Stimulating Long-Tail markets liquidity

RealVovochka · July 25, 2024, 9:15am

I am referring to this grant dYdX Grant: Long-Tail Market Making on dYdX

It’s insane $2.5k per market. Raven Dao is paid ~70k a month

Most of these markets are generating less in fees than paid to a market maker.

If you check for some of the markets like 60% volume is between Xenophon Labs deployed vaults($50k for service and $200k LP) and Raven DAO.

In my opinion it’s insane overpayment

Lorony · July 25, 2024, 7:53pm

Man, i don’t even understand why new niche markets are constantly being added when there is not much interest from traders. Doesn’t that just burn money? Do these markets really attract (new) users?

The low volumes are probably also related to big spreads and lacking liquidity, which you want to improve, yes. But even with deep liquidity and no spread I doubt there will be much „real“ volume for a lot of markets because they have a small community.

I’d suggest to do a poll/research amongst dydx users… like why do ppl use dydx in the first place? How many ppl come here for tokens like BODEN, MOG, MOTHER… ?
Big reason for me was decentralization/no KYC etc. for example. What should we change or focus on instead.
Since we are always generous with treasury, we could also offer a small reward as an incentive to get as many people as possible to participate in a poll😂

shuo · July 26, 2024, 2:45am

It is a worthwhile exploration on add Long-Tail markets, but if we find it to be incorrect, it may be better to stop going on.

RealVovochka · July 26, 2024, 10:03am

The purpose of my post is not to stop the financing of long tail markets, but to discuss specifically its effectiveness and start a community discussion on this topic, after all there is already some data that can be analyzed

RealVovochka · August 5, 2024, 2:52pm

Unfortunately the forum has become a completely useless place, we spend $80k a month on DEP and in 10 days, no one has taken the time to comment on an active grant.
Sick.

LeonoorsCryptoman · August 5, 2024, 6:42pm

When a grant is given, nothing needs to be done anymore since execution is always flawless >.<

RealVovochka · August 6, 2024, 3:30pm

13 hours ago additional $500k was transferred to Xenophon Labs. And RavenDAO got 161k.
Can we finally see any data showing how this money is working for the growth of the project?
@carlbergman @RoboMcGobo @max-holloway @Carlos_Raven

RealVovochka · August 6, 2024, 5:29pm

Why these look very inefficient to me. I use IMX-USD as example. This market is live for exactly 5 month. And the total Net Fees for this market is $1672 (one thousand six hundred seventy two).
$329 in July while we are paying $2500 per month. And it’s just one market which is not the worst

cryptoplaza · August 6, 2024, 5:57pm

I think it’s an interesting conversation, and I believe the real issue lies in the architecture used to generate these markets. It’s important to highlight that the great success of the long tail market, as seen with Uniswap, allowed for market creation at virtually zero cost and without the need for market makers. The order book structure, in my opinion, implies costs that make it unfeasible to have extremely small markets. These markets also have a much higher capital cost, being a derivatives market and requiring sufficient margins to sustain positions. The possibility of not allowing collaterals with the underlying asset seems practically impossible in these types of markets. I think the strategy of paying market makers to start markets would only make sense in potentially large markets. I don’t think it’s a viable strategy for the long tail. I believe that enabling vaults for automated market making could open the door to creating this long tail. I completely agree that the forum does not seem to be a tool where there is interest in discussing these kinds of challenges openly, as the teams or providers executing them are not participating.

carlbergman · August 7, 2024, 4:27am

Hey @RealVovochka,

Thanks for bringing this up – we’re actually in the middle of revisiting our approach. I’ll share some more context below.

First, why bootstrap liquidity on dYdX? Liquidity is the biggest priority for any trading venue. Without liquidity, users trade with poor execution, or worse, don’t trade at all. The user experience is entirely dependent on liquidity. No liquidity, no users.

But why pay for liquidity on long-tail markets? Higher volatility, fewer hedging options, and adverse selection make these markets much riskier for market makers. Since there’s no incentive for market makers to provide liquidity, orderbooks are left practically empty . This hurts the dYdX brand and reputation. What do you think happens when a user onboards to trade a newly listed long-tail and finds an empty orderbook with +20% spreads? I think there’s a good chance they leave and never come back.

