Crypto AMA with Synthetix (6.28.19)

Guests:

  • Kain Warwick

  • Justin Moses

Moderator:


Moderator:

Let’s give a warm welcome to Kain (founder) and Justin (CTO) of Synthetix!

As a reminder for everyone participating, please keep the discussion respectful at all times.

@kain & @justin - could you guys start off by giving us briefs bio touching on your background as well as how you got started in crypto? And then a short overview of your project a brief update on your progress to date? We’ll then be off to the races with questions.

Kain:

Hey Spencer thanks for inviting us!

Justin:

Agreed - it's great to be here

Kain:

Hi I'm Kain, founder of Synthetix. I have been in crypto since 2014. I launched a payment gateway in Australia around then that works with a number of large exchanges including Binance to enable cash deposits in 1300 locations in Australia.

Moderator:

(ps: yes, it's 5am in Australia for Kain, so we're grateful he could make it on 🙂)

Kain:

That's OK not a huge fan sleep at the best of times...

Justin:

Like Kain I’m originally from Australia, having moved to NYC in 2010. I’ve been working as a software engineer since the Dotcom days, most recently as Director of Engineering at MongoDB before joining Synthetix.

Justin:

Kain and I have been friends since high school and I found the work the team were doing at Synthetix (it was Havven back then) really fascinating. I’d only been watching crypto from the sidelines but as I started peeling back the layers I realized just how much potential there was, particularly with the web3 movement and decentralization.

Justin:

5am on Saturday in fact

Kain:

I started synthetix (then Havven) looking for a way to arb the major price premiums in small markets like Australia and Korea on crypto back in late 2016/2017. My view was that regulated stablecoins would not be a thing in the short term so you needed a decentralised stablecoin to be able to solve the problem. Obviously Maker hadn't launched then, nor had the stablecoin boom happened.

Justin:

It's a balmy summer's afternoon in NYC though :)

Kain:

We have since pivotted away from a pure stablecoin model to a synethtic asset trading platform.

Kain:

And here we are 🙂

Participant:

Aside from adjusting the collateralization ratio, what other mechanisms are you considering to help sUSD maintain its peg?

Moderator:

@kain @justin - feel free to get to this in a sec if you still have more intros

Kain:

We think that the C Ratio will be quite effective, but beyond that we are looking at a liquidation mechanism.

Kain:

All good on the inros, if someone wants more background feel free to ask

Moderator:

Okay great—let's open up it for questions!

Moderator:

Let's set the table a bit for everyone. Most people here are familiar with synthetic asset platforms like Maker and UMA. Can you guys explain the core mechanics of Synthetix and it's key differences with those two platforms?

Kain:

The main difference between Synthetix and other platforms is that we used a pooled collateral mechanism.

Kain:

That has trade offs, the negative side is that there is more risk to SNX holders putting collateral into the system.

Kain:

The positive side is that you have a P2C tradinbg model similar to Uniswap which means no liquidity issues and a much better end user experience.

Kain:

Also removes the need for early network efforts since you don't need a counterpary waiting for everyone to come trade (sorry market makers)...

Participant:

Have you considered diversifying the collateral pool in the future?

Justin:

We've been looking into ETH-backed collateral, but don't have a timeline for it just yet

Kain:

Right now supply of Synths is not the main issue we have more than sufficient supply.

Participant:

How do you see the oracle mechanism evolving over time?

Justin:

We've been working with Chainlnk, a decentralized oracle platform that recently launched on mainnet, and as decentralization maximalists, we are eager to move away from a centralized version

Kain:

Ideally we see it evolving to be not our problem to solve.

Kain:

As you might imagine a number of teams working on Oracles have reached out this week, and there is a lot of activity in this area going on.

Justin:

I'm also somewhat bullish on a on-chain oracle, like what Hayden from Uniswap has been proposing...

Participant:

When do you expect to add non-crypto assets to your exchange? Synthetic stocks, commodities, etc.

Kain:

We have gold and silver already, but they are pretty tame. We are doing a regulatory review right now on the potential issues of equity Synths. We are confident it will not be an issue but just want to ensure we have everything in place.

Justin:

In terms of technical implementation, we're targeting the next few months

Moderator:

I'm also going to be adding some questions that people posted from Slido (marked with a *)

* What is the point of pooled counterparties? Why not silo debt like Maker?

Kain:

That is kind of the critical question, the main reason is that it enables Synthetix.exchange to exist. Which is the real value prop of the system.

Participant:

Hi @kain & @justin thanks for doing this! Curious if you could provide some color on what happened with the recent Oracle attack on your network? How was the system attacked and could it happen again?

Kain:

Obviously this introduces a lot more risk, but it means that you can trade between sUSD and sBTC without the need for a counterparty, since the liquidity is shared between all assets.

Participant:

Can you briefly explain how inverse synths work, how trading them can differ from long synths (reset period) and the potential benefit they can bring to the platform by netting down some of the long only exposure?

