Time for a POOL Party! (Part 3)

The first two posts outlined the strategy to grow the protocol by focusing on prizes.

This third post ties it all together! How the POOL token can create the biggest prizes by enabling more yield sources, asset types, and blockchains. The POOL token can unlock PoolTogether’s potential to be a prize savings machine that powers the growth flywheel. :chart_with_upwards_trend:

The strategy is to distribute all prizes in the POOL token itself. :exploding_head:

This enables:

  1. Depositors to save in any asset
  2. Depositors to get rewards in any asset
  3. The largest possible prizes

How this POOL-based prize system works:

To have the largest prizes, we need two things 1) the largest amounts of deposits 2) yielding the highest returns. Right now, the protocol is highly limited because we only support one asset (USDC) and one yield source (Aave).

To transform the protocol into a prize savings machine we need both better yield and the ability to take in more assets. But there’s a problem… if prizes are distributed in the asset deposited, it breaks the prize machine. Each prize pool only has its own small prizes. On top of that it’s a technical nightmare.

The solution is to distribute all prizes in the POOL token.

By doing this, the protocol can easily accept any yield bearing asset (wBTC, USDC, Dai, Ethereum, Matic, etc.). Regardless of what asset you deposit, you are contributing towards one massive prize! The prize can still be displayed in dollars (as it is now) but the actual distribution will happen in POOL tokens. After claiming prizes in POOL, depositors can hold and use their POOL or swap into any other token they want.

The protocol can now scale deposits quickly with massive prizes by allowing many assets and yield sources!

This sounds great but there are a few important questions:

  • What happens to the yield?
  • What happens if people don’t want POOL?
  • How would different assets and yield sources be fairly balanced?

These questions are all related.

Let’s assume the market value of the POOL distributed should roughly match the value of the yield accrued. So if there is $100,000 in total yield accrued then there is approximately $100,000 of POOL distributed. We want to make sure that anyone who receives POOL can easily swap it out for any other asset they might want. We also want to make sure the protocol gets POOL back so that it can keep distributing it. :repeat:

What the protocol can do is let POOL holders return POOL and get back assets the protocol holds at a discount! This discount starts 0% and then increase until the assets are gone.

By letting people get the yield at a discount it ensures there is always demand for POOL and this ensures depositors ultimately get savings rewards in ANY asset.

That answers what happens to the yield and how people can swap out of POOL. The third question is, how would pools be fairly balanced?

If the protocol supports many assets there is a problem. How do you give depositors fair chances? The chances of winning need to be weighted in accordance with the yield generated. If not, the protocol can easily be gamed.

This is an extremely difficult technical problem because it requires an “oracle” for the protocol to use.

The solution is to use the POOL token to weight the odds.

POOL holders can stake their POOL to increase the chance to win for their favorite prize pools. It’s in POOL token holders best interest to create the most yield and therefore the biggest prizes so their incentives are aligned. They will stake their POOL on whatever assets and yield sources are creating the largest prizes. By staking their POOL they will make that prize pool have better odds thereby driving more deposits to it.

What does this mean for the POOL token?

POOL becomes the fuel that powers the prize savings machine, while still serving its core function as the governance token.

By saving your money, you get POOL tokens, you can hold and use those tokens in the protocol or you can swap them for any other assets.

PoolTogether would allow users to deposit in a prize pool, receive the chance to win, and always get back their original deposit. Using POOL as prizes allows more users to deposit a variety of new assets, yield sources, and blockchains generating prizes of increasing sizes.

“POOL wars” will kick off as yield sources and asset creators want to drive usage. Imagine, FEI, Dai, MStable, Dolla, and USDC all competing to stake the most POOL and drive the most odds to their assets.

Next Steps

This concludes the high level strategy behind how the POOL token can improve the protocol and kick off the growth flywheel. This post serves as a primer for the topic but the devil is in the details. How this can be implemented?

If the community is aligned on the desire to pursue using the POOL token in this manner, PoolTogether Inc will propose to build out the infrastructure to support it. Let’s discuss! :ocean::trophy:


Awesome proposal! I feel like being able to deposit any token really helps towards becoming the perfect DeFi savings account.

