What Actually Keeps Capital in DeFi: A Data Breakdown | Eva Weng - Caladan

I'm Kelly Levali Hunt. I'm your MCD for the next couple of hours. Hello. Wake up. Who was partying last night?

Okay, good. We have Ava Wang who is coming to talk about the Ten Commandments of Liquidity and Cinnabs. Who's going to be Okay, I need to hear a whoop.

All right. Come on, Ava. Hi, I was also partying last night, so I also feel you guys down there. Um, but today I'm going to give you a 4 billion education in DeFi capital uh retention. So, what I'm going to share is 10 databased commandments on how to design liquidity incentives.

So what we looked at is 160 programs over five years and what we found is that the median protocol lost half of its liquidity within a month of the incentives ending. And the outcome as you can see is very obvious and it's very binary that there is no middle ground. Either you retain 90% or more of your capital or basically it collapses the law below 25%. So what we know very obviously is that liquidity is easy to attract but hard to retain. This is what we refer to as the mercenary capital that we're so used to in the industry.

So hopefully these next ten commandments can be applied to the way that you design your liquidity incentive programs so that um they won't be it won't be mercenary. It won't be temporary but will stay within your ecosystem. So the first commandment is thou shalt create protocol native utility. So pretty straightforward. Um liquidity only stays when it powers something essential.

Programs where liquidity directly supports core protocol functions retain nearly twice as much capital as those where liquidity is only farming token rewards. Um when liquidity generates real economic value that can be through lending interest, staking yield, trading fees or stability mechanisms, it creates a revenue floor that survives beyond emissions. Without that foundation, liquidity is rented and it's not owned and it will leave very quickly. A second commandment is thou shalt not cliff but taper. Um so how you end really determines how much of the liquidity stays.

Um, as you can see, if you have a click model where there's a drastic drop off um, from the emissions or from the way that you're uh, rewarding your liquidity, there's going to be an abrupt end and then for a panic. Um, however, if you taper, if there is a gradual decline, um, fee revenue and organic yield has time to replace a lot of the subsidies. And then if you want to take it one step further, you can actually combine vesting and tapering together and that actually provides a 50% better outcome in terms of less um exits as well. Our third commandment is thou shalt respect ris risk adjusted returns. Um and what that means is not all liquidity is equal, right?

programs that are built around stable or blue chip asset pairs are going to retain far more capital than those that are built around volatile token pairs. Even when incentive APRs are identical, exposure to permanent loss and asset volatility destroys returns and accelerates exits. And so when these liquidity providers face unpredictable downside risk, no amount of emissions are going to compensate for that long term. Our fourth commandment um which is my favorite is um thou shalt convert depth into volume. So liquidity doesn't create value just by existing in the ecosystem.

It needs to be used to create that value. Right? So programs which successfully convert liquidity into sustained trading volume retain significantly more capital than that just where liquidity sits idle like in the ecosystem. So what you're trying to do is create a flywheel from liquidity into creating volume and then that volume goes into fees what then generates the APR and then the fee revenue can begin to replace the emissions within the without the flywheel uh your incentives are just being are subsidizing the unused capital. And so when your liquidity is deep enough and well positioned enough, traders and aggregators are going to consistently route your trades through pools and it'll start generating more meaningful trading fees.

And then those fees can then going into flywheel contribute to the LP yield, reduce reliance on token emissions, create a revenue floor, and make liquidity rational to keep even after the incentives taper. Idol liquidity is not an asset. it's a liability that's going to exit. Our fifth commandment is thou shalt coordinate supply and demand. So both near and mantle are really good examples of this.

Um liquidity only works when it lands into real usage. So programs which synchronize liquidity incentives with pro product launches integrations um were able to retain nearly twice as much um as those who are only focused on attracting deposits. And so when the liquidity arrives alongside live demand, it has work to do from day one. But when it arrives into an empty ecosystem, it waits for the rewards to end and then it leaves. Sixth commandment is thou shalt make liquidity providers governors.

Um because liquidity becomes much more durable when your providers are actually stakeholders. Right? You want your ecosystem. You want everybody actually involved in how the programs are being run. Whether that is emission control or fee influence, LP voting power.

It really allows them to become stakeholders. And then there's long-term alignment. So governance is able to transform these uh mercenary LPS into invested sake even modest implementations. So an example is Veladrome um were able to beat these traditional models where LPS were just in the system without actually having any say in how the system was being run. The seventh commandment is thou shalt stack yields efficiently.

So um multiple income streams equals multiple reasons to stay. Um, if you're stacking, you know, staking yield plus bribes plus points, trading fees, base APR, token emissions, all of these things together, then you're able to significantly out outperform those that are reliant on a single incentive source. So, token emissions only, right? Um that makes sense because when emissions taper, diversified yield sources um can soften this transition and then single source designs will face binary collapse when the rewards decline. Our eighth commandment is thou shalt match duration to development.

Um so even though like time alone does not create retention but without time demand can't mature. So a short incentive blast, so somewhere less than 3 months as opposed to uh some of these longer programs are 12 plus months um are going to consistently underperform these longer programs because in the longer programs you have a fuller in integration cycle. Your ecosystem matures. Think about how quickly you know crypto is runs and works um within this one month. you're really able to integrate um this liquidity into all the various aspects of your ecosystem um versus it leaving very quickly.

Um when emissions end before the real real usage um can absorb liquidity then the capital is going to exit very abruptly. Um but when programs are structured with sufficient runway organic fee generation can begin replacing subsidies before they taper. Almost there. So n um the ninth uh commandment is thou shalt diversify reward sources. So kind of similar to the one before um single points of failure create single points of mass exodus, right?

If you have your emissions ending very abruptly but there's nothing else um that is continuing to reward your your um LPS, they're basically going to leave. But if you have, you know, token emissions, trading fees, um, all of these, you know, future points and air drops, and then your emissions end, but there's still something that's continuing to give, um, they're going to be able to stay much longer, and there's not going to be so much sudden flight. Um, and our last commandment is, thou shalt use friction wisely. Um, so I think we are always talking about frictionless um, trading, frictionless payments. Um, frictionless is usually a good thing in the industry, but actually friction can support retention, but only paired with real demand.

So, the three times three types of friction I'm talking about are lock periods, um, so timebased exit restrictions, tenure boosts, rewards for staying longer, and then exit penalties, costs for early withdrawal. Um, and when you're able to use all of those with this real demand, um, you're going to keep your users in there for much longer. Um, it can kind of reinforce the commitment that you want within your ecosystem. Um, but then if you don't have real demand, friction is just only going to prop uh postpone the inevitable capital fight. It's not going to keep it from um, leaving when you you do.

Um, force retention without economic value creates resentment, not durability. So, there were 10 commandments that we just went through. Um, and you might kind of be thinking, which ones actually apply to me? And my answer is all of them. So, what we found over these 160 programs, um, is that the programs that used eight or more of these commandments had, uh, 41x on retention.

If you didn't use any of these commandments, right, um you really did not have very much retention. So basically there is a you know a correlation between how much you use uh of these and how both sophisticated and versatile you can design your liquidity programs to be in order to retain that capital, keep it in and reduce the amount of mercenary capital that's within your ecosystem. So, if you are curious about learning more about the actual programs, um, feel free to scan the QR code. There's a full 45page report that really goes into all of the data, all of the information. Um, and with that, thank you very much.