Olympus Pro LP Proposal

[Leighton]

Introduction
Some members of the PoolTogether community expressed interest in the Olympus DAO Bond program (see @rvdeeb post here). Additionally, Olympus DAO recently launched one of the most successful prize pools on the protocol 3,3Together (already over $20 million deposited and 1,000 users in the first 7 days). Finally, our current LP rewards are scheduled to end in 14 days.

Given the growing synergies between our communities and the popularity of their Bond program. I reached out for more details and jointly wrote up this proposal with Tex who is on their team.

Proposal Overview

This proposal outlines a program for PoolTogether to own its liquidity on AMMs. At the time this proposal is written, there is approximately $1.5M in liquidity for POOL/ETH on Uni V2 ($1.35M) and Sushi ($150k). PoolTogether began incentivizing its liquidity earlier this year at 500 POOL per day and currently has tapered emissions to 300 POOL per day. This is inline with an overall push to lower total emissions and use them more effectively.

Motivation
Implementing bonds reduces the need for POOL incentives over time, which only create temporary liquidity and persistent sell pressure. Importantly, bonds incentivize active PoolTogether community members to participate in the program and help distribute tokens to these users similar to traditional liquidity mining programs.

To make it very basic:

• Currently PoolTogether provides 300 POOL per day in LP incentives. This is “renting liquidity” because as soon as the incentives disappear, the LP’s leave. Leaving nothing behind (this happened when incentives decreased from 500 to 300).
• With a bond program PoolTogether can provide a similar amount of POOL per day to people who GIVE their LP tokens to the protocol permanently. This builds long term value and opens a path to reduce LP rewards to zero.

Specification
For those unfamiliar with the Olympus bond mechanism, the bond contract essentially sells a token (ex: POOL) at a discount in exchange for assets (ex: POOL/ETH LP tokens). Bonds operate by initializing the price for LP tokens above market price and then applying a discount function that decrements price until a bond is purchased, which pushes the bond price back up. This mechanism allows the market to determine the optimal price for bonds. For reference, here is the average discount of OHM bonds with their extremely high APY:

Olympus is offering to provide its expertise in bond contract management to support other DAOs interested in owning their own liquidity. This will include providing the UI for bonds and maintaining bond control variables to balance emissions with liquidity accumulation. In exchange for the implementation and community engagement, Olympus would take 3.3% of all POOL sold to back OHM tokens. This will align the success of our communities and allow for cross-DAO governance participation.

PoolTogether’s bonds would have a 7-day vesting period to ensure that discount buyers don’t immediately sell into the market. This helps align the goals of bond participants with the goals of the protocol. In addition to purchasing POOL at a discount, bonders know that they are providing permanent liquidity to PoolTogether’s treasury. An additional benefit of bond programs is that they eliminate impermanent loss inherent in traditional liquidity mining and the discount is locked in at the point of purchase.

Proposed Bond Program:

• Accumulate POOL/ETH liquidity
• Bond 14k-20k POOL over 4 weeks
• Vesting period: 7 days

Pros

• Gets a far higher ROI for POOL than we are now
• Helps build a relationship / increased exposure with another large DeFi protocol
• Builds protocol owned POOL/ETH liquidity
• Allows current liquidity incentives to be tapered/eliminated over time
• Lower POOL emissions should improve tokenomics in the long-term

Cons

• Added POOL emissions via bonds may cause sell pressure in the short-term
• 3.3% fee to Olympus Protocol for facilitating the program

• Support using Olympus Pro Bond Program
• Oppose using Olympus Pro Bond Program

[Brendan]

I think this is a very sensible way to increase and diversify our treasury while supporting token holders.

However, since Olympus wants to take a fee I think we should enable the reserve on the OHM prize pool. Then truly the protocols are intertwined; we are effectively doing a token swap.

We could start with the memetic amount by setting the reserve to 3.3%.

We would then be diversifying into OHM just as they are diversifying into POOL! It would be a strong token partnership with Olympus.

[TheRealTuna]

Great proposal, @Leighton ! Thank you for starting this discussion.

