A twist on intrinsic productivity to maximize INDEX distribution and coop treasury

Another week, another idea from OA…

OK, I think that the polls I started 2 weeks ago have helped us all get a better idea of that people in think community think in terms of lockups, staking and rewards.

Following some more discussions I would like to suggest some parameters based on the out come of, and some changes compared to, the vote. I’ll try to explain my reasoning in this post:

The overall aim I’m trying to reach is:

  1. Practical to implement
  2. Attractive to DPI and INDEX owners but not too generous
  3. All staking rewards based on INDEX distribution - increased INDEX distribution
  4. Build the coop treasury with non INDEX assets.

The third and forth points are important, because I would like to use the DPI treasury (Streaming fees and intrinsic yield captured) to purchase ETH with the intention of building the INDEX liquidity on Uniswap (or Balancer ?) to help temper the volatility of INDEX (without the coop pumping or dumping the INDEX price).

One other comment is that the expected yield from DPI token farming is currently only ~2% (If we exclude SNX staking as being too complex). This is inline with the numbers calculated by Set Labs. So while there is come intrinsic productivity available, this experiment is likely to make a loss due to INDEX distribution (but less than the 15,000 LP mining currently underway). However, it does provide a useful proof of concept which we can build on (addition of Maker vault for DPI, Maker vaults for underlying assets etc).


So, my proposal:

Time scale: 28 days. 01st Feb 2021 to 28th Feb 2021.
This matches the fixed 28 duration preferred in the polls, Running on those dates means that the farming can be deployed after the February reblance, and before the March rebalance, and so makes management much easier.
I understand that aiming for a start of January would result in significant time pressure on the Dev (Staking DPI / INDEX / LP contracts are easy, farming the DPI tokens to AAVE, Compound, and staking is much more difficult).

I am assuming that DPI:ETH LP Mining rewards would continue until the start of this experiment (and may be required while it runs if we get insufficient liquidity on uniswap).

Eligible farms

  • AAVE
  • Compound
  • Governance staking without slashing (YFI and KNC)
  • Governance with slashing (AAVE)

DPI stakers: All rewards in INDEX.
I would propose a fixed issuance based on 10% pa based on time weighted average price of DPI and INDEX in the week last week of January. This is effectively 1% to cover streaming fees and 9% to cover the risk of INDEX price dropping and to to reward for the lock up.

10% PA is 0.8% per month. So, $1,000 worth of DPI at start, would receive $10 of INDEX.

If we get $5,000,000 DPI locked, we would distribute $40,000 worth of INDEX (8,000 INDEX at $5 or 286 per day)

INDEX rewards would be distributed per block to allow farmers to claim and sell over time.

DPI:ETH LP stakers: All rewards in INDEX.
I would propose that LP holders get 50% more rewards than DPI stakers as they are exposed to divergence loss risks, and that this is based on the value of the LP token. i.e 15% pa based on LP value.

We currently have $55 M in the UNISWAP pair earning 88% income from the 15,000 INDEX per day rewards.

$30,000,000 locked and earning 15% would receive (AUV x 1.17%) $351,000 (70,200 INDEX @ $5 = 2,500 per day).

INDEX rewards would be distributed per block to allow farmers to claim and sell over time.

INDEX stakers All rewards in INDEX.
I would propose the INDEX stakers get 10% pa. i.e 0.8% of the number staked for 28 days. INDEX stakers have no price risk as they are reward with INDEX, but do risk slashing.

If 80% of the circulating INDEX (1,000,000) are staked this would distribute a total of 6,400 INDEX.

INDEX rewards would be given at the end of the the staking period.

INDEX slashing

  1. In the event of a loss of farmed tokens, then INDEX staked would be slashed by a maximum of 30% with the slashed tokens being returned to the coop treasury. Any additional shortfall would come from coop reserves (Streaming fee and intrinsic productivity)
  2. In the event of a DPI redemption failure (due too poor management of the DPI farm), staked INDEX would be slashed by 50% to the coop treasury.

Will not be included in staking, rewards or slashing.

Based on the assumptions above
DPI streaming fees = Unchanged
$20,000,000 DPI locked earning 2% pa (0.16% in 28 days) = $32,000
$15,000 ETH in LP = no income.

800,000 INDEX staked = 6,400 INDEX
DPI and LP = ~$391,000 = 78,200 INDEX @ $5.
Total = 84,600 = 3,021 per day (@ $5)
There would also be gas costs for the trades and a significant time to develop and test the smart contracts required.

