MVI Liquidity Mining - June 2021

[mel.eth]

Title: MVI Liquidity Mining - June 2021

Author: OverAnalyser @overanalyser Mel @mel.eth

Created: 2021-06-07; updated June 8, 2021

Link to IIP-24: IIP-24: MVI Liquidity Mining -
Link to May 2021 Update: MVI Liquidity Mining#7 - #4 by overanalyser

Simple Summary
Extend liquidity mining rewards for the MVI-ETH Uniswap pool in line with IIP-24 mandate over the 3rd and final 30-day period of the 90-day total. Total INDEX allocated for the 30-day period will increase from 2,743 INDEX to 3,286 INDEX (approx. $101,440; 7-day MA)

[incorrect May '21 reward value used in original post (increase 4,447 INDEX to 5,328 INDEX), corrected (to actual 2,743 INDEX issued in May '21 with new June '21 reward of 3,286) on June 8, 2021; good catch @verto0912 ; figures and calcs updated (thanks @overanalyser ) in post below.]

Mandate for days 61-90 within IIP-24
Minimum 1,425 INDEX
Maximum 5,700 INDEX
Target $5MM Liquidity

Data Collected
7-day Average Liquidity (Uniswap V2: MVI-ETH Pair): $4,173,548
+2% Depth (As of June 4, 2021, at MVI price of $48.99): $42,405
-2% Depth (As of June 4, 2021, at DPI price of $48.99): $42,278

Calculations:
Percentage of target liquidity achieved (7-day trailing): $4.17MM/$5MM = 83.47%
Further calculations here: MVI Liquidity Mining - June '21 - Google Sheets

Specification:
Extend the current LM campaign for MVI:ETH Uniswap pair for 30-days using the existing contract (as outlined in IIP-24) for an additional 3,286 INDEX (IIP-24 and previous LM Update No. 7 linked above). No action would be necessary for existing liquidity providers.

Methodology:
7-day Average Liquidity / Target Liquidity = Percentage of Target Liquidity Achieved

Given that the Percentage of Target Liquidity Achieved is currently at 83.47%, an increase in the existing incentive by a factor of approximately 1.198 (1/.8347) results in the new reward for the final 30-day period increasing to 3,286 from the previous 30-day allocation of 2,743.

The reward calculated falls within the banding established in IIP-24 of 1,425 (min) and 5,700 (max).

Factors considered:

1. Sustained liquidity is a core driver for broad spectrum adoption.
2. Maintaining continuity of incentives will likely retain current LPs.

Limiting factors:

1. While there have been discussions regarding the broader implications of LM ( Liquidity Mining Strategy ), this proposal is limited to the mandate of IIP-24 and would be the third and final 30-day extension of incentives under that IIP.

Data Sources:
Coingecko
Uniswap

Next Steps:
This post is the formal recommendation of the rewards for days 61-90 as per IIP-24. Implementation is currently planned under the existing mandate; no action is required by the community.

[LemonadeAlpha]

Thanks for this - I think that we are missing a big piece in these reports and that is the actual trade size distribution

We speak about liquidity targets based on servicing a target 1% slippage trade sizing but I think it would be great to arm these with more data around:

Does our trade size target figure come close to being hit or exceeded? Often?

I’d argue we can’t make a good decision about these programs without this data

[mrvls_brkfst]

Always appreciate these reports, thank you @overanalyser and @mel.eth . I very much see this a bringing home the current strategy before the new proposals begin (MVI - Growth Budget Request).

The learning from this LM episode, the research around Uniswap v3, the L2 strategy, and the desire to test whale appeal for MVI mean there’s a lot to consider this month if June LM is to be the last in the classical style, which it feels like it should be. Looking forward to how we power through the summer, whatever it becomes known for!

[overanalyser]

I agree, and to copy some of my comment on DPI:

I see the liquidity mining strategy comprising 3 pillars:

1. L1 Pool size required for trades without undue price slippage
2. Liquidity to maintain market dominace in a sector.
3. Liquidity to promote adoption of L2’s etc.