While the proposed Treasury Management Committee has not yet been formed, the proposed members wanted to weigh in with an opinion and provide some additional insights for the PoolTogether community to consider when reviewing this proposal.

Other protocols have benefited from using Olympus Pro to sell bonds and add protocol-owned liquidity to current treasury holdings. Alchemix discussed their bond sale in AIP-21, and any community members interested in reviewing their discussion can read the relevant forum post.

After discussing the Olympus Pro LP Proposal internally, we believe PoolTogether would greatly benefit from a bond sale.

Rationale

1. A bond sale that targets liquidity where it is deepest and where we offer the greatest incentives (Uniswap V2) benefits the protocol in the long run.
2. Through a bond sale, PoolTogether would diversify into ETH and earn trading fees on POOL buys and sells. With protocol-owned liquidity, our treasury would benefit from short-term traders.
3. With Uniswap V2 positions, we essentially buy more POOL as the price of POOL vs. ETH declines and further diversity into ETH as the price of POOL vs. ETH increases. All the while, the treasury will earn trading fees.
4. Through a bond sale, we give current LPs a choice: are you more bullish on POOL long term vs. short term? If so, sell your liquidity to the treasury for discounted POOL.
5. Establishing a DAO-to-DAO relationship with Olympus could prove to be very beneficial long term. Inter-DAO collaboration is crucial, and working with Olympus on a bond sale would strengthen our relationship with their community, while adding POOL to the Olympus treasury.

If PoolTogether uses Olympus Pro for the bond sale, it would further align incentives between our communities if PoolTogether could add a 3.3% reserve rate on the OHM prize pool, as @brendan has already proposed in this response. Olympus would have POOL in their treasury, and PoolTogether can earn OHM to hold in the treasury reserve.

Any scenario where PoolTogether buys ETH and supplies our own liquidity would cost more when time, resources, fees, etc., are taken into account. This approach would also take longer, as we wouldn’t want to sell large amounts of POOL at once to acquire ETH.

With any other approach, we’d still be discussing the sale of POOL to facilitate the same outcome. Using Olympus Pro to execute a bond sale would be the most cost effective short-term solution with the added benefit of a stronger relationship with the Olympus community.

Review of Uniswap V2 Incentives

Below is a breakdown of current incentives over the next 6 months (182 days).

While current Uniswap V2 incentives will expire in ~14 days, we can use the existing 300 POOL per day to evaluate the expense in terms of POOL over 6 months.

300 POOL x 182 days = 54,600 POOL

Assuming we kept incentives at 300 POOL per day, the impact of a bond sale would be:

Lower bounds w/ 14,000 POOL used to buy bonds: 40,600 POOL / 182 days = ~223 POOL/day
Upper bounds w/ 20,000 POOL used to buy bonds: 34,600 POOL / 182 days = ~190 POOL/day

If incentives were lowered to 250 POOL per day, the impact of the bond sale would be:

250 POOL x 182 days = 45,500 POOL

Lower bounds w/ 14,000 POOL used to buy bonds: 31,500 POOL / 182 days = ~173 POOL/day
Upper bounds w/ 20,000 POOL used to buy bonds: 25,500 POOL / 182 days = ~140 POOL/day

If PoolTogether used a bond sale and sold POOL in exchange for LP tokens, the protocol could reduce emissions by at least 25% over 6 months. With a further reduction of 50 POOL/day in incentives, the protocol could reduce emissions by at least 42% over the next six months.

As a trade off, PoolTogether would be selling 14,000 to 20,000 POOL in the short term to any current LP holders or opportunistic community members. Current LPs don’t have to sell LP tokens for POOL.