Justin:

It wasn't actually an attack as such. It was a cascading failure where one API went down and another feed started sending an incorrect price which the system didn't handle. Then a bot contract that was targeting major shifts exploited the issue multiple times

Kain:

They are similar to an inverse ETF, so you get a positive movement for each negative price movement in the asset it tracks, for example sBTC & iBTC.

Kain:

We have limits to ensure that the profit/loss can't exceed certain thresholds, typically 200%. They are actually a stop gap solution until we have the full synthetic positions contracts in place which will be much more like Bitmex/Deribit.

Justin:

And we have definitely taken steps to ensure it doesn't happen again. Longer term though, we will be moving to a decentralized oracle, which is in line with our larger vision of decentralizing the entire platform

Moderator:

* Currently, collateralization is at 500% with plans to move to 700%+. Isn't this highly capital inefficient? Why not do the Maker model of 150 + liquidation?

Participant:

Makes sense thanks!

Kain:

It is definitely capital inefficient, but because the asset we use as collateral SNX is the only useful as collateral (as opposed to say ETH) there is less opportunity cost. Though of course you could deploy that capital elswhere, but historically within staking systems people tend not to view opportunity costs as highly as they ought to.

Kain:

The main reason for not having 150% is the lack of liquidity in SNX, like Maker there needs to be a specific requirement per asset and we believe that SNX will need to be higher. With the launch of MCD it will be really interesting to see what the threshold are that are set for assets like REP as one example.

Participant:

Another question from me: What is the benefit of only having penalties for being undercollateralized at the time fees are claimed (which are paid out on a weekly basis)? According to the recent Block analysis: "The system could viably be severely undercollateralized for weeks at a time, with issuers simply topping up their collateral on a periodic basis."

Kain:

This is true, which is why we recently voted as a community to lower the threshold to two weeks for claims. But the argument remains that there is still at least a week lag between SNX holder responses to market changes. That said with a C Ratio of 750% that should be sufficient time.

Kain:

In terms of benefits, it is more a limitation of the way that rewards are calculated on chain.

Moderator:

@kain perhaps I'm missing something but is there a purpose to not having the collateral requirements enforced all the time? (i.e. reducing complexity)

Kain:

It would require observing everyone's debt status at all times, which is not possible on chain.

Kain:

So we set a block where we measure it periodically.

Kain: That said recently one of our community members has proposed an improvement where we randomise the windows to not allow SNX holders to game this.

Moderator:

I was literally just going to suggest this!

Kain:

Haha, great minds!

Kain:

There is a usability trade off though, but I think it is worth it.

Kain:

Definintely

Participant:

What is the process for listing a certain asset (sBNB for example), and are there any limits on traders taking positions on individual assets? Theoretically, if there are large outstanding positions on sBNB, and BNB rallied enough, would the system not collapse even if the price of SNX was unchanged?

Kain:

We typically look for liquid assets that are not listed on other platforms like Bitmex. And yes that is definitely a risk, which is why we have the inverse option there as well.

Participant:

What was the design choice behind not having a liquidation process or penalty when collateral falls below a certain threshold?

Kain:

That said BTC and ETH have been rallying hard and in spite of 20% of the debt being in those assets, people take profit etc so the overall system debt has not increased that much and the network is currently around 600%

Justin:

We are planning to implement this in the future. In the meantime we feel like the high cratio is sufficient enough given the market conditions.

Kain:

You can check the state here btw https://dashboard.synthetix.io/

Kain:

We were able to withstand a draw down of the price by 95% over the bear market from the launch of sUSD in July 2018, so we think it is not hugely urgent.

Justin:

Makes sense. Thanks!

Participant:

can y'all touch a bit on the genesis of the project (as Havven) and how you decided to make the switch to Synthetix / what you would've done differently at the beginning if you could go back?

Kain:

We have three weeks here right?

Participant:

i've got nothing but time 🙃

Kain:

lol

Justin:

Definitely one thing we could have done differently is how we distributed the tokens in the token sale. If we could do it again, we'd target holders who were bought into the proposed system and would be more active in our community

Kain:

The main reason for the switch was as mentioned earlier, I didn't believe that regulated stablecoins would be a thing. Obviously that was very very wrong. So we needed to find a way to deliver utility to users of the platform. We launched multicurrency in December 2018, but even that was not sufficient. So when we launched sBTC we realised that volatility of assets was a big driver of usage and leaned into that with the trading platform.

Kain:

This has improved over time organically, but yeah, was one of our biggest issues.

Moderator:

Defending the peg of sUSD is critically important for the system to stay in tact. Cool to see a liquidation process will be set up in the future—but what levers do you guys have now to defend the peg?

Also from Slido: *Is Synthetix.Exchange censorship resistant?

Kain:

As a community we voted to raise the c ratio, and I personally think that will be very effective.

Kain:

The liquidation process would need to kick in very early (~450% right now) to have a meaningful impact on the peg.

Kain:

Really solving the demand side is probably the most important thing right now, and we are definitely seeing this start to happen. The trading volume on the exchange has been 1-2m per day for the last several weeks.