Until now we have used Uni v2 to address #1 and #2.

With the advent of uni v3 and more side chains / L2’s there are many more options to work with.

Looking forward I see #1 as being more of a PWG responsibility and likely utilising Uni v3 , while #2 and #3 migrate towards Growth and BD and using different approaches.

However, there are going to be some interaction we need to consider (what happens when we have more liquidity on L2 than on L1 )

PWG have a work team looking at Uni v3, and how it impacts the coop’s product/liquidity mining.

I’m hoping to develop/discuss and agree on a new framework for liquidity requirements / provision before the end of the 90 day campaigns we have running on MVI and DPI (i.e. before the end of June) for implementation in July.

---

Liquidity Mining Strategy

OK, we had quite a long discussion back in February about liquidity mining, however, I’m not sure we have a strategy that is fit for purpose going forward: So, I’m going to try and restart the discussion: Liquidity Mining Strategy Until now the coop has managed liquidity mining by a combination of: The initial Coop launch IIP’s for specific product LM programmes GWG grants for loopring incentives As time has progressed the calculation of rewards has developed into something based on th…

---

In terms of current trade sizes, I’m currently doing DPI on uni v2 and v3, and will do MVI v2 this week.

[MrMadila]

Just adding to @LemonadeAlpha 's valid point. Would it be worth overlaying/comparing this data with any insight we have over our exchange issuance module usage?

Ie. for large trades > $X, are whales trying to purchase via exchange, or are they sophisticated/motivated enough to be minting and redeeming for themselves to avoid slippage.

Or are we witnessing repetitive trading from addresses building positions?

[verto0912]

I think with MVI there’s very little exchange issuance usage, other than by arb bots. The issue with AXS migration made exchange issuance hard to use for about a month. Plus, as liquidity of the underlying deteriorated during the sell-off, trading through the Uni pool was resulting in roughly the same slippage as exchange issues. Anecdotally, DG chose to just buy on Uniswap ($100k) because exchange issuance had the same price impact.

@MrMadila

Or are we witnessing repetitive trading from addresses building positions?

Not too many repeat buyers.

[verto0912]

@mel.eth

Total INDEX allocated for the 30-day period will increase from 4,447 INDEX

@overanalyser @mel.eth this number is wrong. 4,447 INDEX is the amount that was calculated for May using the incorrect initial distribution as pointed out by Dylan last month. We ended up with below 3,000 INDEX for MVI incentives for May.

Can you please check and revise.

[overanalyser]

Gah, you are absolutely correct.

Ratio is unchanged at 19.8% increase.

First 30 days = 3,810 INDEX
Second 30 Days (-28%) = 2,743 INDEX
Third 30 days (+19.8%) = 3,286 INDEX

Updated calculations here

@mel.eth can you edit the original post?

[Matthew_Graham]

I’ve been pondering something. Are we better off measuring/adjusting incentives based on APR using current INDEX exchange rate.

We refer to APR as the means of attracting capital into the pool. But here we refer to INDEX per day as the means. These are correlated but INDEX price is a huge variable that we are overlooking.

I think we should adjusting target APR to drive the change in the liquidity within the pool. Ie: we want more liquidity, we will offer high APR.

If we have an APR of 25% and the pool is not meeting our target size. We adjust the APR to entice LPs to enter the pool.

Once we know the next iterations APR, we can use the current INDEX exchange rate and determine how many INDEX is needed to attain the desired APR. This way is INDEX’s price rockets up or dumps down, then we incorporate this into how many INDEX we use to incentives users to provide liquidity.

The reason being I always look at yield, not how many INDEX I’ll receive when contemplating being an LP. The math laid out in this forum post fails to incorporate the change in INDEX price.

Also, now that rebalancing has been performed and no issues. Can we amend the pool size target?

$5M was the initial size, and this parameter can be changed - it’s not fixed for 90 days… it is an input for helping determine how many rewards we provide.