Why a Bond Sale Makes Sense

Let’s assume a few things:
Discount Offered to LPs: 5-8%
Olympus Pro Fee: 3.3%

At 14k POOL
Olympus Fee: 462 POOL
5% Discount: 700 POOL
8% Discount: 1,120 POOL

Total LP Owned at 5% + 3.3% = 12,838 POOL worth of LP (~6.34% of Uni V2 Liquidity)
Total LP Owned at 8% + 3.3% = 12,418 POOL worth of LP (~6.13% of Uni V2 Liquidity)

At 20k POOL
Olympus Fee: 660 POOL
5% Discount: 1,000 POOL
8% Discount: 1,600 POOL

Total LP Owned at 5% + 3.3% = 18,340 POOL worth of LP (~9.06% of Uni V2 Liquidity)
Total LP Owned at 8% + 3.3% = 17,740 POOL worth of LP (~8.76% of Uni V2 Liquidity)

Instead of dripping POOL and having farmers sell POOL with no added benefit to the PoolTogether community, we can use existing POOL allocated for liquidity mining incentives to purchase LPs and benefit from short-term traders.

Owning liquidity is a long-term bet that trading fees will outpace the cost of a bond sale, and it allows us to reduce POOL incentives over time.

Protocol-owned liquidity is subject to impermanent loss, but if the price of POOL declines, the LP position rebalances to buy more POOL and sell ETH to achieve a 50/50 ratio. As the price of POOL increases, the LP position sells POOL to diversify into ETH. In either scenario, it’s a benefit for the treasury.

We encourage the community to discuss this proposal, but we are in favor of using Olympus Pro to add Uniswap V2 liquidity to the PoolTogether treasury.

[Torgin]

@TheRealTuna

Any scenario where PoolTogether buys ETH and supplies our own liquidity would cost more when time, resources, fees, etc., are taken into account. This approach would also take longer, as we wouldn’t want to sell large amounts of POOL at once to acquire ETH.

This seems quite unlikely to me. We could just go out and sell POOL or USDC for ETH on the open market over time, then LP with ETH/POOL. This would have the same effect as the bonds but avoid giving a ~5% discount + 3% fee. I can’t imagine the operational cost of doing something like this would be close to 8%.

Bonds have a positive impact if we assume that those buying POOL at a discount through bonding are less likely to sell POOL than the LPs we have now. I don’t know why this would be the case.

If we do go forward with this, it should primarily be to strengthen the partnership between the two DAOs. I can see a token swap style approach as outlined by Brendan work well.

[BRONDER]

As I explained on a previous topic, I’m actually not a fan of this proposal. I really like the idea of the protocol owning it’s own liquidity so that we do not depend on dripping POOL to be able to have liquidity providers.

I believed that I’m 100% in the same boat as @Torgin , and the only advantage that I can see is stablishing a relation with Olympus DAO. However, I don’t really like this relation just yet. Their token prize is unsustainable given how much they’re minting. When that prize goes all the way down to where it is supposed to (each token, is backed by around 40$ if I’m not mistaken), then I can really consider an alliance with them. Until that happens, I’m absolutely against getting OHM tokens for the treasury.

But please, if you end up getting OHM tokens, don’t have them sitting there uselessly, stake them.

[ChrisSavadge]

I would oppose this proposal. Why would you want in your Treasury an asset that can do this? If you think this is an isolated drop, amid a perpetual uptrend, I’m afraid I have to disagree.

I would also like to see a legal analysis, and understand the increased risk of POOL being deemed a security under this proposal. Selling bonds at a discount? Long-term tokenomics? Even if you’re idealogically opposed to securities regulations, there’s no denying that they represent a risk.

I would like to see PoolTogether adopted by state lotteries as a superior system to the status quo. That seems a less likely prospect if you start inviting enforcement actions from regulators, by hitching your wagon to the flavor of the month token.

[Leighton]

I think there is some confusion here. This proposal has nothing to do with the Olympus token. It wouldn’t matter if the Olympus token went to $0. It would have no impact on us or this proposal.

Could you explain where you think the Olympus token fits into this?

[rvdeeb]

Thanks for the write up Leighton, the alignment between the two communities can go a long way especially with the reach of Olympus. The bonding program will also ensure liquidity is sustainable in the future if successful.

