PoolTogether

$1 million weekly prizes & POOL distribution

We are 15 days away from the initial POOL rewards period ending. As such, now is a great time to determine a longer term strategy for POOL distribution. The initial POOL distribution period has been quite successful in

  • Generating ~$200,000 in weekly prizes
  • Growing AUM to ~$237 million
  • Growing reserves by $75,000 - $150,000 weekly (depending on interest rates)

Even with this success, there are a few issues that have become clear with current POOL distribution:

  • The initial distribution schedule was very aggressive (5% of total supply over 14 weeks). We want to move to something that is more sustainable for a longer period of time. Uncertainty about when POOL distribution will end or change discourages deposits so long time frames are helpful.
  • Some large whales are simply farming the POOL and don’t care about winning the prizes. These whales should not be eligible to win prizes. We’ve had many wins by small depositors but there is a perception problem that only whales win.
  • The yield rates on assets are highly varied. Targeting high yield assets with POOL distribution leads to larger prizes and faster reserve growth. Targeting low yield assets helps grow AUM but doesn’t contribute significantly to prizes.
  • POOL distribution is equal across similar asset types (stablecoins both get the same rate) this fragments prizes and prevents the growth of really large headline prizes ($1 million per week).

The goal with adjusting POOL distribution is to address these issues while accelerating the positive outcomes. I propose the following:

  • Reduce POOL distribution to low yield non-stablecoin assets by 80% (50 POOL per day instead of 255 POOL per day)
  • Reduce POOL distribution to the Dai prize pool by 65%, this is currently the pool being most heavily farmed and also has a lower average yield than USDC.
  • Reduce POOL distribution to USDC prize pool by 13%. The idea here is to make a very modest reduction that would make the USDC prize pool the default prize pool with the largest prize
  • For USDC and Dai prize pools, split POOL distribution 80% to sponsored deposits and 20% to regular deposits. “Sponsored” deposits are deposits that contribute interest to the prize but are NOT eligible to win. So the yield farmers should go over to sponsorship and increase the probability of small depositors win large prizes
  • Keep 300 POOL per day LP incentives
  • Keep 100 POOL per day to POOL pool depositors
  • Plan for up to 500 POOL being allocated to new prizes pools
  • Message that these rates will continue for one year. Obviously governance can change them if needed but ideally this is avoided and providing some long term clarity is important.

This plan would accomplish a few goals. 1) Net an immediate 40% reduction in daily POOL distribution 2) remove some yield farming whales and increase odds for smaller depositors to win 3) create a clear single largest prize pool to rally around to get to the largest possible weekly prize. View the details here.

Looking Forward
These changes to POOL distribution ultimately lay the foundation for our next push but get to $1 million in weekly no loss prizes. They are important but I don’t want to lose sight of this bigger goal!

I look forward to all feedback and counter proposals! We should have the goal of getting a proposal up for governance vote 8 days from today.

20 Likes

Based on your spreadsheet, to run this would cost an annual total of 1,387,100 POOL, around 25% of the Treasury. (This includes the 500 POOL per day to future pools)

Would we keep the 300 POOL per day in LP incentives over the course of a year in Uni v2? Is there any issue with keeping liquidity in Uni v2 and not migrating to v3?

I think changing the reserve rate of the USDC pool with this proposal is a mistake and goes against perpetual growth. Reducing the reserved rate to 10% when this launches may provide bigger prize headlines in the short run but slows down the stablecoin ‘vacuum’ we have implemented. This POOL distribution is outlined to run for a year, therefore still attracting liquidity to the pool and growing the reserve. I think it is important to grow the reserve as much as possible while we have the yield farming incentives.

If we want to have a goal of big prize headlines I think that should be a separate proposal in the future. Let the reserve grow more and we will have consistent big prize headlines.

5 Likes

Based on your spreadsheet, to run this would cost an annual total of 1,387,100 POOL, around 25% of the Treasury. (This includes the 500 POOL per day to future pools)

Correct, that would be the maximum upper bound. hard to model the 500 though because we have no idea when it will be implemented. I think keeping inflation relatively high year 1 makes sense as we are still establishing ourselves.

Would we keep the 300 POOL per day in LP incentives over the course of a year in Uni v2? Is there any issue with keeping liquidity in Uni v2 and not migrating to v3?

I anticipate no. We will likely want to move it to V3 once liquidity mining is easy there.

I think changing the reserve rate of the USDC pool with this proposal is a mistake and goes against perpetual growth.