That’s where we come in. To improve the user experience, and make dYdX a more competitive venue, the DEP has engaged two methods of liquidity support for markets in need: Dedicated Market Makers (DMM) and Vaults.

The DMM strategy employs external market making firms, including Raven and Velar Technologies today, to provide their own capital and trading algorithms. These firms take on the responsibilities and risks in return for monthly compensation. The Vaults strategy, on the other hand, includes depositing our own capital into the new AMM-style liquidity provisioning product. Vault deposits are not risk free - they can lose money in certain situations like excessive one-way directional flow. Funding differences aside, the two strategies yield nearly the same results in terms of improving top of book liquidity and tightening the spreads.

To your point on costs, the DMM strategy doesn’t scale very well. With each new market listed, the DEP increases its costs of liquidity support. What happens if 200 more markets get listed next week?

This is why our focus is shifting towards vaults. Vaults are risky, but deposits could also be recovered in full – meaning we spent no money on liquidity. As of today, the DEP has deployed $248k across 20 vaults. The equity across all the vaults currently sits at ~$246k, suggesting a loss of about $2k. Since we started this initiative in early June, withdrawing all of our capital today would mean we spent only $2k for two months of liquidity on 20 markets. At times, this number has been positive – it fluctuates based on activity.

The latest release (v5.2.0) includes support for up to 150 live vaults, an increase from the previously supported 20. With this upgrade, our plan is now for the DEP to transition most of its existing markets supported via DMMs to Vaults. The goal is to reduce our immediate costs while maintaining competitive liquidity across all markets. Eventually, the LP Vaults product will become accessible to external deposits, at which point we hope to see more capital flow in so we can withdraw our original deposits. We don’t expect the DEP to keep bootstrapping liquidity long term.

Finally, why are we paying for markets with little to no trading activity? Any live market should include competitive liquidity to maintain a strong reputation and guarantee a good user experience. What if a new user onboards tomorrow with the intention of trading IMX? We still face the same problem of reputational damage when dealing with live markets. Instead, I think an open question for the community to discuss is whether certain markets should be delisted given their current lack of demand. For example, would it make sense to delist existing markets that have 0 trading volume? As a reminder, the community can delist markets via governance.

cryptoplaza · August 7, 2024, 8:13am

Interesting information, and thanks for sharing the explanation of the strategic decisions. To contribute, I would comment that if the strategy is permissionless listing, we should understand that there will be markets without sufficient liquidity, and this should not be seen as a reputational issue. On Uniswap, there are pairs with extreme illiquidity that are continuously listed, and it is not perceived as a reputational problem for the protocol. The protocol’s measures in this regard are to warn in the interface about these slippages to try to prevent a user from incurring them when we have a situation of limited liquidity. Although we are not yet in that phase, I believe the philosophy should be the same, even now that there is a prior vote that is approved in most cases. I think those measures to prevent such trades and even a filter within the dYdX interface to not show illiquid markets or at least separate them would be a better measure than delisting them or having them become an unsustainable cost source.

As I previously suggested, I believe one of the most interesting ways to achieve liquidity in each market is to make a cash & carry strategy available to users. In general, the complexity of liquidity is providing a counterparty to short positions, especially for highly volatile tokens. When we talk about competitiveness, we certainly talk about slippage, but I believe the funding rate is one of the main costs to consider to be competitive. The cash & carry strategy directly addresses these two critical factors. Another different strategy is to have a fund that covers traders’ losses, which historically is what leverage produces and could be captured by dYdX instead of being given to the market. This should be a more general Vault of the project, but in any case, I believe it is something so strategic that outsourcing this type of activity is really a bad decision

RealVovochka · August 7, 2024, 8:29am

Hey @carlbergman ,

Thanks for your reply.

I don’t have any particular complaints about Vaults at the moment, apart from some transparency issues. It would be helpful to publish:

A list of active Vaults
Their trading statistics
Any overlap in markets between DMM and Vaults
The percentage of each market’s volume accounted for by DMMs and Vaults

I personally found a list of active Vaults through the API, but not all community members can do this.
Why are Xenophon Labs suddenly experts in market making? Maybe they are, but there’s no public information about it.

Regarding DMM:

Which markets and for how much time DEP pays for
Whether there has been an increase in volume in these markets (from regular users)
How many new users started trading these pairs
The requirements for companies providing such services
The origin of the $2.5K per market monthly price
Why Velar and RavenDAO were chosen (You forgot to mention COI in the blog, as always)
Why there was no open tender to choose such market makers

The community needs data.

