[Discussion] Launching New Markets

[tongnk]

There’s been a lot of discussion around launching new markets on Perpetual Protocol but one of the things that we need to start understanding is how do we decide which markets to launch?

One of the key items that we need to consider here is the potential trading volume - if there isn’t enough traded volume then this could lead to a drain in the insurance fund (if the index price vs market price deviates too greatly then the funding rate may increase significantly).

Some questions then arise:

• How do we determine if a market has enough potential volume?
• Is the token mainly traded on central exchanges and will those users use DEX exchanges?
• How do we quantify the potential trading volume?

I’m sure there are many more questions we would need to answer but I assume most of you will get the gist of it. The end goal of this discussion is for governance to try and start to formulate some form of framework that we can use to evaluate markets and then launch them as the community wishes!

[Lanre]

It’s quite hard a priori to work out how much volume any other coins we list but, perhaps as we get more historical data for the uncapped volume of the BTC/ETH pairs, we can analyse what % of DEX volumes our BTC and ETH pairs have captured and then extrapolate that to potential pairs to get an estimate of their volume. For example, say we know that our exchange consistently has 10% of the volume of ETH across DEXs, then we can assume that it’ll likely be a similar proportion for other coins.

I imagine our first users will be DeFi heavy users who tend to stay away from CeFi and have large sums on metamask etc. I think it’s unlikely we’ll be able to garner much of the CeFi trader audience at least at first.

One question I have: what is the lower bound on volume for a new pair before there’s a tangible risk of insurance fund drain?

[shumwhere]

Instead of having a governance system to vote yes/no on pairs, is it possible to build into the native staking protocol a way for PERP holders to also stake their PERP on pairs they want to see? In order to get listed(In guarded mode with OI cap), a pair needs to reach a minimum staked threshold, The PERP staked will then be locked for a period(Say 1 year? We can discuss length separately). During that time:

• Only stakers of the new pair are exposed to insurance fund shortfalls but only they receive the fees. What this means is that they bear 100% of the insurance fund risk where they will get fully liquidated of their staked PERP in order to keep their pair collateralized but in return they get 100% of the fees in that pair. If after getting liquidated it’s still not enough, that’s when governance kicks in to determine next steps.
• On top of getting 100% of the fees of their new pair, they still receive fees and staking rewards as a normal staker from all other pairs. This way it’s not an opportunity cost of having to pick one over the other and end up being locked for a year with less rewards but rather it’s a bonus action to earn more fees at the risk of taking on more personal risk vs just normal staking.
• Regular PERP stakers can therefor choose to move their PERP into any of the new pairs staking pool at any time but will be locked for 1 year minus whatever time has already elapsed from when the pair was created.
• As more people lock additional PERP into new pairs, their OI cap increases accordingly.

Example:

• Alice stakes 100 PERP and receives fees and rewards from ETH/USDC & BTC/USDC pairs as normal.
• Alice thinks XRP/USDC pair could be profitable due to trading volumes on CEXs and wants to see it in PERP Protocol. It needs 200 PERP in order to get listed in guarded mode with a 200 PERP value of OI cap. She signals her interest with some method that sets her 100 PERP aside for the XRP/USDC pair(until she decides to move it back to normal staking meaning she can still change her mind until the threshold is reached).
• During this time Alice still receives fees and rewards as normal. She is only signaling she’s willing to risk 100 of her PERP to bootstrap this pair but her PERP can get locked at any moment if it triggers.
• Bob also stakes 1000 PERP and receives fees and rewards as normal and sees that 100/200 PERP has been funded and he likes the idea as well so he moves 100 PERP over to the XRP/USDC pair triggering the event.
• The XRP/USDC pair is now live in guarded mode. Alice and Bob are each solely 50% responsible for any insurance fund shortfalls of this pair for 1 year and their PERP gets locked. During this year, they will each receive 50% of the fees collected as well as whatever fees and rewards are associated with normal staking.
• 6 months later the XRP/USDC pair turns out to be very profitable. Bob decided to move another 100 PERP to this pair. It gets locked for the remainder duration of 6 months.
• So from now on, Alice is now 33% responsible for the pool, Bob is 67% for the remainder of the 6 months. They receive fees from this pair according to this new ratio.
• Others can also join at any time for 1 year minus the amount of time that has already elapsed. E.g the 1 year lock is not at time of staking but rather from when the pair was created.
• After 1 year, governance decides what to do if the pair is unprofitable or risky. If the pair is profitable, governance also decides if we want to remove the cap and consolidate the pair under the normal staking. If so now everyone staking takes on insurance fund risk and get the fees along with ETH and BTC pairs.