Definitely valid reasons for keeping it high. And perhaps we keep it higher on the Dai prize pool than the USDC. I wasn’t able to get into it too much but underlying my logic here is that I think we should make an all out push to get one of the prize pools to the largest possible headline prize ($1 million weekly!). Ultimately, large prizes are the key differentiator of the protocol so delivering on that unique value proposition is really important. That will establish us as THE player in the space.

9 Likes

Getting to $1 million dollars weekly will definitely draw the most attention and headlines. It makes sense to concentrate on one pool to get there. Given the higher yield on USDC, choosing it over DAI is rational. Unfortunately, the way the media operates they won’t care about a “measly” $160,000 weekly prize that’s fragmented between 2 pools. They want something clean and big like $1,000,000. I think our growth is incredible and I’m not downplaying anything, I just think that’s how media outlets would perceive this. I’m all for “an all out push” to increase one of the prize pools.

Decreasing the reserve rate from 50% to 10% will help us get higher weekly prizes, but I share @McOso’s concern, it kind of undoes the perpetual growth machine we have created. Of course, I see the benefit in lowering it and how that 50% reserve rate might even be bogging us down a bit and maybe even hampering that reflexive growth. The fact that we’re still getting these large prize pools with a 50% reserve rate is awesome! PoolTogether is reflexive and higher prizes will attract more users and deposits. Now is probably the time to focus on the reflexivity and lower the reserve rate, though. The compromise to keep the DAI reserve rate high is a nice little addition.

3 Likes

Can we separate these proposals when they go to vote? I don’t like the idea of having to vote on the POOL distribution and the change in reserve rate in the one vote.

4 Likes

Overall I agree with most of what’s suggested. I think the two points I am unsure about are:

1). USDC reduction to only 2000/day
2). Reserve rate reduction from 50% to 10% on USDC and DAI pools.

I talked in the discord about having good inflation and bad inflation. Good inflation offers incentive for people to buy and hold POOL while bad inflation is just inflation that goes straight to the market. We do need bad inflation of course since that is the motor for driving TLV.

Right now, we are contributing 400 POOL/day towards good inflation through the POOL pool and AMM campaign while 2900 POOL/day is directed towards bad inflation (not including the 500 POOL reserved for future pools). I would prefer to divert some of the bad inflation towards good inflation. Increasing incentive to hold POOL and reducing POOL coming to the market through bad inflation will lead to a higher market price which could either offer a higher or lower APY to yield farmers; there is a sweetspot somewhere. I suspect we are far from the sweetspot given the distribution is significantly skewed towards bad inflation.

I’m going to give everyone a hypothetical example to consider. Taking the USDC pool that has 79,779,971 TLV today, IF the liquidity mining campaign was 100 POOL/day less than what it is what impact would that have had on the APY. Over 14 weeks this would equate to 9,800 POOL. Bearing in mind the spread for selling or buying 1,000 POOL on Uniswap is nearly $1, let’s assume that the market price would be $1 higher than it is right now. Without doing some complex matlab analysis I think this is a fair assumption to make and most would accept it as a reasonable, even conservative, estimate.

Current:
2295.9x(20.42) / 79,779,971x365x100 = 21.449032% APY

Hypothetical 100 POOL/day less and assuming a $1 higher price
2195.9x(21.42) / 79,779,971x365x100 = 21.5194425% APY

So you see, a reduction in the yield farming reward can even be in the yield farmers interest, excuse the pun.

I think 2,000 is too high and probably counter productive. I do like the idea of staggering it down slowly as someone in discord suggested when we were having the discussion a few weeks back.

7 Likes

I think reducing emphasis on one of the stablecoin pools and focus everyone on using one bigger stablecoin pool makes sense. If prizes get big enough, then I could see Twitter discussing the weekly prize as a semi-big event each week.

5 Likes

A couple thoughts on reserve rate:

  1. How much more is a $1 million headline prize really worth, compared to a $650k prize? Sure we might get a couple news headlines when we hit it, but after that?

At the current rate we’d be giving up around $80k/week to reach it by reducing reserve rate to 10%. We could spend (part of) that money on marketing campaigns, which would potentially give us more exposure than a headline prize that is 1/3 larger.

  1. As others have noted, a reduction in the reserve rate gets in the way of our perpetual growth mechanism and makes the treasury smaller.

I would argue that treasury growth also goes into the category of “good inflation”, as it increases the fundamental value of a POOL token. As @ageless pointed out nicely, POOL price is a big lever on yield farming APRs.

  1. Even with a 50% reduction, yield farming will still be the majority of return when depositing into Pooltogether. Even more so, if we see a POOL price increase as a result of reduced inflation.

