The Liquidity Conundrum

Apologies folks, but this is going to be another long one…

Ever since analysing the profitability of our pools here, I’ve been unable to shake the feeling that we’re doing something drastically wrong. That $500K/week figure has been rattling around my head all weekend. After digging a little deeper, I’m kicking myself for not doing this sooner.

We’ve had a string of good news (new strategies, audit, WETH rewards, marketing push etc.) but still seem unable to cross $150M in TVL on a sustained basis and the price of PICKLE continues to stay sub-30. After a lot of discussions in the Discord and a bunch of math (to follow), there’s a simple conclusion to be made:

We have misallocated our PICKLE rewards and they remain too heavily skewed to the Power Pool. As a result, we do not have enough rewards available to incentivize fresh TVL to come over to us. We launch a new strategy, people pile in, APY drops, and we stagnate.

Before pulling out the pitchforks, please remember that I’m heavily invested in the Power Pool myself, so I hope the LPs will hear me out :blush:

What is the function/purpose of Pool2?

At a protocol level, the basic and ONLY function of the power pool is to provide an exit ramp for users of the protocol. Basically, our customers (Jar depositors/TVL) should be able to sell their farmed PICKLEs for ETH with minimal slippage. So, let’s consider that for a minute.

Our current weekly emissions are 27458 PICKLEs/week or 3922/day.

Every day, 2475 PICKLEs are earned by Pool2, and just 1447 PICKLEs are shared between ALL the pJars.

This shows us that even the biggest whales in our Jars can’t be responsible for the large trades we’ve been seeing. For example, even having $18M in 3pool (a la Harvest) only yields 118 PICKLEs/day. That’s less than $3000 per day of selling volume, which is a bug on the windshield to our 23M of liquidity.

Consequently, the large trades (2000+ PICKLEs) have to be coming from one of two sources:

  • Large traders who are using our liquidity to speculate, with no real interest in PICKLE. It’s just another token for them to trade, and our deep liquidity lets them move large size with minimal slippage.
  • Large participants in our own power pool, who are accumulating and dumping because they have no incentive to stake, considering that the staking APY is half that of the LP.

With me so far? Before moving on to the numbers, here’s a brief primer on how Uniswap works for those who may not know.

Uniswap is really elegant in its simplicity. It works by pooling 2 assets 50:50 and maintaining a Constant Product between the quantities of each token in the pool. That sounds more complicated than it is. For the purposes of this discussion, it’s enough to know that the movement in price caused by a trade is directly related to the size of the trade relative to the size of the pool. You can read more here.

Basically, big trade make number move more :stuck_out_tongue:

Now you’re wondering…all this to tell me something I already knew? Please bear with me…

Liquidity Analysis of the Power Pool

I used the Uniswap equation to model the levels of slippage (price movement) that would be expected at different levels of liquidity. For simplicity, I’ve ignored the Uniswap 0.3% swap fees, as that would be a constant (in relative terms) in all the scenarios. Here’s what I found:

As you can see, with our current $23M of liquidity, it is possible to buy or sell 10,000 PICKLEs ($250K) while incurring slippage of only 2%. Put another way, it takes 10,000 PICKLEs to move price 2%.

We often joke that PICKLE is our favourite stablecoin, and this is the reason why. It takes a LOT of buying/selling pressure to move the price.

We have suppressed our own price volatility by maintaining (and rewarding) far too much liquidity in relation to our needs. Not a single user of the PICKLE ecosystem has a requirement to trade these quantities.

Our TVL/Liquidity ratio is 123M/23M = 5.35. That’s way too low IMO.

Now, contrast this to Barnbridge, who currently have the opposite problem. Their rewards are skewed towards their stablecoin pool, and as a result there isn’t enough liquidity in their Pool2 to absorb all the selling from their weekly harvest. TVL/LIQ = 273M/5.4M = 55. Way too high, hence the weekly dumps.

Harvest seems to have found a good balance between offering higher yields to users, and enough liquidity to absorb average daily volume. TVL/LIQ = 400M/17M = 24

If we apply Harvest’s ratio to Pickle, we should only be looking to maintain $5M in liquidity.

Even at just $5M in liquidity, 5K PICKLEs can be traded with 5% slippage.

