Zeus
First I agree with the ball park number of 10%.
As far as meme goes I would suggest:
10.288 = cube_root(33)* cube_root(33)

    tex I agree with the idea and in my view the allocation shouldn’t be large at this stage. Anything in excess of 15 percent will be too high. My general preference would be to have the treasury allocation at 10% as a starter.

      Great job for spearheading this, tex. Agree with starting small and re-evaluating. 10%.

      In favor of this in a general sense. As earlier commenters have expressed, I'd prefer to begin with a conservative allocation (relative; perhaps 10%) and increase after an initial evaluation period if the DAO decides it's in the treasury's interest.
      It would be helpful to discuss other near-term plans or general strategy for excess reserves to better understand how AAVE would fit in to the larger plan. Whatever minimizes risk for OlympusDAO and assists in raising OHM's RFV should be considered!

      This is indeed a great initiative. As of the treasury allocation, I agree with those who have expressed conservative numbers (someone said 2sqrt(33) which is ~ 11.5% and a reminder of our favorite meme).
      We want to diversify to mitigate the protocol risks, so we have to keep in mind that we will come up more partnerships in the future (remember the rari fuse pool that is being discussed ).
      Also I think that it would be smart to always save some gunpowder in the treasury, which would give us flexibility and reaction capabilities in case of a blackswan. So if we want to keep building partnerships in a sustainable way, I wouldn't exceed 15%.

      Thanks for the good initiative!

      xh3b4sd I concur. In the future I would also make different % allocations to other protocols depending on the risk they pose (eg RGT). 10% seems like a good place to be but I would refrain from scaling higher in the short or medium term.

        around 10% seems reasonable here ...

        <=10% sounds reasonable to me.

        What Metrics are we considering, analyzing and what is are thw key objectives for doing this? +RFV?
        A data driven approach + structured high-level considerations can help us here.

        Good thinking. I'm onboard with a conservative amount. 20%-30%ish...? Hat's off to the big brains at Olympus.

        Question: should we be thinking about this as a percentage (i.e. 10%), or as a nominal value (i.e. $1m)? If it is as a percentage, we need to constantly re-balance to maintain the correct allocation; if it is a nominal value, it remains there until decided otherwise. Upside of nominal is less management, upside of percentage is it remains consistent portion of the treasury.

          Zeus I think percentage makes most sense, since the percentage value is an answer to how much of the treasury we're willing to risk and allocate to Aave. When we grow the treasury x10, the strategy should follow IMO.

            Nominal with quarterly rebalances via governance could be interesting. It would appear to offer more flexibility if new opportunities arise and allow for a more streamlined process to move funds around should a crisis event / black swan occur.

            Zeus So it looks like the consensus of the community is building somewhere around 10-15%. If we think of this as a percentage of excess reserves, there will indeed need to be some active management strategy in place to manage the reserves. Perhaps we could start with a nominal value to begin with... something round like $1M which represents roughly 13.33% of the ~$7.5M in excess reserves right now. In order to target a steady percentage of excess reserves, we could develop a strategy of offering aDAI bonds when under that percentage and aDAI->DAI redemptions when over percentage. aDAI bonds was my initial thought for this proposal anyways and a small targeted amount to maintain percentage of excess reserves invested in Aave will allow us to gauge interest from aDAI holders and grow the treasury at the same time.

              I agree with the 10% allocation
              these meme numbers are good - square root of 33 rocks

              I also like the sqrt 33 approach, but mainly because I think 10% is a good starting point. It's enough to matter, but also seems like an acceptable level of risk taken.