This is just my crude idea, I’m sure it can be refined in many ways as I’m sure there’s a lot of stuff I didn’t consider and think about. But the overall idea makes sense no? Instead of having people come in and complain about adding pairs, they can put their money where their mouth is and bootstrap it themselves(assuming there’s an oracle for it.)

[tongnk]

@Lanre

It’s quite hard a priori to work out how much volume any other coins we list but, perhaps as we get more historical data for the uncapped volume of the BTC/ETH pairs, we can analyse what % of DEX volumes our BTC and ETH pairs have captured and then extrapolate that to potential pairs to get an estimate of their volume. For example, say we know that our exchange consistently has 10% of the volume of ETH across DEXs, then we can assume that it’ll likely be a similar proportion for other coins.

What stats would you want to see? We have some but probably need someone to take a look over the queries to make sure they are all correct

@Lanre

One question I have: what is the lower bound on volume for a new pair before there’s a tangible risk of insurance fund drain?

I’m not sure actually. Apparently the funding rate is calculated on closing of the position vs during so this is something we might have to consider

[tongnk]

Yes I really like this idea and we probably should explore it more - you’re effectively using staked perp as an indicator of interest and then the OI is capped based on the amount that is staked.

Given 1 year lockup is a long time I would almost say you’d want some benefit there as well.

Not to derail this train of thought as I think there’s massive value in this but I’m wondering if we can potentially shift our focus to more immediate term. So I am thinking maybe how do we:

1. Launch markets as governance
2. Then afterwards how do we create a framework so anyone can launch markets

I think your point describes #2 really nicely but not sure if it can be deeply thought through and then implemented in the short term.

However as I write this I’m wondering if there is something quick we can do for a simple lockup as almost an interim step 1.5 for launching markets

[shumwhere]

Yeah, I threw 1 year in as a placeholder as I know it would require a discussion to determine a fair length of time.

In terms of a short term solution, I don’t know how feasible or long it would take but what about using the Balancer 80/20 LP that currently exists that will presumably be significantly reduced once native staking goes live. Until a longer term solution can be created, PERP holders can stake natively or stake in the LP where they can use their BPT to stake in creating a new pair. In this case the opportunity cost will exist however, this also indirectly solves the reduced liquidity issues that will result from when the native staking goes live and everyone exits the BAL pool. I don’t know how the reward mechanism will work but the weekly BAL rewards in this pool will go towards the insurance fund of the new pair. You will need to fix your smart contract though because it currently doesn’t have the ability to claim BAL rewards.

[tongnk]

I’m just having a further think and potentially there are some interesting metrics that we can look at. Maybe some of these can spark some thoughts. I just took YFI as a potential topic of discussion and used Santiment (which is free and nice) to dig into some data.

Number of wallets + Distribution

Ideally there is a good distribution and what does token distribution look like across the chain. It’s all well and good if there are say 10,000 holders but if it’s all concentrated in 3 wallets then maybe not so great

In this one if I look at YFI token distribution on Santiment, it looks like there’s 5k wallets that hold < 0.1 YFI (or ~2.5k usd) and then the largest distribution is with ~160 wallets that hold most of it. Is this a good thing, or is this a bad thing? Will this mean that those ~14k wallets holding less than 0.1 YFI will want to leverage up?

Coin supply on exchange as a %

Can we derive a hypothesis here that the more tokens that exist on an exchange the less likely those tokens will trade?

Trading volume

How much is enough trading volume?

Social dominance

This is a santiment metric which is quite interesting. The description is basically:

Shows the share (or %) of the coin’s mentions on crypto-related social media, compared to a pool of 50+ of the most talked-about projects online.

[will_sheehan]

I wouldn’t overthink this, AAVE, YFI, UNI, SUSHI, LINK, and a few others have proven defi trader market fit.

But you can trade all of those on CeFi futures platforms, if you want to be bold do something super DeFi native. Some super interesting futures would be tokens like ESD and DPI, both of these have no CeFi derivs platforms but tons of spot liquidity on uniswap to handle liquidations

[almccann]

I really like @shumwhere idea of staking PERP or BPT to ‘vote’ for pairs but I worry about the impact long-term.

My concern would be the effective split in capital and insurance fund risk- it’s effectively then pools of pairs with different risks and fee incomes, and I would rather the risks and fees were shared amongst the whole platform. Ie I’d rather the higher volume pairs subsidise the lower pairs and so forth.

I also agree with @will_sheehan that simply copying CEX assets may be missing a big DEX opportunity in long-tail or index assets.

In short I think forecasting volumes is relevant, but also systemic risk and in turn DeFi/Uniswap user preferences and market opportunity (as opposed to CEX)

[tongnk]

@will_sheehan

I wouldn’t overthink this, AAVE, YFI, UNI, SUSHI, LINK, and a few others have proven defi trader market fit.

