As a GMX delegate i will be voting in favour of this proposal, reasoning and discussion as follows.
As stated in the proposal, the intention of this change to the rewards system is to encourage a strong buy pressure on GMX token. A more straight forward way to implement this is to include the option to swap accrued ETH/AVAX rewards into GMX when claiming/compounding. Metavault, a once friendly GMX fork but has now transitioned into much more, implemented this within their forked version. I believe the request from the community for such an implementation was more for convenience rather than creating this strong buy pressure, and after some thought while it would bring buy pressure it would be likely largely negligible and wouldn’t achieve the entirety of the proposal.
Having read the comments above, across telegram and twitter i see a couple of common concerns, which i will comment underneath each with my own thoughts:
1) Is buy back and distribute still considered real yield?
So long as the distributed tokens are fully liquid and not locked in any way, then yes i would consider this real yield still. I think it must be made abundantly clear that the distributed GMX are not newly minted tokens and that these are 100% purchased with real revenue collected from GMX platform.
2) Yield is now in a more volatile asset, potentially more risk associated?
I would argue that as a GMX holder/staker you already accepting of the risks of a more volatile/lower market cap asset and that this should not matter in that context, but the option to convert GMX → ETH/AVAX in claim/compound is a must.
3) Users will lose $ value if having to convert GMX → ETH/AVAX on claiming
Of course this part is hard to mitigate. Ideally swapping would happen through gmGMX market, but this may take time to become liquid enough that price impact does not significantly reduce the net earnings. Having the option to route through aggregators could help to mitigate the fees paid on swap, but will require development time and potentially introduces further smart contract risk. I believe longer term that the value brought to GMX token through BB&D would negate the fees lost in swap vs being paid in ETH/AVAX.
4) What happens when people just chose to swap GMX → ETH/AVAX on claiming? Surely that will offset the buy pressure?
Yes, to extent, but if faced with two options 1) ETH/AVAX → GMX in the claim/compound process and require the user to be more conscious in converting rewards to GMX or 2) Distribute as GMX and give users the option to swap into ETH/AVAX, i believe the latter will result in far more net buy pressure than the former. If there is no significant buy pressure after implementing BB&D, which indications from other protocols that implemented such designs recently (eg gains) suggest this would not be the case, then the only thing that would be lost is time.
To conclude, so long as implementation does not require significant development time and does not add unnecessary smart contract risk i see very little downside with this proposal.