There is some concerns around Olympus and the OHM token which may be valid. It’s worth mentioning Olympus Pro is a separate protocol from Olympus DAO so the smart contracts or bonding program would not be affected by any price action on the OHM token.

I’m in agreement with Brendan also that we should look at switching on the reserve for the OHM pool.

[Leighton]

To clarify, this proposal doesn’t have anything to do with the Olympus token. This would be for PoolTogether to use the Olympus Pro platform (see here).

This is similar to how we use Uniswap or Sushiswap, our usage of those protocols doesn’t have anything to do with the UNI or Sushi token.

[BRONDER]

It is clear for me that this proposal is just to use Olympus PRO (although it has been mentioned the possibility to get some OHM tokens in exchange for POOL if I’m not mistaken).

Just taking into account the use of Olympus PRO, I’m still in favor of approving the purchase of liquidity for the treasury without the use of this bonds. That way, we would save around 8% of costs (bond discounts + Olympus fee). The downside is that we won’t be promoting the relationship with Olympus.

Assuming that we will buy around 1M liquidity, the question for me is: do we want to give up 80K $ to promote the relationship with Olympus?

[Brendan]

This seems quite unlikely to me. We could just go out and sell POOL or USDC for ETH on the open market over time, then LP with ETH/POOL. This would have the same effect as the bonds but avoid giving a ~5% discount + 3% fee. I can’t imagine the operational cost of doing something like this would be close to 8%.

That’s what is so cool about the bonds- it’s a very useful mechanism to sell tokens. Let’s consider other options to sell:

• AMM: Requires no coordination of buyers, but impacts the market significantly.
• Gnosis Auction: some coordination required for the public auction event, but impacts the market less (if purchase price doesn’t result in immediate arb opportunity).
• OTC deal: requires significant coordination, but the whale buyers are unlikely to sell so it has zero impact on market.

This is where bonds shine, IMO: it’s a set-and-forget continuous process. Very little coordination is required, and the sale doesn’t impact the market. The discount isn’t absolute either; the discount drops until someone buys the bond. And- the tokens vest so they can’t immediately dump.

It’s a great way to sell tokens, so that’s why it’s been receiving a lot of praise. It’s a great mechanism to add to our quiver, and Olympus has all the infrastructure set up.

[Brendan]

• Bond 14k-20k POOL over 4 weeks

Want to add that I’d support the upper-end of this: 20k POOL over 4 weeks.

Edit: I also want to add that I thought we were going to do a $1m or so; $200k of POOL is a very small amount. This is a good way to dip our toes in Olympus Pro and try it out. I think we can hold off on the reserve switch.

[ChrisSavadge]

I was reacting to the comment from TheRealTuna:

Olympus would have POOL in their treasury, and PoolTogether can earn OHM to hold in the treasury reserve.

Thanks for the clarification.

[Torgin]

Why do you assume that selling bonds has a smaller impact on spot price than selling continuously on an AMM? The vesting period is only one week.
Selling POOL at a discount makes it more likely to be mostly dumped as an easy arbitrage imo.
Maybe with a longer vesting period I could see it.
That still leaves the 3% fee of Olympus.

[qwercus]

ill start with a disclaimer that i am both financially and ideologically aligned with both PoolTogether and Olympus, so i have obvious bias in wanting to see the two align themselves.

i think everyone agrees that protocol owned liquidity is worth pursuing for the long term health of the protocol and as a benefit for its users. the question we are asking ourselves is the same question other DAOs face: is this service/frontend Olympus offering worth the 3.3% of bonded POOL tokens their treasury will accumulate? I’ll leave that decision up to the individual, but Brendan brings up a great idea in turning on a reserve for the sOHM prize pool to offset the 3.3% of POOL going into the Olympus treasury.

@Brendan

However, since Olympus wants to take a fee I think we should enable the reserve on the OHM prize pool. Then truly the protocols are intertwined; we are effectively doing a token swap.

I think this is brilliant, but the reserve rate should be carefully considered to compare to the amount of POOL that would be going in the opposite direction. At Leighton’s upper bound of 20k POOL, Olympus treasury would be taking in 660 POOL. At current market prices, that is approximately 10k USD.

