Governance: Liquidity programs should factor time

[julian-molina]

The liquidity programs need to be improved to eliminate an obvious attack vector:

People may add liquidity the last day of the governance process, get rewards, remove liquidity immediately after, and potentially dump the rewards in the market. Because liquidity rewards don't require voting or any manual process whatsoever other than setting up and submitting a profile, the danger is quite real. There is no community oversight that would prevent such things from happening.

Even if no bad actor is involved, it is not fair to reward all liquidity equally.

The liquidity programs must check the date the LP tokens arrived in the wallet, and factor in the time spent in the calculation of rewards.

I suggest time should not be a linear factor. Longer periods should be rewarded more, so as to incentivize long-term stability of liquidity. There should probably be a cap at 90 days where the factor stops increasing, so as not to privilege excessively long-term liquidity providers, discouraging new comers.

[BastianMuc]

I'm currently experimenting with this.

@julian-molina - I have two follow-up questions on how we imagine the target picture:

1. If, after the implementation of this, a user's liquidity token reward gets reduced because the user has not held the liquidity for a sustained period of time: What do you expect to happen with this user's "unclaimed" tokens? Should they
   a) remain undistributed, or
   b) only affect the user's share of the total pool, so that the pool will still get distributed in full, but to the benefit of other liquidity providers?

2. I tend to disagree with your suggestion not to make time a linear factor - mainly because I prefer things to be simple. My suggestion: If I held liquidity for 2 days before distribution, I'll receive 2/90 of the reward. If I had held the liquidity for 90 days or beyond, I'd have received 90/90 (=all) of the reward.
   If you insist on having it non-linear - do you have any idea on how steep you'd like the curve to be?

[julian-molina]

Sorry for the late reply, I somehow missed the notification!

1. b) -- The pool should be distributed regardless, mostly for simplicity. Any other formula would add complexity that few people will understand.

2. You're probably right. Simplicity is desirable.

[julian-molina]

@BastianMuc have you got to look into this?