For reasons of efficiency and sustainability, the current rewards system should be replaced with a new one, and that all further rewards payments should go to a new smart contract.
The level of rewards is currently too high and the $1 million Polygon fund is running out too fast. We went through approximately 400k in 8 weeks and one of those weeks had zero payouts and another week had only partial payouts. The solution is to end the bonus rewards (that is given to early LPers who stay in the market until the last hour) and to reduce the general rewards rate by 70%. This would give time for the Augur DAO to arrange additional rewards and to increase volume. This might add another 24-30 weeks to the program (I’m loosely estimating - someone could do the math on this).
The No Contest and No Contest/Tie options need to be eliminated to reduce the cost of impermanent loss to LPers. This loss can be from 1%-2% (and possibly even more). It is especially bad for NFL pools where a tie is not possible as the price can drop from 0.02 to as low as 0.001. For instance, if you have a NFL pool where the price falls from 0.02 to 0.003 there is an impermanent loss of approximately 2.09%
This effect can be countered by initializing the price for No Contest at 0.002 to 0.004 (and for No Contest / Tie at perhaps 0.007 to 0.009) - however there will still be some IL over the course of the 1-5 days leading up to the NFL game as cancelling becomes less likely. And the price of No Contest could fall as low as 0.0005.
This week alone, 3 out of 13 NFL games were delayed due to covid. Fortunately people weren’t providing major amounts of LP, otherwise the LP losses would have been huge (perhaps 20% of LP capital for NFL games???).
This is also a problem in MMA and other sports markets.
This LP loss is only sustainable with a large rewards system. For LPing to work in the long term and to make the rewards program significantly more efficient, it makes sense to eliminate No Contest and No Contest/Tie and to replace them with a method that pays out 50/50 in the event of No Contest or a Tie.
Excessive Incentives for Tiny LP Amounts
Currently if someone is the only LPer they can get 100% of the rewards for a market while potentially only providing $1 or less of liquidity. This is a waste. I recommend scaling the LP rewards to the total amount of liquidty. For starters, I would suggest linear scaling the LP rewards based on a target of $5k. So at 5k of LP the LPers would get 100% of the rewards, whereas with $500 LP they would only get 10% of the rewards and so on.
If rewards are paid to the existing contract, a super majority of the rewards will be distributed to people who provided $20-$200 of liquidity in markets that had $600 or less of liquidity. This will also eat up a lot of money (up to $70k/week since Nov 21) and leave very little left which could be used to attract 20k+ liquidity on each market. All rewards should go to a new contract and this should be publicly announced so people stop gambling on whether or not the rewards will be paid to the existing contract.
Declining Incentives for Large LP Amounts
While small amounts of LP are overly incentivized, the opposite is true for large amounts. A possible solution is to scale the rewards linearly to a larger amount - so you would get 100% at 20k and only 10% at 2k. Another alternative is to scale them non-linearily - ex. you could get 40% at 5k, 80% at 10k, and 100% at 20k.