I’d like to propose a fairly dramatic change to our liquidity mining framework.
It’s relatively hot off the presses, but Olympus have released Olympus Pro, a mechanism for protocols to purchase the liquidity that they wish to promote for certain tokens rather than rent it through liquidity mining.
I think this is a great idea, and one that we should seriously consider.
I’m going to give one of my famously lengthy chats about the why, and then the how.
Disclaimer: I do not hold any OHM, and have not done for several months now. @Glueeater is a member of the Indexed Sigma Committee, and is heavily involved with Olympus.
Absolute beginner summary: Protocols often use the power of their treasury (in the form of their native token) to reward third-parties that provide liquidity on DEXes: this is often in the form of giving out token X for providing liquidity for X/ETH. This is known as a ‘pool2’ for historical reasons, and it’s a double-edged sword: it makes the token liquid, but at the cost of downwards selling pressure on token X, as people claim their rewards and trade them for ‘safe havens’, primarily to offset impermanent loss within their LP tokens that is manifested from other people doing the same thing.
More importantly, this can only ever be a temporary solution: once the rewards of token X end, there is no incentive beyond trading fees for people to continue providing that liquidity. So, it disappears. The net result is that liquidity is gone, and you’ve handed out potentially millions in token X with nothing to show for it beyond having turned a profit for some providers, and subjected token X to an absolute beating in the interim.
Indexed is no stranger to the idea of liquidity mining: we have an ongoing scheme that will have handed out approximately 600,000 NDX between May and October of this year for providing liquidity to our index tokens. We don’t engage in pool2 rewards (i.e. rewarding NDX for liquidity for NDX/ETH), but we do rent liquidity for DEFI5, DEGEN, CC10 et al.
Separate to this, we handed out 2,500,000 NDX (25% of the total supply) very early on in the lifetime of the protocol in liquidity mining, however the ‘true’ purpose of this was twofold: to wrest control of the protocol from the founders (who had 20% reserved) and to advertise its launch.
The side-effect of this is fairly obvious to see on a price chart, and while it is not my intent to make this post about the price of NDX, I have natural concerns about both securing the DAO (you don’t want to make it too cheap for someone to buy up enough tokens to start making hostile Governor Alpha proposals and be close to quorum by themselves - see Venus), and maximising the spending power of the Treasury for things other than liquidity.
For the most part, Indexed has engaged in this liquidity rental because there was no alternative: liquidity mining, for better or worse, has been the default way of ensuring that tokens - in our case, our index products - can be easily traded by third parties: sure, you can mint them directly from the site, but that incurs a swap fee that’s intentionally quite high because the goal of the indices is to minimise IL rather than to maximise inter-pool trading: if you can establish liquidity that incurs a slippage on a decent-sized trade that’s lower than the 2% swap fee when minting, you’re winning.
I digress: it looks like there’s a different way now.
The key idea behind Olympus Pro is that of offering bonds in conjunction with/in lieu of an emissions program. Here’s how this would look:
- Indexed offers some amount of NDX for sale in exchange for liquidity that they want to acquire such as DEFI5-WETH or FFF-WETH.
- Olympus advertises these market/s on the Olympus Pro website, and Indexed core integrates the market into the Indexed webpage. All necessary audited contracts and infrastructure has already been built out by Olympus.
- Interested counterparties can trade the desired LPs in exchange for NDX, receiving a discount that defaults to 5% (i.e. selling $95 of DEFI-WETH for $100 of NDX).
- 3% of the purchased NDX is kept by Olympus as a facilitation fee, and retained by their Treasury in perpetuity.
- The remaining purchased NDX vests over the course of seven days (as a default) and be claimed over time (i.e. 14.28% can be claimed after one day).
- The liquidity tokens are sent to the Indexed Treasury.
The net result is that the Indexed Treasury has bought liquidity outright in exchange for NDX, where it can be treated as a revenue-generating asset (from swap fees). Short of a Governor Alpha vote from the DAO to sell, exchange or otherwise dispose of these LP tokens, this can then be considered a permanent liquidity floor for our indices.
In light of the upcoming Balancer V2 upgrade for the Indexed protocol, this liquidity would also be interest-bearing, as one half of the LPs would be earning best-in-lending-market rates due to their integration with the respective underlying Nirn vaults.
It’s worth pointing that the default figures presented here (5% discount and 7 day vest) are configurable by Olympus, but are subject to the amount of NDX being offered to bond: I’m looking for some clarity from Olympus as to what exactly constitutes ‘enough’ to change these numbers (and by how much), but the key idea remains the same.
My take is that as an index protocol, we should be utilising our Treasury to dogfood our own products as much as possible, while simultaneously ensuring that the NDX that we do spend for liquidity isn’t going into the grinder.