I do have some thoughts though regarding the implementation of POOL returns and prize allocations though, that can hopefully contribute to future discussion:

1) The voting system for prize allocations seems to create more incentives for depositors to vote for the asset that they have deposited, rather than the asset that actually provides the protocol the most yield. My fear is that in such a system, especially with other protocols building on top of this, the amount of prizes we can give out would decrease in the medium to long term due to the majority of prizes going to a less efficient asset (leading to more users wanting to deposit that asset, and so on…).

2) Would there be any mechanism in place to stop protocols and/or automated scripts from simply instantly returning POOL for the pool’s underlying token as soon as there is even a minuscule amount of profit to be had? This may be from my misunderstanding of how the returning functionality works, but it would seem that following the basics of game theory would essentially mean that such a mechanism would be borderline unusable for regular depositors. Of course given enough liquidity users can still trade POOL for any asset of their choice, but that would incur fees and also create sell pressure for the token.


Excellent questions @Ncookie!

1) The voting system for prize allocations seems to create more incentives for depositors to vote for the asset that they have deposited, rather than the asset that actually provides the protocol the most yield.

Liquidations are essentially buybacks, and with less buybacks the POOL token becomes less valuable, which means the value of the prizes goes down. The assumption here is that POOL token holders want the biggest prize possible, so they’ll vote for pool that produce the biggest returns.

That being said, we could have the voters for a prize pool earn a portion of the liquidated POOL tokens. Let’s say 10% of liquidations go to the voters. This will help align voters, because:

  1. They earn more ownership as a result of their voting
  2. Good voting produces more buybacks, keeping the POOL token healthy.

I need to do more investigating as to how complex that would be to build. As long as the system is modular, we could ship the MVP first then add that feature if it proves needed.

2) Would there be any mechanism in place to stop protocols and/or automated scripts from simply instantly returning POOL for the pool’s underlying token as soon as there is even a minuscule amount of profit to be had?

Truth be told, we expect bots to be capturing liquidations. This is actually a good thing; bots mean that the arbitrage is as efficient as possible. In the liquidation simulations we achieved 99% efficiency, which means that we are maximizing the value of the buyback. The expectation is that this volume will create healthy markets as a side effect.


So does this mean that essentially the POOL token price would be tied to interest rates?
Meaning that if DeFi interests rates go down, so does POOL (because less prizes are being generated). Would that be the case?


I don’t want to speak for Brendan, because he will likely have a better answer, but from my understanding of the system we’ve been discussing is there will definitely be a coupling between POOL price and assets available for auction.

However, generally the pressure should remain upwards for the long-term.

In the short term (literal seconds) there will be pressure to buy POOL for the asset auction. That’s the intended goal. To have a healthy protocol, where value is constantly circulating. Stale assets that sit for too long are not optimal and generally lead to breakdown of systems.

Sooo while the result is regular, upward pressure for POOL, the objective is not that. The objective is to ensure the system remains healthy with deep liquidity POOL<>Token pairs so Depositor’s can liquidate their Prize positions. If desired.

I’m not nearly smart enough to get all this math, but Vitalik published an article several years ago, about why if we choose to take this route we NEED to ensure constant circulation.

Now the long-term thinking…

For example let’s say you buy POOL for $4.00 and trade it for 4 USDC at a price of $3.90 in auction in real-time. No problem. You’re a good arbing bot.

But let’s say you acquired POOL when it was worth $2.00 in the open markets (ideally through the PrizePools). You could now be getting 8 USDC for $3.90 at auction, because you acquired POOL when it was much cheaper.

Thus, the incentive is both available for short-term arbers and long-term savers.

We’ve again created a dual-side market, where both parties are serving each other. Arbers create constant demand for POOL. And depositors help create long-term and DEEP liquidity pools for the necessary volume to be circulated.

If interest rates drop, sure yes their will be less short-term demand for POOL. Maybe effecting the price. Maybe not.