In my eyes, we’ll still be in the “aggressive yield farming” phase and are in a prime position to keep reserve rates high.

11 Likes

Wow, some strong arguments back and forth in here!

For the record, im leaning in favor of sentiment to keep the reserve rates high. The treasury needs to be grown as much as possible in this early stage of the growth of PT to build a strong foundation of underlying capital for the Pools.

I prefer the ‘slow and steady’ approach to $1 mil via growing the reserves, rather than incentivizing fickle whales. While I agree that $1 mil weekly prizes is an awesome headline and will drive users to the protocol, marketing the protocol can also be done in other ways.

That being said, the general structure of 80/20 POOL yield to sponsors and regular depositors is a good adjustment to the stablecoin pools and recognizes that different users are in the Pools for different reasons.

6 Likes

I agree with most of this. I echo the sentiments of others worried about lowering the reserve rate. everything else looks good to me. A reduction to 10% feels very drastic. I would like to keep it at 50 for now. if we want to lower it then 10% feels like much too large of a drop. I think that keeping the reserve rate at 50% is a better long-term strategy. Also, if we reduce the reserve rate that much it is possible that we do not accomplish the goal of the 80/20 drip split. If the prizes shoot back up then some “medium whales” may not move their funds to sponsorship. they may be more enticed by the prizes.

3 Likes

PTIP-13: Treasury Diversification has wide support as evidenced by the earlier vote, so I believe it will pass. This means that PT will soon have about $6m USDC in the treasury. That’s enough runway to build for years.

I’m going to split my feedback and address the reserve rate and distribution rates separately.

Reserve Rate

The reserve is intended to do two things:

  1. Capture value for the protocol
  2. Retain funds in the pools for perpetual growth.

We will soon have a massive treasury, so #1 isn’t pressing right now. Let’s consider #2:

A reserve rate of 50% is currently capturing ~$21k per week in the USDC pool. That $21k is essentially sponsorship until it is withdrawn. Per year it’s about $1m.

Let’s say the reserve was constant throughout the year to keep the math simple. This means that the APR was boosted by the sponsorship by: $1m / $80m or 1.25%.

So an APR of 5% will effectively be (5 + 5*1.25%) or 5.0625%

The reserve’s actual contribution to perpetual growth is negligible

So that strikes point #2. This leads me to believe that we should drop the reserve rate to 0% to maximize prizes and growth.

POOL Distribution

I want to provide feedback on some of the above points:

Reduce POOL distribution to low yield non-stablecoin assets by 80%

100% agree with this. We should still have emissions, but lower them as they not very valuable.

Reduce POOL distribution to USDC prize pool by 13% and Dai by 65%

I agree with rebalancing them, but I’d like to propose something else:

This tells me that USDC is basically 3x bigger than Dai in terms of the addressable market. The APR between them is very similar, so it seems to me we should reflect the POOL distribution according to the market size.

I propose we redistribute the current emissions of USDC + Dai at a 3:1 ratio to match market size.

Currently the combined POOL emissions is 2295 + 2295 or ~4590 POOL per day. At 3:1 each pool would receive:

USDC: 3442.5 POOL / day
Dai: 1147.50 POOL / day

I think this distribution better reflects the addressable market.

split POOL distribution 80% to sponsored deposits and 20% to regular deposits

The intention here is to separate farming from prizes: in theory whales will migrate to pure farming so they won’t be eligible for prizes.

You can reduce this idea down even further: it’s a choice between consistent APR and prizes.

This is exactly what Pods introduce: a spectrum between consistent APR and prizes. More people will win.

Initially I was a proponent of surfacing the sponsorship mechanic, but now I hesitate as it adds complexity to solve something which will likely be addressed by Pods.

My inclination here is to instead surface Pods as a way for users to win more, and let all users earn the same amount of POOL.

This does have some benefits:

  • there won’t be any liquidity thrashing from the farming migration, which should reduce the impact on TVL
  • Promoting Pods is more sustainable long-term, as it’s core to the product. If emissions stopped completely, there would still be value in being eligible for the prize.
  • Much less complexity!
6 Likes

I have to disagree with lowering the reserve rate to zero. that is the only source of revenue that the protocol has. The numbers you list are as of now. what if the TVL goes to a billion dollars? that reserve would be about four times more than it is right now. if the reserve is zero then the money we receive from PTIP-13 will slowly bleed away, and we will not be replacing it with anything.

4 Likes

Overall, as @ageless pointed out. I think reserve rates should be settled in a different discussion. It was a mistake for me to put the two together. I did want to point out though that your numbers are not correct.