Exchanging NDX for index liquidity in this way is good for our users: those who buy our tokens to hold in the long-term (they always have a way to trade our products, even if they’re not familiar with the act of minting and burning Balancer LPs via our site), and it’s also good for the protocol, for two reasons:
- Permanent liquidity ensures that there’s always a reason to arbitrage the indices between their NAVs and their AMM values, thus generating exit fees, and
- As alluded to above, the swap fees and [soon] Nirn revenue drawn from the purchased liquidity enriches the Treasury.
All of the above is not to say that I don’t have some concerns: some of which we’ve imposed upon ourselves, some of which are more ‘market-forcey’ in nature. They are:
We need an answer to the question “why would I sell my blue-chip index/Ether liquidity [i.e. DEFI5/WETH] in exchange for the NDX token?”.
We’re seeing various generous third-party farming opportunities arise lately for NDX/WETH (such as the Onsen program by Sushi and the upcoming Balancer-sponsored rewards for the 80/20 NDX/WETH pool on Arbitrum) and the protocol-revenue sharing program is imminent: are these opportunities ‘enough’?
If they’re not, what kind of discount would we need to offer for this to be appealing enough to convince people, given that the Olympus setup is such that the buyer would have to wait seven days for the bond to fully redeem? This isn’t a question that any one person can answer, but it’s something to bear in mind.
If we engage in a bond program like this in any meaningful fashion (i.e. offer a bond of 100k NDX for liquidity), we’re effectively releasing that much to potentially be market-sold as soon as the bond period is over, rather than emitted slowly via liquidity mining (that much NDX would take just under five weeks to emit at the current stage of our liquidity mining program).
Do we consider this risk - that of non-trivial amounts of NDX potentially being dumped in quick fashion - worth the benefit of acquiring permanent liquidity? In an ideal world the NDX would only be purchased by fans of the protocol who want more say in governance (or protocol revenue sharing through dNDX), but we can’t enforce this.
How much liquidity would we want to acquire, and do we want to acquire it for all of our index products or just the most popular ones? Achieving 2% slippage with a US$10,000 purchase on a 50/50 AMM needs ~US$997,000 in liquidity at US$3,425 ETH, and 5% slippage needs ~US$400,000.
It’d be wonderful to have enough liquidity to guarantee as low a slippage as possible for everything, but that would pretty significantly dent the Treasury. What’s best (in the eyes of the protocol) here?
The crux of what I’m getting to here is this:
I propose that the DAO consider routing a significant portion of emissions going forward towards bond markets to purchase liquidity for the protocol, with mind to potentially terminating the current liquidity mining scheme entirely in favour of this.
The currently active liquidity mining scheme is set to emit 1,500,000 NDX over two years on a linearly-decreasing curve. The amount that was initially transferred is set to be fully allocated around mid-October, after which the DAO needs to vote via Governor Alpha to transfer the remaining 900,000 to the Masterchef from the Treasury.
In the interests of having a concrete proposal for people to chew on, here’s the pitch:
Reduce the amount to be transferred via Governor Alpha to the Masterchef in October to 200,000 NDX, keeping the remainder within the Treasury for now.
Of the withheld 700,000 NDX, utilise 100,000 NDX to create a bond market for DEFI5-WETH and FFF-WETH liquidity for the Indexed DAO Treasury via Olympus Pro, bringing forward the end-time of the liquidity mining program appropriately.
Note that due to the way that the curve for our Masterchef works, agreeing to this means that the DAO subsequently needs to vote on-chain to terminate the program at the new end-time: the block at which 1,400,000 NDX has been emitted, assuming that there was a subsequent vote to move the remaining 600,000 NDX into the Masterchef (earlier still if there are later decisions to allocate more of this NDX elsewhere).
This is because token rewards are allocated on a per-block basis, and if we let the scheme run to completion without transferring the full amount, liquidity providers will end up attempting to withdraw more NDX than is available, resulting in later withdrawal attempts failing. It would get ugly.
might will annoy a large chunk of the current liquidity providers, but I truly believe that everyone that engages in liquidity provision knows that yield-farming in its present form is simply - as I have already said - rented loyalty that results in a slow bleed for the reward token. I don’t think any of us want that to continue for NDX.
We have time to see how the initial wave of partners for Olympus Pro report back on their experiences using the bond program for their own purposes, and use that to gauge whether we want to go forward with this.
If this proposal is accepted and deemed a success, I would be minded to put forward proposals for additional bond markets drawn from the withheld Masterchef allocation, further bringing forward the end of the liquidity-mining scheme.
So, at 1,866 words including this… thoughts?