But you can trade all of those on CeFi futures platforms, if you want to be bold do something super DeFi native. Some super interesting futures would be tokens like ESD and DPI, both of these have no CeFi derivs platforms but tons of spot liquidity on uniswap to handle liquidations

I think this is a super good point BUT I would say we need to consider risks here:

• What happens if the token effectively dies?
• This is a perpetual market so how would we shut it down? When do we shut it down?

@almccann

I really like @shumwhere idea of staking PERP or BPT to ‘vote’ for pairs but I worry about the impact long-term.

My concern would be the effective split in capital and insurance fund risk- it’s effectively then pools of pairs with different risks and fee incomes, and I would rather the risks and fees were shared amongst the whole platform. Ie I’d rather the higher volume pairs subsidise the lower pairs and so forth.

I think the benefit of the split model would be the ability for individuals to more freely spin up riskier markets that are isolated from the current insurance market.

So the 2 markets right now I guess are shared so in that case we would have to be more mindful of risk? But then how do we determine said risk?

[almccann]

I can definitely see the benefits with the split approach… see what peps want to trade and go from there. Rather than completely open could be a series of juiced competitions and volume and OI wins.

Biggest issue even with OI caps which protects PERP somewhat is if users have a bad experience for whatever reason. Even if the UI says “this is tier 2 asset it’s risky” users will blame the product for mishaps.

[tongnk]

Though don’t you think that Nexus Mutual has already solved this design pattern? So people are used to seeing this? You could almost just not even let people press the button because it’s full. But I digress I guess.

So if we take @will_sheehan 's initial idea - how would we evaluate say ESD to see if we want to launch it as a market?

[shumwhere]

@tongnk

I think this is a super good point BUT I would say we need to consider risks here:

• What happens if the token effectively dies?
• This is a perpetual market so how would we shut it down? When do we shut it down?

I guess I’ll ask, since CeXs do shut down low volume pairs, how does the process work in PERPs case? Does PERP have some kind of global settlement process like Maker?

In any case, I originally offered up a joined solution but as mentioned it’s a longer term solution. The short term solution splits up capital and insurance fund risk but I think it’s fine because it only matters when that risk manifests and has been realized that you need to dip into the fund. And as time goes on and the fund grows, it becomes less of an issue that the capital is split. I would also argue it’s the PERP holders choice what they want to do with their token. Looking at PERP holders now, we already see not a lot of them stake PERP as it is on Balancer. Naturally some will be more comfortable staking on native and some will be more comfortable staking on BAL.

[tongnk]

Here’s a curveball for you.

Would we want to list XRP, LTC, BCH and other non ERC20 tokens?

XRP has a ridiculous number of active addresses - 45K

PLUS there seems to be a lot of outflow

But what happens if it gets delisted from everything due to SEC - will the token die? Would we still want to list it?

[shumwhere]

I only used XRP as an example but I don’t think it’s a good idea to list it at all. At least not right now when they are under a microscope. PERP is not decentralized enough to put itself in front of the crosshairs of regulators IMO.

[tongnk]

Well we wouldn’t make a decision there. I think any markets going forward have to be voted on in some mechanism by community as we start to decentralise more.

Question is how should we ask people to judge if we should or should not list a market

[bardur]

I also like @shumwhere idea of staking PERP on new trading pairs. This will give the PERP token more utility and demand.

I think it should be possible to spin up experimental markets that are isolated from the main trading pairs. Experimental markets should have their own isolated insurance fund and should not be backstopped by PERP token holders/stakers (or only by dedicated stakers as @shumwhere proposed). When the DAO thinks a market is mature, it can be included in the main insurance fund. When launching a new experimental market, the DAO can decide to bootstrap the isolated insurance fund with capital from the main insurance fund.

When we are talking about trading pairs, we should also talk about collateral types (long term). USDC is the best collateral right now, but this will change when regulators start looking at defi. I don’t think we should do a ESD/USDC pair, but rather talk about ESD as a supported collateral (way) down the road, e.g. BTC/ESD.

I’m also a proponent of native collateral types, e.g. USD/ETH and LTC/ETH. This would be fully resistant from regulators but is like v3+ of PERP.

[yenwen]

@shumwhere @bardur We are actually working with Delphi Digital on the model of creating new markets permissionlessly. They have created the Aave proposal to create money market permissionlessly and that’s a really thoughtful one. Here are links to their proposal:

Aave Proposal Walkthrough - Delphi Digital

Proposal: New Aave Token Architecture

Hopefully we can release the proposal in Jan.

[yenwen]

Yes, we can shutdown a market and settle every traders with a average price. Although the mechanism is there, it’s a last line of defense and we don’t want to tigger it often.

[yenwen]

I think trading XRP doesn’t increase our exposure to security risk because we don’t trade XRP directly or in any form.
But, the derivatives itself is a huge risk already.