Justin:

Strictly no as synthetix.exchange is a dApp that we've written and published. However, the contracts themselves are accessible everywhere Ethereum is so in that sense the Synthetix system itself is censorship resistant. And given the quality of Etherescan these days to write to contracts with Metamask the dApp is almost replaceable with just the contracts (well, except maybe the UX :) )

Participant:

On the topic of demand, what steps are you guys taking to see that sUSD attracts the same level of adoption/integration as DAI?

Kain:

The main use case for sUSD is trading on synthetix.exchange at the moment, so we have been working hard to improve the user experience there. Also looking at adding more assets, XTZ, MKR, TRX (equities soon hopefully).

Kain:

We have also been having trading competitions internally (in our community) for the last month

Kain:

We are also working on integrations with other DeFi projects. Something that was not possible last year due to the mechanism in sUSD when it launched. Which has now been removed (fee collection)

Moderator:

* what does risk management of the debt pool look like? Is there a team/process that addresses asset selection and asset mix

Kain:

We look at a number of criteria to determine whether an asset can be added,. The main risks are with crypto assets, so we look at average daily volatility and liquidity and other metrics.

Kain:

We will definitely be sticking with the top 25 cryptoassets for the foreseeable future.

Moderator:

Can you tell us about your team and community?

Kain:

We started with a community in TG, in early 2018 we had the largest tg in crypto 160k people across two channels. As you can imagine it was an utter shitshow.

Kain:

We made the decision to migrate to Discord about 6 months ago and we had about 50-100 people follow us there from telegram. It has since built up to over 3k iirc and we have an extremely active and vibrant community of users who think deeply about the project.

Kain:

moving to discord was probably one of the best things we did last year.

Justin:

Our team is split between Sydney (the majority) and NYC. Engineering is managed across the two timezones (sometimes painfully), with myself here in the states and four engineers in Sydney. We also have a community manager in Sydney and a VP Partnerships (Jordan) who works here with me in NYC. He just recently moved to the states; I'm teaching him everything I know ;)

Justin:

https://discordapp.com/invite/AEdUHzt

Participant:

How much fees have stakers earned to date? Do you have projections for how much they can reasonably expect to earn over time?

Moderator:

Wow, that's an outrageous amount of people to have in a Telegram channel...

Kain:

it was the worst...

Kain:

We have two mechanisms here. Fees from exchanges and staking rewards.

Kain:

The exchange has generated around 150k in fees so far, and some of our users have calculated the yield on fees alone to be 2-5%

Kain:

For the staking rewards (paid in SNX) the current inflation rate this year is 75%, so the yield is a little higher as only around 70% of SNX holders are staking.

Participant:

And those issuance rewards are locked for a period of time?

Kain:

Yeah the staking rewards are locked for 1 year and then vest on the issuance date, so from year two there will be a weekly vesting in the network of around 1.4m tokens.

Kain:

But you can stake the rewards immediately you just can't transfer them.

Moderator:

What was the structure of the token sale? Any plans to do an additional sale to target productive holders?

Kain:

We are talking to some funds at the moment to allocate tokens out of our treasury, but there is no formal plan for another sale.

Participant:

Have you considered working with any layer 2 solutions?

Kain:

We are actively looking for L2 solutions, so if you know any that are live now we would love to meet them.

Moderator:

Interesting, why's that? I'd be curious to hear what specific requirements you have for an L2 platform.

Moderator:

15-minute warning folks! Get your last-minute questions in

Justin:

We are looking at an L2 solution for the oracle specifically. As you can imagine, our current oracle requires a lot of ETH for gas and could update much faster if on L2

We're also looking at running some of our exchange on L2 as well.

Kain:

Interms of specific requirements, existing is a good property to start.

Moderator:

Haha tell me about it. Being live in production is so underrated these days 😉

Kain:

lol

Kain:

BTW someone was asking about synths listed, https://www.synthetix.io/tokens

Participant:

70% is quite a high staking participation rate. Can you get into how you engaged your community to get involved with staking?

Kain:

It sort of speaks to my earlier point, if you are holding SNX then you are significnatly incentivised to stake through the current staking rewards, but also you are only really holding it is you think the mechanism and the system can work.

Kain:

I think people (including myself) really got this wrong with fixed supply tokens.

Kain:

Personally I say all the blame lies with Fabian 😉

Kain:

power of defaults

Moderator:

Heh. Alright guys, looks like we're running up on time. @kain @justin what's the best way to stay up to date on all things Synthetix? And what's the best way for folks to get in touch?

Kain:

Thanks so much, best bet is in discord for sure, we also have a mailing list.

Kain:

https://discordapp.com/invite/AEdUHzt

Kain:

Also twitter @synthetix_io

Moderator:

Awesome. Thanks for coming on guys!!

Kain:

our pleasure - thanks for having us on

Kain:

https://blog.synthetix.io/

Kain:

is another good place for info.

Kain:

And of course check out our dapps, mintr.synthetix.io and synthetix.exchange

Kain:

Thanks Spencer!

Participant:

Thank you, Spencer & Synthetix!

Moderator:

Good stuff!

Moderator:

Crypto AMA: thanks for tuning in and asking great questions! Stay tuned for an announcement about our next AMA