The vast majority of our users trade far less than that, and trades of up to 1000 PICKLEs will still only incur slippage of 1% or less.

By reducing our liquidity, we accomplish the following:

  • Disincentivize large speculators from playing in our pool…if they want to trade big, they have to pay up. Buying 10K PICKLEs will move the price 10%, not 2% as it does now.
  • Have PICKLE rewards to allocate to profitable strategies that earn fees. Currently, 70% of our dwindling PICKLEs are generating no fees. There are only roughly 300K PICKLEs left to be minted, and we’re currently directing 19K EVERY WEEK to maintain liquidity that we aren’t using. Our average daily volume is less than $2M, and having 10X the liquidity is an unnecessary waste of precious resources.
  • Shake out opportunistic liquidity. The reduced number of participants in the Power Pool will enjoy a higher APY, whilst the additional fees generated by putting those PICKLEs to work will increase the funds available for stakers.
  • The goal is to get LPs to move to staking, taking those PICKLEs out of circulation and creating a supply squeeze. The only way to do this is to maximise staking rewards, which is derived from the additional fees generated. We make it attractive to ‘Stake and Stack’ over ‘Farm and Dump’.

Fee Projections

Here are some scenarios I worked out, diverting 5, 10 and 15% of PICKLE rewards to the pJars

Since we’re already discussing pDAI and REN, I used these pools as examples.

The numbers assume that our current APYs are our stagnation point, and so the projected TVL is what would be required to bring the new boosted APYs back down to current levels.

As you can see from the data, we can monetize our remaining PICKLEs in a big way by reducing liquidity.

More fees -> Higher staking yield -> Higher PICKLE valuation, which benefits all of us.

LP yields are already converging towards staking (103 vs 57), and after a few weeks of dropping emissions, a higher PICKLE price is the only thing that’s going to support the ecosystem.

I think $5M is too low since we want to keep a buffer for stability…but around $7.5M seems to be a sweet spot. That’s a 65% reduction from current levels. Just imagine what we could earn in fees with another 20-30% of PICKLE rewards being allocated to the pJars.

I started working up a plan for gradual reductions to try and minimize price disruptions but realized that this would be a very tough proposal to pass, so I’ll wait to see the reactions from the team and community :blush:

Thanks for reading! Feedback and criticism welcome!

Bring on the pitchforks!

Here’s the Excel file if anyone wants it :

11 Likes

Here’s the raw analysis, please feel free to check my math.

Current

$15M

$10M

$7.5M

$5M

3 Likes

Excellent analysis! It would be great to hear from the Devs about this issue. Was there a logic behind this allocation that we are missing?

At the beginning (the hyperinflation phase) it’s important to have very deep liquidity because of farming and heavy dumping. To incentive people to keep providing liquidity, lots of pickles were necessary. As we mature as a protocol, that need becomes less and less. We’ve arrived at a point where the Pickle emissions could bring more value elsewhere.

5 Likes

Fully agree!
I believe LP rewards were originally set at 50% and were boosted to 75% to curb exceptional volatility in the early days? That was before my time, so I’m not sure what the prevailing sentiment was.

I fully agree with the sentiment of this analysis and strongly urge the community to carefully read this analysis and consider reducing rewards to pool two. This will be a contentious vote as currently only users in pool two can even vote and thus are likely to want to preserve their share of the emissions.

As @yyctrader mentions, this pool is just too large for anyone outside a whale’s needs and is suppressing price movement in either direction allowing for a great farming experience to the detriment of token holders. Additionally, this high allocation of emissions to pool two puts PICKLE the most PICKLE in the hands of those paying $0 in fees to the protocol.

Pickle emissions to jars at a higher rate should make those jars more attractive to new TVL, existing farming protocols like Harvest, and will benefit the existing jar users who are the backbone of the ecosystem. I know there is a camp that will argue that pool two is that backbone – however I urge you to question why that is the case? Pool two solely provides an on and off ramp for users and while that is a necessity, there are many other projects out there with non-incentivized pools, or with significantly lower MC / Liquidity ratio pools doing just fine.