[mel.eth]

@verto0912 thanks for reviewing and flagging this; @overanalyser thanks for updating the calcs on the sheet! All figures and calcs now updated on the original post.

[verto0912]

I believe we actually did target 35% APY in the original IIP. Now we are just assuming that the current APY of 22% is the equilibrium which it is not Especially in the down market and given the risk profile of the MVI.

[puniaviision]

@Matthew_Graham

$5M was the initial size, and this parameter can be changed - it’s not fixed for 90 days… it is an input for helping determine how many rewards we provide.

But that’s how you’re doing it @overanalyser right? We target the APR necessary to hit a certain pool size. We assume an APR the market resolves at, target a liquidity pool size, and determine the number of $INDEX at the current price.

[Matthew_Graham]

@puniaviision

But that’s how you’re doing it @overanalyser right? We target the APR necessary to hit a certain pool size. We assume an APR the market resolves at, target a liquidity pool size, and determine the number of $INDEX at the current price.

I believe the forum post adjusts the INDEX/Day metric based on current liquidity.

I think it would be more effective to adjust APR (%) based on target liquidity. Then later work out how many INDEX we distribute per day to achieve this.

In this cycle, pool size is less than target >> adjust APR >> calculate how many INDEX is needed.

I genuinely feel we can amend the $5M target and go higher. I believe the LM framework is all about determining what rewards based on set goals. We can change the goals based on where the product is in its life cycle.

Is targeting $7.5 or $10M an option - As in can rebalancing handle that size product. It makes sense to boost liquidity whilst it’s easy to do on V2 - hopefully unincentivised supply then tracks like DPI.

[overanalyser]

As a LP, when considering entering a farm, I would agree that I look at the %APY when considering a farm. However, in some ways, and High APY indicates higher risk, a new farm, unfavorable tokens or I’m missing something. While the market isn’t efficient, the trends are there.

For me the process in setting rewards (for uni v2) is:

1. Estimate an APY that the market will settle to.
2. Identify a target pool size
3. Calculate the number of tokens (using a TWAP INDEX price)
4. After the farm has had time to settle, the APY calculated in step #1 is kind of irrelevant - The market has decided what rewards it will accept. So, we should just look at the ratio of pool size vs target and adjust

Notes:
a) for DPI and MVI, the LP market appears to be settled around 25%.

Shushi swap is a little less, so either the market prefers to be paid in $Sushi, or $INDEX price is outperforming a little.

08jun21	Fee APY	Rewards APY	Total
MVI	3.7%	22.0%	25.7%
DPI v2	15.0%	11.0%	26.0%
DPI Sushi	6.6%	13.7%	20.3%

b) Large pool size has a number of benefits including: less price impact for large buys, gives holder confidence in exit liquidity, allows Arbitrage closer to NAV, should increase AUM.

c) Rather than INDEX per $ liquidity, it may be more proper to consider INDEX per LP token, so we ignore the impact of price changes for the set token and ETH.

I could propose that we do liquidity mining based on a target of 7,000 MVI:ETH LP tokens. Which would mean that it would ignore the effect of token prices and should give a reasonably constant slippage when trading MVI:ETH. – This is arguably a better metric to use. However, I don’t know anyone who thinks in terms of LP tokens (We currently have ~$3.5 M liquidity, and 5,576 LP tokens, so 1 LP is ~$630 USD)

@jdcook - a suggestion: add LP token supply to the KPI dashboards. We can see where the Set token sitting in the LP, but issue / redemption of LP tokens isn’t easy to see.

d) Liquidity is sticky (and likely slow to add). Some LPs will react to the market APY, but a lot will be passive. Personally, I decided to LP MVI at launch and I wouldn’t consider adding or removing liquidity during the initial 90 LP campaign. Likewise, I don’t think my DPI LP position on Uniswap v2 has changed since November. Gas, attention, tax and overall portfolio balance considerations mean I’m relatively insensitive to APY.

e) We could remove some of the uncertainty by giving rewards in DAI / ETH / MVI. However, some people are in the pools to accumulate INDEX, if we gave them something else they may not buy. We have a treasury full of INDEX, if we exchanged it for other tokens to use as rewards, we would tank the INDEX price.