However, there is always the pull in the other direction to SAVE. To compound your investment and to have the opportunity to liquidate for other assets when the system is running at the most optimal levels. Generally, there should always be a demand via the auction, even with low interest rates. It just becomes a matter of scale.

This long-term saving potential should ALSO bring demand for POOL, because ideally you’re always buying assets for a discount with POOL if we reach an optimal PrizePool Network size.

Any ways… that’s at-least my understanding of the proposed system and how we can balance both the short-term executing and long-term vision to create a healthy protocol via price stabilization and thus the system as a whole.


High interest rates => high inflow of POOL (from liquidations)
large prizes => high outflow of POOL (from prizes)

Adjusting the prize size will be a critical monetary policy for PoolTogether. As interest rates drop, we will need to lower our prize size to balance POOL flows. If they increase, we can increase our prize size to balance POOL flows. If we wish, we can adjust the prize size to have greater outflows, or greater inflows.

PoolTogether becomes a prize machine based on token flow.

1 Like

How is this discount being paid for?

1 Like

Good question!

Probably easiest to think of it as “slippage”. If the protocol wants to swap assets it holds for POOL tokens there is going to be some degree of slippage to facilitate that. So you could think of the protocol “paying” it but it’s more just like slippage any time any trade is made.

1 Like

Thanks for the answers, the Vitalik article you linked is very insightful, some paragraphs I had to read them a hundred times but eventually I think I got it haha

For this to happen, the amount of value you can swap your POOL tokens for has to increase over time, and to my understanding, this happens when the revenue by the prize pools goes up over time, meaning that the opposite can happen too:

-Buy POOL at $2 (Currently can be exchanged for 2 USDC)
-Revenue of prize pools decreases (either because of a decrease in TVL, or lower interest rates)
-Meaning less Buybacks
-Now that same POOL can be exchanged for 1 USDC (and you bought it at $2)

So essentially, in this scenario, when you hold POOL, you are betting that TVL is gonna increase over time (good incentive) but that could easily be off-set by a decrease of interest rates at the same time, those being completely outside our control (Say the prize pools have a TVL of 100M earning 4% APR = 4M a year in buybacks, then because the protocol is doing well, TVL increases to 150M, but the interest rates dropped to 2% = We are down to 3M in buybacks, thus POOL value drops)

However, isn’t your point in direct conflict with @Brendan’s response?
He says that whenever prize pool’s revenue goes down, we can reduce the size of prizes (reduce POOL supply) and thus bringing back the price, then when revenue goes up, we increase POOL supply by growing prizes, thus leading to a somewhat stabilized POOL price (assuming that’s the goal of our monetary policy)

So then, what you said would never happen, as the token price would presumably never go up or down in the long term, in this scenario the incentive to hold POOL would be instead to earn a % of liquidations (Resulting in alignment with TVL growth, as other factors would presumably be mitigated by the monetary policy)

That said, responding to Brendan now:

Given the critical impact that adjusting prizes can have in POOL’s price

Who will executing those decisions?

An obvious answer would be simply POOL governance, however given the critical aspect, these adjustments have to be fast and precise, governance is simply too slow for that, so does this mean a new working group has to be formed? Such group would have immense influence over the token’s price (because that’s their goal) and therefore immense trust is required. Governance could somewhat limit what they can do but then that would be counter-productive as it would also limit our ability to quickly respond to market fluctuations.

And what market factors would PT have to consider in it’s monetary policy?
How it would track them?

As I explained earlier, TVL and interest rates are key factors to the prizepool’s revenue, and therefore our monetary policy.

Are we gonna track TVL across this many different blockchains? If we choose to ignore it, then each time our TVL goes down (or up), so does the POOL price (I would be ok with this! Just pointing that it’s an “extra layer” of volatility that wasn’t before present, at least not in such a direct way).