3 Likes

Right. I agree with that but it doesn’t explain the specific allocations up to this point as other allocations could have been chosen. We also had the farming bonus initially and as @yyctrader indicated that we are way above what would be sustainable and above other projects so I’m wondering if there is something that we are missing in terms of the allocation. I think the Devs have a better understanding of the numbers from the beginning so it would be good to understand their plan a bit better. (I feel the “white paper” for Pickle is also severely outdated and uninformative)

Do we really have an idea of the size of Pickle holders? Given the length of the price stagnation, to me, it is possible that there is at least one ~100,000 Pickle whale in either farm, staking or combination. If possible, and this person were to sell due to a significant change, that would significantly impact price. Is this the reason for the allocation?

I’ve been here since the beginning as well. There was a need for deep liquidity and we needed a higher number, so 75% was chosen and it seemed to work. A lot of decisions were made really quick, because of the fast nature of farming.

2 Likes

I’m strongly in favor of reallocating some Pool2 rewards to jars. The analysis is incredible. Also, I’m apparently terrible at optimizing my profits because I moved from Pool2 to staking a few weeks back. I guess I just hate impermanent loss, and feel like staking is better in terms of predictability…

1 Like

Yeah I get that too (I have also been farming since the beginning). But it doesn’t explain why it’s still like that two months in and it appears Devs haven’t really been concerned. So I feel there is something we’re missing aside from needing deep liquidity initially.

Would love to hear from the dev regarding this new approach. :smiley:

1 Like

I think the analysis, albeit infused with a sense of numerical certainty, is biased towards existing systems. It is not about the math, it is about the assumptions.

There is a lack of exploration about the drawbacks of reducing liquidity (e.g. price volatility). I find it hard to think our community can put up with YFI-style volatility. Simply, too many people would get rekt and sentiment would take months to recover.

Naturally, it’d be very remiss of me to pass this opportunity to bring forth the Smart Treasury mechanism as a superior way of capturing protocol fees for price growth which is a goal shared by all PICKLE holders: passive and active. Of course, this was pointed out, but from the point of view that more staking leads to a better price which is debatable (remember, there’s ETH leaving the system). If this is a contentious vote, we should not assume that farmers are refusing to vote against their interests (other reductions have been approved). In fact, we now have more information, and different directions in which we can go to achieve protocol health and invest in growth.

3 Likes

I support the idea of reducing PICKLE emissions to the Pickle Power (pool2) in PIP-17 and I also support it here (but probably less so for now). I urge that we redirect emissions step by step. PIP-17 redirects 5% of emissions to our pJars (strategies) and in turn generate more fees. It will bring the current split to 65% to LP and 35% to pJars.

I say we take a moment to digest that change (assuming it happens) and then, if deemed successful, we could redirect another 5% bringing the split to 60/40 and keep liquidity over $10mill.

Other things to focus on:

  1. New pJar strategies? (UNI farming may be ending)
  2. Smart Treasury (as @leekuanjew points out and we have been discussing)
  3. UI / UX push
  4. Voting power for PICKLE Stakers

I feel strongly these other areas will also drive TVL, make the platform more profitable and mature the platform for all users.

4 Likes

Agreed, I should have spent more time on the negative effects.
TBH I wasn’t expecting a positive response from the community, which is a welcome surprise :slight_smile:

@leekuanjew Would you mind helping me with some volatility analyses? We can add it to the discussion to present a more complete picture.

This is a good idea and the numbers look good but there are so many variables in play, it’s impossible to predict the knock-on effects. We shouldn’t rush into it, and as @Mah_Dills alluded to, there are more pressing concerns. In particular, I think the end of UNI farming, and evolving the staking / value accrual mechanism need to be top-of-mind.

My impression is that this conversation is being driven by the perception that PICKLE is stagnating. For those who are worried about current ‘boring’ PICKLE price, let me just remind you that the fundamentals are already very promising. Maybe we just need to communicate this better. The rest of this post may be a little off-topic…


:cucumber: :rotating_light: :cucumber: :rotating_light: :cucumber: :rotating_light: BEGIN FREE ALPHA DUMP* :rotating_light: :cucumber: :rotating_light: :cucumber: :rotating_light: :cucumber:

We currently generate 290ETH weekly for stakers. That’s 15K ETH net profit returned to PICKLE holders annually in hard crypto. Not food tokens, not some wrapped pseudo-dollar, but cold, hard ETH.