Yes, I agree that liquidity is crucial for new markets and a great user experience. But the question is at what price. I’m sure “what if” scenarios are not a basis for business decisions (like your example with IMX), and in this case, the spread is more of a decorative indicator; the depth of liquidity is very small anyway.

Most of the new markets lately have been proposed by one user, and the current governance features mean validators will vote in favor without even examining them. Is that a reason to pay $2.5K a month to maintain liquidity right away? I think there needs to be a real indicator of users’ willingness to trade.

I’m all for DMM if dydx can create a new market faster than its competitors. If some hot new coin had TGE and we can start offering a trading pair immediately, it makes total sense to use DMM. But now this is not possible; the voting period is quite long, etc.

Paying for liquidity in half-dead markets with no sign that users want to trade them is money burn. Who said it is reputational damage? Totally agree with @cryptoplaza on this topic. And who said that the main competitive advantage of dydx is a high number of markets?

I never suggested withdrawing liquidity from Vaults, but I would like an explanation of why another $500K was sent to Xenophon without any prior discussion of the intermediate results of this experiment with the community.

As for delisting the markets,

Who is going to do that and propose such proposals? Don’t tell me that you will pay Chaos Labs again, as it was with the grant where you paid $900 for a report on a new market that took 30 minutes max to create.

I’d be happy to continue the discussion on this topic if you provide more specifics and respond in a reasonable timeframe.

max-holloway · August 7, 2024, 2:17pm

Hey man, long time no correspondence . Hope you’ve been well.

Why are Xenophon Labs suddenly experts in market making? Maybe they are, but there’s no public information about it.

Tell me how you really feel . I wouldn’t describe us as experts! But Xenophon Labs has done some research on market-making related stuff in the past; analyzing MM on uniswap, looking at dYdX HFT-related trading efficacy, and thinking through market maker experience w.r.t. MEV on dYdX chain. But honestly we’re not experts, just willing to put in the work to analyze the vaults performance. And at the end of the day the vault performance isn’t that complicated to measure: number go up?.

I don’t have any particular complaints about Vaults at the moment, apart from some transparency issues.

Hmm, it’s all on-chain, so people can take a look at it if they’re interested.

Would you mind sharing the query you used to find the vaults list via API? That would help anyone who’s interested in finding the vaults addresses, and they can then look at the vault performance via https://dydx.trade/portfolio/overview?address=<VAULT_ADDRESS>.

I don’t have an aggregated dashboard for vault performance, and I don’t intend to make one. However, I still intend to release research on efficacy of vaults, which will obviously have the performance numbers.

RealVovochka · August 7, 2024, 3:18pm

Thank you for your reply. I am doing great.

Correct me if I am wrong because there is not much detailed information about how Vaults operate. As I understand by analogy with market making bots , there is a strategy with a certain set of parameters, in this case what is the role of Xenophon?

To define the Vaults strategy? Or just do money management, collect statistics and evaluate efficiency and safety, if the latter then my question about Xenophon’s competence is cancelled. If the former, then I would like to see some credentials that Xenophon has successfully (profitably) engaged in market making business.

As for API, you can use the following request

grpcurl -d '{}' dydx-grpc.publicnode.com:443 dydxprotocol.vault.Query/AllVaults

And the API response where owner is the address of the Vault as I understand

{
  "vaultId": {
    "type": "VAULT_TYPE_CLOB",
    "number": 10
  },
  "subaccountId": {
    "owner": "dydx165fla2xss0wrg8rr04x3e46mneat4fckrcf429"
  },
  "equity": "AgIpHwoU",
  "inventory": "AwL68IA=",
  "totalShares": {
    "numShares": "AgJUC+QA"
  }
},

grpcurl can be installed using brew

But as I wrote in the previous post I have almost no questions about Vaults, but a lot of questions about DMM. But thanks for the comments in this thread

RealVovochka · August 7, 2024, 9:56pm

For those who is interested

Links to vaults portfolio

CipherLabs · August 8, 2024, 3:34am

There are straightforward solutions to many of these issues. For instance, implementing autonomous daily trading competitions for each new market listed, running for three months post-listing, seems like an obvious way to bootstrap liquidity and encourage trading activity. Numerous simple solutions like this could be easily implemented. However, it appears that self-interests have always been more prominent within DYDX, hindering such initiatives.

carlbergman · August 8, 2024, 9:04am

All good questions. I’ll share brief responses below, but keep in mind this is all very much in motion at the moment. We’ll be sharing a longer update on everything done at a later date.