And interest rates? The goal of this strategy is to not rely in price oracles, but wouldn’t we need some kind of “interest oracles” instead? This is basically the same question I had with DPR, but unlike that case, we cannot just “set and forget” this parameter as it’s crucial to the token’s health, it’s data that needs to be constantly pushed if we don’t want it to constantly affect POOL’s price.

On other note:
When we make prizes smaller, we reduce POOL outflows as Brendan says, therefore reducing supply and therefore (one would guess) pushing the price upwards.

So if we start with 10 POOL prizes and then we adjust them to 1 POOL, then 0.1, 0.01, 0.001…

It would seem like we could raise POOL price to infinity.

I believe here is where appears another (crucial) factor that Vitalik describes in the article, it being the velocity of money, or “holding time”.

Basically when we “artificially increase” POOL’s price in this way, nothing really changes for depositors, the amount in US dollars they receive it’s the exact same, it’s just denominated in less POOL, the equilibrium is kept.

However, to POOL holders, they just got big (or weren’t that affected by a downwards price trend) so then, if suddenly many decided to sell, they could deplete the protocol’s liquidity in that moment, pushing the price downwards again, this is because their POOL is now valued above what it “should” be (hopefully I’m being coherent here). Essentially, if for any factor a “POOL run” occurred in moments like those, we would be much more vulnerable. We could require a locking period for staking, but then you have to deal with non-staked holders.

So then, that’s another factor I believe a monetary policy would have to consider when doing prize adjustments, and I’m not sure how we would even measure it.

Overall, I feel like the monetary policy it’s actually the most important part of this new strategy, and I would really like to see that aspect expanded on. Especially to see defined what would be the exact goal of having a monetary policy, if its goal is to keep all these factors from having a direct impact in POOL’s price, then perhaps it would make sense to “off-load” many of these functionalities over to a secondary token minted solely for this purpose, then we can let the market forces act freely on it instead.


I always appreciate Vitalik’s articles. The low-low-level thinking often goes over my head, but I think the big takeaways help ask the right questions about new tokenomics systems. The microeconomics lens helps distill some of fundamental elements that are needed in a sustainable tokenomics system.

For this to happen, the amount of value you can swap your POOL tokens for has to increase over time, and to my understanding, this happens when the revenue by the prize pools goes up over time, meaning that the opposite can happen too:

True. It’s a double edged sword. It changes User expectations about how prizes work. I suppose there will probably be a best “game theoretical” time to convert POOL for the desired underlying TOKEN, but for most people that’s not going to matter.

Regarding the example scenario you presented.

A gut feeling, but I would bet the protocol will need to use Protocol Owned Liquidity to hedge against market inefficiencies and rapid price movements still effecting crypto markets. I don’t know exactly what that would look like, but it would probably have similar properties of Aave’s Safety Module. The protocol can provide guarantees (even with shortfall event) about liquidity to Depositors. And that position only strengthens overtime.

Just as a hypothetical… but to purchase Assets in the Auction the protocol could require a POOL<>USDC LP token. Buying assets at auction would mean the protocol owned liquidity for Depositors is always growing. Arbing means creating an ever increasing floor for Depositors. Generally speaking I think peak times of interest capture would offset times of low interest capture over a long enough time.

He says that whenever prize pool’s revenue goes down, we can reduce the size of prizes (reduce POOL supply) and thus bringing back the price, then when revenue goes up, we increase POOL supply by growing prizes, thus leading to a somewhat stabilized POOL price (assuming that’s the goal of our monetary policy)

I don’t think it’s a contradiction. Probably just a differ understanding of the same problem. Yes, we can change the size of Prizes directed towards PrizePools. However, with the new gauge model this process could potentially be slow moving and likely non-optimal in the early stages.

For high-traffic PrizePools (e.x. USDC) the incentive is strong for quick rebalancing.

But for new asset types and growing PrizePools this incentive might not be strong enough.

I agree the ideal situation is a fluid rebalancing of POOL to all the gauge target addresses. But in a MultiChain world with a lot of moving parts I personally don’t see a clear path forward to go from 0-1 yet.

1 Like

My god, this is a good idea!!!


awesome curation, thanks for sharing the info