Now let’s look at earnings multiples. Earnings multiples are a blunt tool but then so is a hammer, and like a hammer, earnings multiples are fine if you keep in mind their limitations. Also, literally everyone uses a hammer at some point when they open their toolbox, and the same goes for earnings multiples (and shaky metaphors).

PICKLE generates $6M at current ETH prices ($400 * 15K ETH) and is currently hovering around $20M market cap. This implies a multiple of… less than 4x(!). This is laughable. Once speculative (retail) money starts flowing back into Defi (as opposed to degen farmer money or predictable-returns-seeking whale money), PICKLE will look like an absolute no-brainer.

** Note: As pointed out by @dafacto these numbers are overly optimistic since the current rewards are boosted by accumulated treasury overflow. See my post further down for more realistic numbers. **

:cucumber: :rotating_light: :cucumber: :rotating_light: :cucumber: :rotating_light: END FREE ALPHA DUMP* :rotating_light: :cucumber: :rotating_light: :cucumber: :rotating_light: :cucumber:

*May contain more dump than alpha.


Conclusion: Current reward rates are not factored in to the PICKLE price, there is still plenty of room to grow without tweaking rewards distribution - if the price is depressed currently, it’s a reflection of market sentiment more than anything else. Instead, let’s focus on more pressing concerns. IMO the top two:

  1. Replacing the UNI jars to maintain TVL and thus keep the rewards flowing to stakers.
    • EITHER including stakers in governance (this would likely imply a lockup period à la veCRV)
    • OR replace staking mechanism with value accrual through a smart treasury. IMO this is the big question that will determine the future of PICKLE.
4 Likes

Think you nailed it here, @brinosaurus_rex.

1 Like

In financial markets liquidity is one of the key values. If you destroy liquidity you destroy the value of the asset. If you punish people for wanting to buy in size, then they wont buy. And if you punish them for wanting to sell in size, they wont buy in. If you want to dissuade whales then your asset will be small and so will the value of your asset. Harm liquidity and viscous circle ensues.

Beware hurting liquidity. Note the whole worlds financial assets are being propped up by governments injecting liquidity. If you kill Pickle liquidity you will kill Pickle.

DeFi value is foremost about trust and trust is won through stability which is in turn underpinned by liquidity.

Trying to get rich to quick with Pickle is a recipe for getting poorer fast.

2 Likes

Isn’t this temporary, for four weeks, as we pay out the excess treasury funds above $500k? Does the rest of the analysis assume this level of staking rewards over the period of a full year?

2 Likes

This may sound incredibly stupid because I have a limited knowledge of how this would actually affect overall liquidity/volatility, but has there been any consideration of alternate Pickle/TokenX pools? Like, for example, halving the rewards from the Pickle/ETH pool and adding, say, a Pickle/WBTC pool with an equal amount of emission rewards? I don’t know if that would have the intended effect or not, I’m just spitballing here.

Has a Pickle burn been proposed before? Nothing major, just a small burn per tx to reduce the overall supply of Pickle and therefore (presumably) increase the price of Pickle?

Again, I’m not working with super fintech knowledge here so I’m just curious to know if these options would/could “work” in whatever capacity.

2 Likes

Yes I just multiplied the weekly rewards by 52 (weeks in a year). You’re right to point out that this isn’t stable. Please don’t mistake my back-of-the-envelope calc for an in-depth profitability report! And please treat anyone giving you Exlusive Alpha Leakz © with a healthy dose of skepticism.

Something a bit more detailed / realistic / pessimistic:
I think it’s 4.5% of rewards going to stakers. (Right?)
So at 100M TVL in pJars and assuming 15% avg pool performance, jars make 15M annually.
4.5% of 15M is 675K per year.

This suggests a longer-term earnings multiple of ~30x, which sounds sensible enough depending on what you consider sensible in defi-land… Either way I stand by my main point which is that we need to focus on the fundamentals rather than tweaking the reward parameters.

I’ll amend my earlier post.

EDIT: See here.

Yes, this has been proposed and discussed here. Also check out this thread. TLDR there seems to be consensus that buybacks are useful but that we should buy back dynamically using a smart treasury instead of burning the bought-back tokens.

Re. multiple Pickle/TokenX pools, I think this would fragment liquidity too much - it’s better to have one deep pool rather than a bunch of shallow pools (puddles?).

2 Likes