Jare's Proposal Four for Yield Generating Opportunity: add Uniswap to on-chain loans

[staccDOTsol]

Why

Uniswap, as I undertand it, is a liquidity pool I can put funds into so that market makers and other parties can execute trades at leverage or for 'free-ish.' These funds then produce yield as the market makers have paid fees which gets added to the pot, then available for the liquidity providers to later withdraw. I may be incorrect in my understanding or I may have forgotten something.

Solution

https://github.com/Uniswap

Considerations

I'm certainly hoping their contracts are open and allow these types of on-chain interactions from other smart contracts.

https://etherscan.io/address/0x09cabec1ead1c0ba254b09efb3ee13841712be14#writeContract

Uniswap DAI (and everybody else born from this factory - 1000s exchanges) do indeed have 'addLiquidity' 'removeLiquidity' onchain :)

Additional Info

This appears to be something that would allow for yield :) Could RAY diversify and balance tokens thru exchanging them for Ether on the Uniswap exchanges should RAY put a new % into a new basket of top-performing Uniswap exchanges? This would also add fees to the pool, but cost some gas. I'm unsure how RAY balances things in on-chain loans, is it possible to have that kind of behavior (rebalancing without new gas) on Uniswap? @dpurhar27

[dpurhar27]

Hey Jare, you are correct, supplying liquidity to Uniswap pools is a means of generating yield through trading fees accrued from Uniswap (0.3% per trade).

We've considered adding Uniswap in the past, and still wish to in the future. However, there is an extra consideration for most pools named "impermanent loss" which you can read about here

In short, it requires a hedging strategy to cancel price volatility between the pairs you're supplying too. This can be incorporated into RAY, but current on-chain solutions are too expensive (gas and capital wise) which cancels the potential extra yield.

It's probable in the future RAY will let users opt-in to opportunities which aren't fully passive, such as this, and then we could offer it without a hedging solution.

There are some Uniswap pools we can natively now without any hedging strategy, such as the ETH/WETH pool, since there is inherently no price risk across that pair. However, another consideration is posed - how do we evaluate the future returns of the pool? Returns are driven by volume on the pair meaning we'd need to predict demand. This is bit more complex relative to lending returns, but can be done.

The next steps to take for the ETH/WETH pool is determining if it generates enough returns to beat out what we currently support.

[staccDOTsol]

@dpurhar27 what are the existing on-chain solutions to hedge against impermanent loss?

+1 ETH/WETH and other pairs we could incorporate.

As far as projected returns are considered, all that people would want is an average APR for all the last available month's performance, excluding outliers - with a stern gambling-esque warning that 'prior performance is no indication or assurance of future performance.'

I think gambling terms would help support more Uniswap pairs, too - given you guys are considering off-chain solutions with a lot more inherent risk, I think that medium-risk investors will be happy with 'this is onchain BUT impermanent read about it here.'

[staccDOTsol]

Closed by mistake!

[dpurhar27]

You can form custom hedging solutions by using DEX's to open short/long positions - ie. dY/dX or bZx. The idea is to remove price risk.

Unfortunately, that won't work for RAY's focus. The job of RAY is to optimize yield on behalf of users. RAY doesn't leave it up to the users to accept risk they may miss gains b/c it's funding speculative opportunities. RAY's current core product is to be passive, and not speculative in any way.

That being said, RAY will be offering more speculative opportunities, which a user can use to put together a higher risk/higher reward portfolio. The idea being, opportunities used still don't allow capital loss (only yield opportunities) but they offer a potential higher return by being speculative. Here, we'd still need to optimize and therefore attempt to predict future Uniswap pool gains, but including such a disclaimer makes sense.