This idea would work for many of the tokens supported by aave. And would have all of the opportunities there accessed
The main gist of this idea is for deposits of a single coin (say ETH) to be spread across all the aave lending opportunities. The spread would be based on the yield opportunities therein. However, if the vault got too big it would saturate all the lending opportunities found there.
I saw that KNC was currently yielding much more than the other pools on account of low lending to borrowing ratio. I thought perhaps, even though I am not really interested in holding KNC it was a good opportunity. The difference here from just supplying to certain pools is that each vault user would only be exposed to a portion of any single asset and only for a certain time frame.
I don’t know the coding that would be required to pull this off but this is the flow of things as I see them.
1 Depositors will deposit the ETH, (or whatever single asset that is deemed best for this use)
2 The code looks at the current aave yield patterns and calculates a ratio of all, or the best, yielding stats of the aave pools. Pools that have higher yield, and thus have more space in them for added liquidity would gain a higher portion of the funds in the vault.
3 The code converts all the ETH in the vault to those tokens in the calculated ratio.
4 The code deposits them and monitors the yield or saturation levels. (the time frame is a part that needs brainstorming)
…leaving the tokens longer exposes vaulter to perhaps undesired tokens.
…leaving the tokens longer also exposes vaulters to capital losses of tokens they are not following.
…leaving the tokens longer creates a better gas to reward ratio, assuming all other things stay constant.
5 rewards are converted back to ETH and perhaps all the pools can be emptied out and converted back to ETH also for redistribution (but this appears wasteful but might offer some security against capital losses if it is )
Now this proposal certainly is not fleshed out but I think it has some potential.