A reserve rate of 50% is currently capturing ~$21k per week in the USDC pool.

Since the 50% reserve has been implemented USDC reserve growth (inclusive of this week) is averaging $46,000 so this is actually $2.4 million per year and that is assuming the protocol doesn’t grow at all. I think we can all agree that is a very pessimistic assumption :). I do understand in the near term the reserve doesn’t radically impact prize sizes. I also think there is an argument that optimizing for larger prizes over reserve growth makes sense early on. But I wouldn’t discount the importance of a reserve either.

Regarding your suggestion on POOL distribution, I think it’s a good angle to think about for how to allocate between the two stable coins but your suggestion doesn’t actually lower POOL distribution at all. I think a substantial reduction in overall POOL distribution is a non-negotiable here.

3 Likes

Good points! I mentioned this in a separate comment but I’m going to edit out my thoughts on reserve rate.

We should handle one governance issue at a time and although reserve rate & POOL distribution have impact on each other it would be much more clean to JUST handle POOL distribution and then revisit reserve rate if it is needed.

3 Likes

The numbers you list are as of now. what if the TVL goes to a billion dollars? that reserve would be about four times more than it is right now. if the reserve is zero then the money we receive from PTIP-13 will slowly bleed away, and we will not be replacing it with anything.

This next round of funding is $6m USD. It will take a long time to spend that money. At our current burn rate this money would last years.

Priorizing growth so that we can achieve $1b is our goal!

Maximizing prizes will help us get to $1b TVL, at which point we can enable the reserve.

2 Likes

I like your reasoning here and I agree – at the end of the day POOL distribution doesn’t matter to depositors, they only care about effective APR so optimizing for the highest effective APR with lowest emissions make sense.

I also like the idea of staggering it down but I want to be careful with that. Gas costs are insanely expensive and people want a level of certainty when they make deposits. We may need to make another adjustment after this initial one but I’d hate to make multiple in the coming months.

I’m brainstorming a bit here but one possible option here would be to publicly target an effective APR on deposits. We could then adjust distribution rates every ~30 days to true up to that APR target.

Overall, I’m open to reducing the USDC POOL distribution a bit more but I do think it’s important to preserve one prize pool that is the “primary one”. I also think an advantage of this approach (making a major reduction to Dai and a much smaller reduction to USDC) would be a good A/B test.

4 Likes

Since the 50% reserve has been implemented USDC reserve growth (inclusive of this week) is averaging $46,000 so this is actually $2.4 million per year and that is assuming the protocol doesn’t grow at all. I think we can all agree that is a very pessimistic assumption

Growing the pool size doesn’t really matter; it’s the proportion of pool that is reserve. Even doubling the reserves would only push the previously calculated effective APR from 5% to 5.13%.

We’re optimizing for large prizes so we shouldn’t kneecap that growth by capturing any portion of them for reserve. We will have plenty of funds to operate with.

I think a substantial reduction in overall POOL distribution is a non-negotiable here.

Why is it non-negotiable? I don’t see the logic for this.

Edit: had an out-of-band conversation with Leighton:

He explained that the general idea for lowering distribution is that it would let the treasury last longer and reduce the sell pressure on the token.

Makes sense to me; and my follow-up thought is that we should increase the LP incentives to stabilize the APR for farmers.

3 Likes

Great! I agree with this proposal. I think it will contribute to the long-term growth of this project.

2 Likes

I think that the size of the pool matters. That’s why TVL is such a commonly used metric in defi. Imagine our TVL reaches 10 billion. (Most likely will happen at some point) and let’s assume we make even 5% on that. That’s 500 million per year in interest/52 weeks = ~9.6 million dollars per week. Even if we only take ten percent in reserve rate that’s close to a million dollars per week in protocol revenue. Additionally in the traditional financial world in a large market like bonds a 13 basis point increase is statistically significant. I don’t think looking at the effective Apr as you have here provides enough context. Think about what the protocol could do with a million dollars every single week. It’s not necessarily about do we have enough to operate. In my mind it’s more about can we earn enough to make a difference in the world? PoolTogether foundation for charity could make a difference with that kind of money. There could be a PoolTogether stadium for some sports team. There could be a PoolTogether University. There are so many possibilities. On top of this I think investors are more interested in protocols that generate revenue.

I respectfully strongly disagree with a zero percent reserve rate.

6 Likes

This is brilliant and will help increase expected value for smaller fish. Why didnt we come up with this before? Overall I very much support the proposed changes but am also in favor of some reserve capturing.

3 Likes