The sOHM prize pool has awarded prizes around 1m USD every 3 days in the form of sOHM/33T. 3.3% of a million dollar prize pool would be moving approximately 30k USD to the protocol controlled reserve every 3 days. It should be noted that sOHM is currently very volatile and will continue to be for a while until it leaves the growth phase and enters the stability phase.

But based on some of the price concerns above, i feel the need to clarify that the sOHM token is staked OHM and is rebasing itself approximately every 8 hours. This token in the reserve would be gaining interest at a variable rate, roughly determined by the below table found in OIP18

The way i see it, an initial bond program is the start of a mutually beneficial relationship. PoolTogether has use in owning its own liquidity. Olympus has use for POOL as a reserve asset. PoolTogether has use for sOHM to rapidly grow prize pools to draw in users to headline prizes. Olympus has use in those prize pool participants for growing its ecosystem. PoolTogether will lock up sOHM in prize pool reserves and Olympus will lock up POOL in its treasury reserves. I have participated in community calls for both protocols, and can tell you that both protocols have flywheels that perpetually grow community and value. This partnership seems like an obvious acceleration of both flywheels.

[tex]

When you’re incentivizing liquidity via rewarding community members, selling continuously into your AMM liquidity has the effect of dumping on your community. This decreases token price, exacerbates IL, and generally has poor optics.

In our experience at Olympus, the majority of bond purchasers tend to roll their bond over week-to-week to increase their effective ROI. This actually aligns the long-term incentives of the protocol (acquire liquidity) and bonder (grow POOL stack). While it may seem like easy arbitrage opportunities, it’s even easier for PoolTogether community members to snatch up smaller discounts and compound them over the course of the bond program.

[TheRealTuna]

I assumed the reserve would be turned on at some point and what better time to do it then in conjunction with this proposal? I don’t think a less than 5% reserve on this pool is going to ruffle any feathers. We need to make sure our TVL is actually working for us and providing some return.

[monstrosity]

Just throwing a few things out there:

1. We don’t have create a POOL-? LP. One purpose is to diversify our reserve with assets we want to hold another thought is to own our own a portion of our liquidity.
2. xdai network is offering some rewards if we did a POOL-STAKE or ETH-STAKE LP in return for discounted $STAKE if POOL matches rewards on the xdai protocol.
3. or we could do the reverse POOL-STAKE in return for discounted $POOL and we could have some matching $STAKE incentives on the Pool protocol on xdai.

Essentially POOL could receive 1 for 1 matching of rewards as long as we bring the incentives to xDai. One more piece to this is with an xDAI launch for OlympusPro our Bond wouldn’t get lost in potentially a long list of other projects on eth mainnet. There maybe a half dozen pages to scroll through within the next month or two.

@Torgin

The vesting period is only one week.

We could make the vesting period 14 days 30, or even 60 days. Obviously you’d expect a higher premium on the discount but arguably you’d be attracting long term holders of the token.

[BraveNewDeFi]

I’ll give my personal 2 POOL about the Olympus DAO proposal:

There’s a definite benefit of using Olympus Pro for bonds regardless of what anyone thinks of Olympus as a protocol or the OHM token itself. We’re just using a DAO for a service that can provide value to the PoolTogether community.

Bonding vs. Current LM Incentives

We’re currently giving away 100% of the liquidity mining incentives to LPs on Uniswap V2. At 300 POOL/day, we’re giving 2,100 POOL/week to Uniswap V2 LPs.

If PoolTogether were to use Olympus Pro and exchanged POOL at a discount of 5% over a 4-week period with a 1-week vesting period, then:

14k less the 3.3% fee would be 13,538 POOL, which represents an increase of 50.42% of existing monthly POOL distribution to LPs.
20K less the 3.3% fee would be 19,340 POOL, which represents an increase of 114.89% of existing weekly POOL distribution to LPs.

