of its price publishers.

This section was adapted from our original thread, here.

About zkLend

zkLend is an L2 money-market protocol built on StarkNet, combining zk-rollup scalability, superior transaction speed, and cost-savings with Ethereum’s security. The protocol offers a dual solution: a permissioned and compliance-focused solution for institutional clients, and a permissionless service for DeFi users — all without sacrificing decentralisation.

Website | Twitter | Telegram | Discord | Spotify

--

--

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store
zkLend

zkLend

zkLend is an L2 money-market protocol built on StarkNet, combining zk-rollup scalability, superior transaction speed, and cost-savings with Ethereum’s security.