@Matthew_Graham

$5M was the initial size, and this parameter can be changed - it’s not fixed for 90 days… it is an input for helping determine how many rewards we provide.

No it’s not fixed at $5 m. However, I think that is more of a Growth discussion. At the moment I / PWG has been targeting L1 liquidity and AUM. I’m hoping that the conversation will move more toward slippage (customer need) and AUM growth.

I’m aware that there are plans to launch/incentivise L2 / side chain liquidity over the next month or so, so it will be interesting to see what happens to the liquidity/trade volumes.

[overanalyser]

@Matthew_Graham

In this cycle, pool size is less than target >> adjust APR >> calculate how many INDEX is needed.

I struggle to see how this calcualtion is different:

Current cal:
AUM is 4.17 m = 83.4% of target.
INDEX rewards are 2743
x1.198 INDEX = 3286 INDEX for next 30 days.

APY based calculation:
INDEX price is $34.59 (20 day TWAP used for May rewards)
INDEX rewards are 2743
APY = 2743 / 30 * 34.56 * 365 / 4,170,000 = 27.7%
AUM is 4.17 m = 83.4% of target.
Increase APY by x1.198 = 33.2%
Index rewards = 0.332 * 4,170,000 / 365 *30 /34.59 = 3292 INDEX tokens over 30 days.

[verto0912]

@overanalyser

I struggle to see how this calcualtion is different:

I think the calculation is different because the target was initially set at 35% APY for a $5m pool.

Index rewards = 5,000,000*0.35/12/34.59 = 4216 INDEX (you can do the same with 365 and 30, instead of 12).

[Matthew_Graham]

@overanalyser

Index rewards = 0.332 * 4,170,000 / 365 *30 /34.59 = 3292 INDEX tokens over 30 days.

I would use the “target pool size” in the above calc.
We know 27.7% retains $4.17M. Therefore, 27.7% at $5M plus the extra amount.

I don’t think the relationship between incentive APR & pool size is linear, but without data to say what the relationship is - It is very hard to say what the “extra amount” should be. Perhaps a nominal amount, 5% or some arbitrary selected number will do, until we have a data set to drive decision making.

@verto0912

Index rewards = 5,000,000*0.35/12/34.59 = 4216 INDEX (you can do the same with 365 and 30, instead of 12).

The market isn’t as buoyant now as it was soon after MVI launched. Based on this observation, the above calculation may be a little lite on. That is my gut feel though. I do agree, the target pool side should be used in the calculation though.

[verto0912]

@Matthew_Graham

Based on this observation, the above calculation may be a little lite on.

Would agree here. We targeted 35% during a very different market environment. We saw that during the more up trending market, LPs were okay with 25%-28%. This was still driving meaningful unit supply growth and we were overshooting on the liquidity target by good 40%. We see that now, incentives are not meaningfully affecting unit supply growth and liquidity is under target. Again, DFC and myself think about this from unit supply first, liquidity second perspective. Unit supply drives our revenue and incentives should be used to affect unit supply first and liquidity second.

I would love to experiment with this and see what LPs want to add to their LP positions in the current market environment. My guess is that it’s 50% APY. However, I do understand that this might not be acceptable as we are trying to scale back spending on LM incentives.

[Matthew_Graham]

@verto0912

I do understand that this might not be acceptable as we are trying to scale back spending on LM incentives

From the income statement perspective: Liquidity Mining is not cheap. But without on-chain liquidity product integrations are not viable. If we think long term, we need a lot more on-chain liquidity for integrations with Tier 1 lending protocols like Aave.

The spend on MVI is a lot less relative to DPI and the DPI product is a lot more mature in its lifecycle. Capital efficiency would be to shift LM rewards from DPI to MVI and still taper our expenditure.