Compound strategies (USDC and/or ETH)

I allow myself to create this topic in order to discuss the opportunity of exploiting the strategy offered by pJAR 0.88 (pDAI farm).

According to its results and the various opinions concerning it, it is a success :

  • a steadily increasing TVL
  • a very competitive yield
  • use by other protocols, most of which wish to preserve Pickle’s stability (Yearn, yAXIS, Stabilize)
  • the support of the community as favorite for the attribution of rewards

In this context, and if it is feasible, I would like to see proposed a new jar with a single asset, like USDC for example.
I made a spreadsheet based on the values ​​of at a time T:

  • USDC being used as collateral
  • DAI as a loan.
  • These two assets offer interesting $ COMP rewards.

I simulated the values ​​obtained with a collateralization rate of 80%, and a value of USDC equal to 1.01 DAI. The spreadsheet has only 15 rows and the starting bet is 1,000,000.
This is what it looks like (EDIT : picture above show you USDC one asset strategy):

If I was correct in my calculations, the rate of return would be 17.96% APY ! with a leverage of 3.79x.
Decreasing the collateralization rate lowers this leverage (2.92x lev@75% coll = 14.93% APY).

Regarding the redistribution of interest, I’m not sure how that would apply for Pickle … autocompound USDC + reselling $ COMP for USDC or DAI ? or against PICKLE ? :thinking:

I would like your opinion on this idea as well as on the method I used.
Am I wrong in my calculations?
Have I forgotten some data or information?

Is this the method used by the pJAR 0.88?

To be honest, I am waiting for the availability of the borrowing option in to test this strategy individually.
(This application allows you to group various transactions into one, and saves on gas costs!)

Assuming that this is achievable, what would you think of a “pJar” using this strategy for ETH?
A collateralization ratio of 50% would provide a return of almost 9% APY, based on previous data.

I leave you access to my spreadsheet if you wish to make any comments whatsoever. I can give you access in copy and on request (by email).

Thanks for your feedback :roll_eyes:

(November 14)
I have updated my spreadsheets.
It seems obvious that a strategy with a single asset is more relevant!
The first sheet presents my initial idea (I could not borrow the same asset that I collateralised in Compound, before today)
The 2nd sheet considers only DAI, the 3rd only USDC and the last one presents a strategy with ETH as collateral up to 50%, with USDC as borrowed asset.

If anyone knows how I could automate the values ​​of a cell (% APY, ETH value …) in a Google spreadsheet, I’m a taker :wink:


Beautiful work! The other key benefit you didn’t mention is that the yield is not TVL dependent, and very profitable! So for me the pJAR 0.88 should be a primary focus for future strategy development.

Last time I looked at the USDC stats they were very similar to DAI, albeit lower liquidity, so I think the team was already looking into that as an option.

Everyone would love an ETH option but not sure it’s as feasible/more risky.


Well thought out proposal!
The numbers seem to make sense…I support exploring this further.


Thanks guys !

I digged a little more into the calculations:
Imagine the returns only come from COMP harvests, for this the accumulated DAI debt must be less than the accumulated USDC rewards.
By observing the variation in rates over a period of time (a few hours at the moment), I found that the maximum borrowing limit was between 67% and 70%.

This gives an annual return of approximately 12% APY. Which corresponds roughly to the cumulative returns of COMP.

By setting a limit at 67%, the excess USDC rewards produced is equal to 0.25% of the invested capital + a total return of 11.61% APY.
This surplus could be reinjected to increase the value of provisions (and therefore encourage “staker” longer).

  • With a limit of 60%: surplus = 0.9%, APY = 9.81%
  • With a limit of 65%: surplus = 0.46%, APY = 11%

From 70%, the debt outweighs the rewards (excluding COMP), which seems unsustainable to me over time.
Knowing that I am only using 15 lines of calculation, it must be possible to maximize returns up to a point.

I think that only the yield provided by an allocation of emissions (from a UNI pool, or by a “levy” on PICKLE/ETH LP to please certain people ^^), would be influenced by TVL.

I cannot imagine the influence that such a strategy would have on the Protocol Compound (or another) …

Here is the start of a Discord conversation started on the subject:

@Sifu uses this kind of strategy on a personal basis, and @Bulleye was talking about this a while ago (but never showing interest with data).
It seems possible to use only one asset as collateral and loan, DAI offering the best rates. This is what Dapps like Instadapp or DefiSaver will offer …

  • Using only USDC as a provision and a loan, I get better returns: around 13% APY with a limit of 71.8%.
  • Using only DAI as a provision and a loan, I get better returns: around 15.6% APY with a limit of 73.5%.

Checking out how to apply all of this is beyond my skill at the moment.

Regarding ETH, such a strategy would appear viable and relatively secure by limiting the borrowing coefficient to 0.5. Which today would give a liquidation risk if ETH drops by half. At the slightest warning, it would be possible to repay the debts. Well, I believe…

Anyway, I’m going to stop with the numbers lol, if that’s all that is infeasible and I would have done it for nothing :sweat_smile:


You’d have my vote. Seems the simplest and most elegant way to offer Jars for these USDx coins.


@Tomato_Juxx maybe you could approach the devs with this proposal


He has by posting here, the devs will read it when they get chance

1 Like

This subject does not pretend to be a proposal, but just a line of thought on possible opportunities.
I will be delighted if this results in the development of a clear and well-defined statement.

The advantage of a forum post is that people have time to read and study every word, unlike in Discord where most ideas get lost in the flow of conversations.

As specified, this is only an approach as I may have omitted certain details or data that would invalidate my results. It is above all this point that I wish to clarify and discuss because it is also possible to optimize and improve such a strategy.

I noticed that as of today I can use one only asset on, as collateral and as loan (which I couldn’t do before from UI).

Today to test such a strategy, it would cost me too much in gas costs, whether to apply it manually on Compound, or by using Dapps like DefiSaver which require fewer transactions but much more expensive.

Furucombo should soon (according to a dev on their Discord) allow all the necessary transactions to be grouped into a single (proxy-contract): Supply + Borrow + Sypply + Borrow + … for a much lower cost.

The fact that Devs interact less often on Discord and Forum leads me to believe that they are already busy or preoccupied.

From memory, it seems to me that Larry had justified that USDC was not profitable enough, and that ETH was too risky …

1 Like
  • Harvest are doing it (at least I think).
  • Our devs are the right persons that could evoluate the risks.

Harvest Strategy

How can they offer lower interests than those offered by Compound?
Even with 30% fees on dividends it should be (2.54% x 0.7 =) 1.78% (not a big difference though)

And they don’t seem to be adopting the strategy I’m studying. Just look like they’re stacking USDC on Compound as a simple vault (without leverage).
Most of the returns they offer come from what they add in FARM’s rewards (0.79% of week 11 emissions).

If Pickle implemented a Compound-USDC strategy as I present it, and with current rates, we could claim a return of 11.17% APY, net of fees (27.5%), with a 10% collateralization margin as applied to pDAI, and with no emission’s allocation.
At the moment, the rates are more interesting for USDC than DAI in such a strategy!
Maybe I’ll end up creating a real proposal about all of this, because I don’t see any downside or any other protocol offering it … :roll_eyes: :cucumber: :star_struck:


I agree, I think we should also offer a pJar 0.88 using a Compound-USDC strategy.

1 Like