Hi everyone. Thanks to @adamdelsol for summing up the various points made so far, which is very helpful. I have some (fairly rambling) thoughts/comments on the above posts, and some little ideas to share.
I did my best to organize this into sections, consistently numbered throughout the post so that replies can refer to parts and ideas more easily.
#1: Regarding JET on-chain liquidity and PMF
I respectfully disagree with @ezio that the most important thing in this thread is JET liquidity depth. Rather, the most important thing by far is meaningful activity on the protocol, AKA product-market-fit (“PMF”). We want to see data showing an uptrend in product usage (meaning that they are useful for users, and PMF exists).
On this thread, here are a few specific metrics that seem important (and that a good LM program would want to help push):
- An uptrend in absolute value both deposited and borrowed (obviously)
- An uptrend in users (makes the assumption that each pubkey is a user which is obviously faulty but still not useless imo)
- An uptrend in a metric like average amounts borrowed and lent by each user on a per-capita basis
- An uptrend in the amount of fees collected by the protocol (from interest paid to lenders)
Sidenote: With that said, whenever the initiative begins to increase JET AMM liquidity, it seems like a good idea mentioned above to apply for ORCA rewards for the pool as @ezio suggested (perhaps in addition to potential JET rewards). Jet has had a good relationship with Orca, built leverage swaps atop of Orca pools, there have been joint AMA’s between the teams, etc. And Orca incentivizes many pools - so it seems very natural to bring this up with Orca’s team.
#2: Attracting other DeFi users
As stated above, growing users and usage is paramount. From Adam’s bullet points, I’d like to go through a thought experiment with one of them –
- reward users of other DeFi protocols - to attract other DeFi customers
This seems like a great place to start because we know for fact and can see / analyze the fact that there is plenty of value being put to use on chain.
Let’s look at a larger lending and borrowing competitor – Solend. The main pool on Solend has a total of ±$420M supplied and ±$156M borrowed. Although there are other pools with material values of assets supplied and borrowed, this main pool is by far the bulk of the platform’s TVL, so we can proceed to look at solely the main pool.
Looking at the main pool and sorting by amount supplied, the top assets supplied are:
- USDC - $105M
- SOL - $100M
- USDT - $30M (note the significantly lower amount than USDC)
- mSOL - $70M (incentivized by Solend and Marinade)
- stSOL -$76M (incentivized by Solend and Lido)
Now let’s look at the amount borrowed for each of those same assets:
- USDC -$63M (incentivized by Solend)
- SOL - $67M (incentivized by Solend)
- USDT - $19M (incentivized by Solend)
- mSOL - $520k (tiny amount compared to supply!)
- stSOL - $4M
I want to mention a couple things that come to mind looking at this data, which we can learn from:
The largest incentives in the main pool are incentivizing USDC borrowing - users are able to borrow for only 0.06% net rate accounting for SLEND rewards at the time of this writing.
As USDC is by far the most-used asset in DeFi, I suggest that JET should consider following suit and using the LM program to boost USDC borrowing the most out of any borrowing. Solend seems to confirm this experimentally.
USDT borrowing also logically makes senses to boost imo, but it is apparent from combing a platform like DefiLlama that USDC is used much more ubiquitously and in higher concentration in DeFi (this is again evidenced on the amount of USDT vs USDC captured by Solend listed above, as well as the fact that Solend incentivizes the USDT pool with less than 50% of SLEND emitted to USDT borrowers as opposed to USDC).
In the case of stables, I believe it makes more sense to incentivize borrowing, as we can expect the average user to have more demand for stables than borrowing other assets and potentially more complicated strategies. Sure, there is a segment that want to borrow all assets. And that’s great!
But imo boosting stable borrowing only (and not lending) limits the cost to the DAO treasury and is still helpful. Lowering costs to borrow stables is the most apparent case of a “borrow incentivization fly wheel” (boosted APYs for borrowing stables → cheaper net borrowing costs → more borrowers → higher rates → more depositors flock to the platform as deposit rate rises.
The staking derivatives on Solend have much more supply than borrowed. mSOL has almost nothing borrowed relative to supplied! And yet…Solend is paying out 2.42% APY to StSOL depositors, and 1.06% for mSOL. These rewards are made up of SLEND rewards (analogous to JET rewards for the sake of discussion here), and to a larger extent proportionally with incentives from Lido and Marinade.
Consider that Jet will earn protocol revenue from a chunk of the interest paid to lenders. Thus, if Jet had a similarly massive supply of staked SOL derivatives, the protocol would earn a lot of steady fees from the interest paid to depositors - The protocol would still make good revenue here, even without much borrowing activity!
To me, this means that the JetDAO should indeed consider boosting supply APYs of mSOL and stSOL by eventually facilitating those projects to juice rewards, either with core team development or as a community developer bounty if there are not enough resources in the medium term.
The important question here imo: How much do we need to boost APYs for depositors on those assets? How can JET easily drain some of the liquidity from Solend?