Sharing two quick dashboards that can be used for live tracking :

Vaults: https://vaults.dydxgrants.com/ (hacked this together very quickly, prone to crashing if reloaded too often)
DMMs: Grafana

Responses:

Active vaults and DMMs can be found in the dashboards. DMM markets are being updated and transitioned out at the moment. DMMs started in March, Vaults in June.
We did overlap DMMs and Vaults for a few weeks on 2-3 markets, mostly as an experiment to see the impact it had on liquidity. Ultimately, we don’t think this redundancy is needed, and it can lead to issues with the DMM and Vault trading against each other.
Trading stats are not actively maintained at the moment. Our goal is to provide liquidity on all markets in need, not to scrutinize markets based on their activity. Again, we have no insight on future trading trends or behavior, only what’s live today on dYdX. We’ll assess trading data and possible liquidity impacts as part of our review.
Raven and Velar were chosen based on their experience and existing participation on dYdX v4. Both teams were already fully integrated with the exchange. We did engage with other teams as part of our selection process. The DEP had no conflicts at the time of approval.
I agree with you on costs being tricky here. We definitely run a risk of paying for support on a market that never ends up getting any trading activity. This is why we’re excited to see the progress done on Vaults, and are now transition most (if not all) of the support to Vaults funding.
I don’t have a strong opinion on delisting at the moment. Just an open question for the community to consider given the negative experience of trading on a market with limited liquidity – if a market isn’t attracting much interest, maybe it’s better to just delist it?

RealVovochka · August 9, 2024, 5:19pm

Hey, @carlbergman
Thanks for the information and the dashboards

This is a good first step towards transparency

I think it is good to add in the information about Vaults : the address of each Vault and the market in which it is deployed
Also I noticed oddities on Vault 73 (Boden) yesterday it was written Capital Contributed: 20k Equity: 0 and PNL = -20k.
Today I see Equity: 5k and PNL: -15k, it looks strange. The total PNL is not counted correctly.

Also I would like to ask @max-holloway what happened to this Vault that it was deleveraged and liquidated?

And remind you the question who is responsible for the Vault strategy which was asked in my previous reply

The Velar and RavenDAO addresses should be added to the DMM information.

I could be wrong but the RavenDAO address is dydx18p7nz5rqezkyscdz9pv9rchnsesjyjjjyfe92t3

The Velar address is unknown to me.

As for your answers on DMM I think it’s worth looking at question #3, I think statistics are the best way to show effectiveness in this case, I hope the DEP gives more detailed information in their report

RealVovochka · August 23, 2024, 10:14pm

Since Xenophon Labs @max-holloway conducted research in dydx V3 on optimizing trading rewards, which led to their gradual reduction, I have a few questions in the context of the program from Chaos Labs:

The account that interests me: dydx1u07sdcdrlxyjl9z9l34t8t0ysy5ratycz8u6zq

Is this account strategy is only farming rewards (high loss before rewards and profitable after)?
Can we consider the volume generated by this account as organic?
Can the entity behind this account can be considered as an expert in automated trading strategies
An the final question. Is this account controlled by Xenophon Labs?

Thank you

RealVovochka · October 20, 2024, 3:40pm

@carlbergman @max-holloway
Can we see the report on the performance of the vaults? We have some data now.

It looks like over -10% APY . How will we motivate users to be LP in vaults if they are losing money?

Topic		Replies	Views
New Listings due diligence (foxy) dYdX Chain	7	183	August 22, 2024
Incentive Design Ideation Thread dYdX Chain	7	334	September 3, 2024
Liquidity at new markets dYdX Chain	11	817	March 13, 2024
DEGEN Market Liquidity and incentives Grants	0	48	July 24, 2024
Trader Retention in dYdX Governance/DAO Discussions	1	247	September 11, 2024

Stimulating Long-Tail markets liquidity

Regarding DMM:

As for delisting the markets,

Related topics