However, instead of giving away POOL at a 100% discount, we’re closing that to 5% in exchange for LP tokens, which provide exposure to both POOL and ETH. And then you can add the 3.3% fee Olympus Pro charges to handle the sale of bonds.

Using bonds would mean we’d earn 95% of the value back in LP tokens instead of giving away 100% of POOL through LM incentives. That is a huge benefit.

Olympus Pro Bonds vs. Selling POOL from the Treasury to Provide Uniswap V2 Liquidity Directly

There’s a couple ways to look at this:

1. All things held equal: we’d still be dripping 300 POOL/day AND selling POOL to buy ETH over a period of 1 month.
2. Reduce incentives and use POOL to buy ETH: I’d assume we’d see more LPs leave if we reduced incentives significantly all at once. A step down in LM incentives could be done but that would exacerbate sell pressure if we’re assuming all Uniswap V2 LPs are selling POOL.

If we use 20k POOL and bond that over a 4-week period in exchange for liquidity, we’re buying out liquidity and giving LPs an incentive to do so. While the sell pressure could double within a 1-month window, it will likely taper off over time.

Based on today’s volume on Uniswap V2, 4,835 POOL/week represents $68,947.1, or 23.72% of the daily volume.

POOL is up ~30% on the monthly. A 5% discount + Olympus Pro’s 3.3% fee is $13.07, where POOL was trading on 12 October.

If POOL were bonded and vesting for 7 days, which caused 4,835 POOL to be sold every week, we’d see ~$69k in added volume each week if we assume that every single bonded POOL is sold at current prices.

Instead, if we decrease daily rewards for LPs to 140 POOL/day and offering a buyout through bonding, we’re essentially creating demand to sell LP tokens for bonded POOL. It would cost PoolTogether 20k POOL over 1 month, but over the course of the next 6 months, the protocol would save 9,780 POOL given (182 days * 300 POOL/day) - [(182 days * 140 POOL/day) + 19,340 POOL] = 9,780 POOL

That’s a 16.7% reduction in sell pressure over the next 6 months from Uniswap V2 LPs given increased sell pressure over the first month when bonds are sold.

If instead, POOL was sold to buy ETH over the next month, then we’d be selling 10,000 POOL (~$142,600 at current prices) for ETH, everything held equal. If this were in addition to current rewards, then there would be ~19,000 POOL sold within the next month. Given that 660 (3.3%) of 20,000 POOL would be held in the Olympus DAO treasury, we’d only see an additional 340 POOL being sold within the same time frame if we assume every POOL drip to LPs is being sold and every POOL bonded in exchange for liquidity is being sold as well.

If POOL is sold over a longer timeframe, then we run the risk that the price could drop or rise. If the price were to drop by $1.18/POOL or if POOL fell against ETH by 3.3% within 2-3 months, then we’d be in the same place as we would be if we bonded and sold POOL in the short term.

The biggest factor from my point of view: PoolTogether is dripping POOL to Uniswap V2 LPs at a 100% discount; whereas, PoolTogether could sell bonded POOL with a 7-day vesting period at a slight discount with a fee to Olympus and own ~9% of the existing liquidity on Uniswap V2 where the deepest liquidity exists, while closing the existing LM discount from 100% to 8.3%.

[monstrosity]

@BraveNewDeFi

Instead, if we decrease daily rewards for LPs to 140 POOL/day and offering a buyout through bonding, we’re essentially creating demand to sell LP tokens for bonded POOL. It would cost PoolTogether 20k POOL over 1 month, but over the course of the next 6 months, the protocol would save 9,780 POOL given (182 days * 300 POOL/day) - [(182 days * 140 POOL/day) + 19,340 POOL] = 9,780 POOL

We could marry that with the xdai network matching 1 for 1 in post above yours. This double rewards would put some additional buy pressure on the LP and could be used to attract further use of the protocol. As a negotiation it would probably make sense to do a ETH/STAKE as a reserve asset or POOL/STAKE LP to offset sell pressure on POOL while diversifying into the STAKE token a bit. The main benefit of course is POOL acquires assets in our reserve and double the rewards to incentivize.