Final thoughts

  • This proposal is intended to stimulate some more debate and is in no way fixed.
  • I’ve tried to reduce the uncertainties for the stakers (we could guaranteed a fixed $% reward paid in INDEX, but I feel that would be significantly more complex).
  • In the future I would expect higher yields from the farm as more opportunities become available (DPI Maker and or yDPI vault to allow double dipping by farming DPI).
  • In the future, I would expect more of the rewards to be paid from DPI intrinsic income.

One thing I wanted to add to the intrinsic productivity discussion is that there is INSANE returns if we are able to LP on Sushiswap. For example, Aave, COMP, and SNX are each yielding >60%.

On first blush, the challenges to achieving this yield are two-fold:

  • Inability to acquire ETH: The Index does not own ETH and thus is not able to provide the other side
  • Impermanent Loss Risk: Even if we were able to acquire the requisite ETH, we risk IL of ETH - which is difficult to manage.

It turns out that many of the solutions to these issues are coming live. In particular:

  • Aave Credit Delegation: For trusted counterparties, people can lend tokens on an uncollateralized basis to someone else. In this case, it is now possible for 3rd parties to lend ETH to the DPI.
  • Put/Call Straddle on ETH: At the end of the day, impermanent loss risk is the risk of price volatility. In the traditional markets, volatility can be bet or hedged against using options products. In the particular case of LPing, the right product is the Put/Call straddle, in which the value of the options structured product increases in value as DPI/ETH moves away from its desirable target.

This is all to say that there is another step in the roadmap for compelling intrinsic productivity opportunities we should continue to explore

1 Like

Although I LP and farm LP’s I’ve not really considered them as intrinsic productivity methods. I think I’m echoing yEarn who avoid LP based strategies (until they get Deriswap working…).

Divergence loss is a risk (but not huge - New Blog cumming soon TM).
I think the main problem is that the opportunities are short lived, UNI rewards was a couple of months, Sushi does a weekly menu. So we could spend lots of Dev time, tying to code a LP based farm and then have it removed.

I think that there may be longer term benefits from targetign Maker vaults.

  1. White list on Maker oracle for tokens
  2. Fork the old yEarn ETH strategy
  3. Mint DAI and farm on the best opportunity (yDAI / cDAI etc)

Maker have Vaults for YFI, Comp, LRC and KNC.

In any case, at the speed of DeFi, there may well be better solutions for all tokens in the near future.

Agreed - chasing high rewards on something like sushi makes sense for whales that can move quickly here - by the time we shipped the ability to support something like this, i think returns will likely have fallen back down to earth significantly

1 Like

I think it would be interesting to look into streaming the intrinsic productivity rewards back to index and dpi stakers as stablecoins. This would make us one of just a handful of protocols that distribute dividends using stablecoins, rather than just by printing more of our own token. The one problem I see with doing this now is that for a trial period of just 28 days, and with yields being relatively low, gas costs would most likely outweigh claiming these rewards for most. This may make it not worth it for now, but once a more permanent program is put into place, I think it could help boost the value proposition of index and dpi.

Stable coins makes sense in the long term (or income as DPI so they get a direct offset of the streaming fee without any divergence risk).

My goal in this experiment is to use some of the INDEX reserves and to build DPI as a reserve for the coop so we have a more diverse and less volatile treasury.

IN addition we benefit by getting INDEX into the hands of more people, if we maket sell INDEX for stable coins and then issue stable coins as a dividend, the DPI stakers may use the USD to do other tthings, if we give them INDEX, they may decide to hold them as “free money”.

28 days is a short period, but it was the community preferred option (and yes I ignored many of the other poll results :wink: ). "8 days has the advantage that everything can be run between two rebalances, so we remove a complication for the Devs.

I think that many were concerned that a 3 month lock in would be too long for many DPI holders as lots can happen in 3 months.

But the gas impact of staking and unstaking / claiming is not lost on me.

I wonder if v2 AAVE gives us a better option to farm DPI:

  1. Propose AAVE vaults for everything
  2. Stake DPI
  3. Remove all 11 tokens and deposit into relavent AAVE vaults.
    4.Earn interest in AAVE
  4. Combine the collateral and borrow stable coins for further income generqation.