Simply by ensuring that the rate paid to depositors is higher than on Solend and blasting a marketing campaign about it. Ideally by getting some “juice” from Lido/Marinade, but also by pushing the rate another ±0.5 - 1% (or even 2%+) higher than Solend’s rate by a JET LM program as James suggested. This would require additional research into the historical rates at Solend to see what amount makes the most sense to incentivize for Jet without needless waste. We could also dynamically adjust rates based on what the rates are at Solend but I think that may create some deeper risks/attack vectors.
That would be a hell of a targeted marketing campaign though, and if users were depositing for idle yield, why wouldn’t they switch for higher yield from a respectable audited competitor? We all know how serious wide-eyed yield-hunters are about squeezing as much as possible from their capital (particularly in a bear market where lending on a protocol is a relatively safe option for yield).
I just wanted to specifically draw attention to the fact that there is a material amount of protocol revenue in something as simple as being the default DeFi lending and borrowing platform for staked SOL derivatives.
Note that Solend incentivizes SOL borrowing in proportion less than USDC, but more than USDT. Again, I think this is a well-made and calculated move by Solend that Jet can learn from. As the native asset to Solana, SOL is obviously the bread-and-butter of many active and non-active users of the chain. It makes sense to incentivize SOL one way or another (again, perhaps assuring that Jet rates beat Solend rates?) to try to pull some of those SOL borrowers and depositors towards Jet.
Considering how much TVL is on platforms like Solend (any many others, but this post is already too long) – What about airdropping users of other lending protocols JET proportional to their historical usage of those lending protocols (relative to other users)? This would be a “targeted” LM program that would aim for the heavy hitters in DeFi. I still think the LM should reward the little guy as well, but this is one angle to consider (esp since research shows that a relatively small group of DeFi power users control most of the capital). Further, the LM could also give a slight additional bonus to heavy borrowers since that is the ideal behavior that Jet should encourage the most (borrowing drives up rates, attracting more lenders to the platform).
#3: A time-degrading boost that is replenished by optimizing ideal behavior for the protocol
This covers ideas like adding a JET reward base multiplier based on staking a relative value of JET compared to value of a user’s deposits/borrows, including a time component that degrades the bonus continually unless the user takes action to replenish the full bonus through a behavior that benefits the protocol, etc.
The above would look to the end user similarly to how AAVE LM rewards worked, but AAVE just gave equal AAVE (now OP) rewards them indiscriminately to every single lender and depositor - Jet can optimize for more specific behavior that helps the protocol by considering each app user’s relative value of JET staked, or by considering their voting history, etc.
I think the secret sauce here is to include a bonus/incentive for borrowers and lenders that is dependent upon them having JET staked (and perhaps on them being active voters, at least in part).
Just throwing some casual numbers out – for example, if value of the JET they stake is equal to ±10% or more of their borrowed (or deposited/lent) amount, they get a JET bonus for borrowing and depositing, thus reducing their borrowing interest costs and/or boosting their interest earned from lending. Could do tiers as well (ie a tier for 5% of app value staked, 15%, etc). Again, these numbers are totally random and I am using them to share an idea. The actual values would require some deeper thinking, consideration, and discussion.
Ie, to keep the max boost for your LM incentives on borrows and deposits, the user needs JET staked equal to x% of your borrows or deposits, AND this bonus degrades slowly over time but is replenished (in part) by votes (or staking more JET). I’m imagining more or less a veCrv-like system here that would incentivize users to stake and restake continuously + vote for maximum benefit.
I also think that this ties in nicely with @wil’s point above about providing a multiple to users based on their staking duration.
#4: Partnering with other protocols vs a community developer bounty
From Adam’s bullet points:
- partnering with other protocols to incentivize lenders
Marinade has stated their openness to add to rewards for mSOL depositors, so why not start with them since Jet has a good relationship?
The main constraint here is developer resources.
But at the same time, it is self-evident that the juiced rewards from Lido and Marinade would pair very nicely with Jet’s own LM program and help pull in users.
Perhaps a good starting place is to test the waters of community contribution by offering a grant for the dev work necessary? While bringing up community grants this is a slight digression from the thread at hand, it’s actually related to this discussion because partnering with Lido and Marinade (and potentially other projects!) for the sake of attracting more users and usage would reduce the strain on JET token incentives from this LM program.
#5: Which Jet products should we focus on discussing LM for?
I understand that the upcoming Jet Fixed Term product will be the most novel from a DeFi perspective and most useful to the largest amount of users.
So perhaps our discussion should expand consider whether the LM program that is created should devote to both the current lending pools as well as the fixed term product? In what proportion? Should we should consider thinking ahead to bolster participation in the new app when it is ready, in addition to the traditional “pools” model that exists now?
I want to make it clear again that these ideas are just early stage ideas and not thoroughly researched yet; but rather ideas I wanted to share for the sake of discussion.
I’ve been loving the activity in this thread and wanted to contribute to it and do my part to keep the discussion rolling. I’d be happy to see these ideas tweaked into something even better for the protocol, or shot down completely for better ideas. If any of these inspire thoughts (or even better, a post) from any community members, I will be glad I posted. Thank you for reading!