Basically we use the same strategy for all 10 tokens. That should make the strategy much easier to code and manage. We may even be able to use the AAVE collateral swaps for rebalancing

Obviously if we borrow we face a liquidation risk, but we can use a very high ratio to play it safe.

@bweick @setoshi, any thoughts?

1 Like

Do you have a sense of the APY that can be yielded from earning interest on Aave?

I have no idea.

The interest for adding tokens as collateral is likely low (maybe higher for SNX), so it really depends on whether we can borrow stable coins and farm them for profit.

I’m just wondering if it becomes and easy / generic strategy for intrinsic farming.

Maybe compound is better due to COMP rewards.

we need to have DPI as usable collateral on Compound or Aave, this simplifies the process as no need to remove all tokens.

collateral then used to borrow stablecoins. then use Curve.fi stablecoin pools(circa 15%) return with CRV rewards or Yearn Yield (12% on compound strategy or 14% on curve’s pool3)

What about the added risks and loss of agility in a stressful moment when compounding rewards across different protocols? I can’t even formulate a proper question but I hope someone can meander around the point. It would be nice to get an easy way to assess those risks.

I think an important thing to keep in mind (if I’m understanding correctly) is that we would be increasing risk locking up the underlying tokens individually as collateral on Aave (or Comp etc…). If any one of those tokens has a black swan event, it wouldn’t just tank the DPI, it would destroy a portion of our funds.

Only is some circumstances would we face a risk:

If Token XX collapses, then aXXX will also collapse, but DPI would only drop by the weight of XXX and the coop should cope with any difference in the drop between the two (as both will be worth less)

If a protocol has a flaw / exploit and the xUNI token become worthless (but UNI still has value), then INDEX stakers / the coop would be liable.

If all of token XXX is borrowed, then we may be unable to remove the tokens when we need to. That could cause a liquidity crunch on DPI redemption - not good. However, the lending protocols are designed to keep some liquidity at all times (it gets very expensive to borrow at high utilization).

The key risk that we do avoid is any rug pulls / draining of DPI. The smart contract structures mean that 1 token collapsing does not drain the pool


An updated income table. Some changes, but still works out at about 2% pa.

I still think that we should progress with this experiment, mainly to prove the concept and to develop the codes required for intrinsic productivity as I see this as having potential to be a significant income source for all coop products.

I know that a few people have commented that we should be rewarding INDEX stakers with Streaming fees. I can see the merit in this and would like to discuss other structures for the allocation of income.

1 Like

I know it’s easy for me to say as a community member that won’t be building the contracts for this, but I think it’s worth doing. As more products come on stream in future we can use the same contracts to break apart and farm the constituents so it becomes more valuable over time.

We still have a small AUV relative to our goal so it might also be a good time to try experimenting. As per the poll the other day it seems DPI holders are in favour of building intrinsic productivity.

It’s also more generally something super cool that doesn’t exist in tradfi. Just my two sats

@overanalyser @DarkForestCapital may be at this point we take it to a vote and engage in an IIP. As per the current Product Onboarding Process, intrinsic productivity will follow the same steps as a new product due to the amount of dev and risk work required here.

We can get the ball started by having a proper IIP we can then take to voting.

1 Like

Understood. @overanalyser happy to work with you on IIP but think you’ve done most of the hard work already :wink:

1 Like

I wasn’t sure if this should be posted here or in the DPI streaming discussion. I think they kind of overlap.

I do think that there is an opportunity to add utility to INDEX through staking. However, it’s challenging to see how staked INDEX can be best used for the benefit of the protocol. One idea I had that hasn’t been explored previously has to do with insurance. This might make more sense down the line as we add more products. At that stage, staked INDEX could be used to provide a pool of asset to underwrite insurance for Index Coop and its associated products. We’d have to get it onto Nexus or a different insurer though. For the time being, we could use staked INDEX to underwrite insurance for the Set Protocol on Nexus.

I realise that this is somewhat half-baked. I think there’s an opportunity for intrinsic productivity with insurance, so wanted to share this idea with the community.


Half baked or not, I like the sound of it.

1 Like

I may have missed the mark but my image of DPI is having a simple token to invest in that obfuscates the compounding complexities of defi such as yield and governance tokens but still grants access to them. The fact that it’s not a promisary note but that DPI can be split into its constituent coins by the owner and the ability to vote in governance on the protocols is massively appealing. Why not siphon off a small % of streaming income to a smart contract risk mitigation fund? Another minuscule % could perhaps go to index holders.