Problem: The protocol is out of money and has no means to control the BNT price in the short term, thus the protocol is at the mercy of the external market at all times.
Proposal: TKN price mimicking
Build up a reserve and adjust the BNT price when a larger IL occurs on a pool:
-
TKN crashing (large TKN pools) against BNT: stop the burn, sell off some BNT for TKN x (DAI, ETH, wBTC, LINK - examples)
- this decreases the IL by mimicking the TKN price
- multiple large pools should be taken into a consideration. If only one pool is monitored and mimicked the other ones could accrue more IL when this strategy is applied, this could lead into deficit [mostly when a blue chip TKN is mooning the other blue chips are mooning too]
-
TKN mooning against BNT, or BNT crashing: Buy back BNT, start the burn → increase the BNT price until a certain TKN/BNT->IL threshold is reached
-
Neutral state (TKN against BNT is in a certain range):
- Burn BNT or vBNT in batches, always have a reserve for rapid price changes [possibly fill up off curve pool reserves if in deficit]
-
This could be done first manually, after more data on when is it the most beneficial to swap and experimentation with TKN:BNT price ranges. Later it could be automatized.
-
The goal is to reduce the protocol expenses on paying off IL by adjusting the BNT price to TKN pools with the most IL.
The protocol could monitor the TKN/BNT price deviation (deficit) in large pools, or anticipate large withdrawals and act accordingly to minimize the amount of IL compensations.
Afterthoughs:
BNT would basically become an index of large TKN pools.
I would discontinue the BNT and vBNT burning. It has not much usage when the protocol is transitioning to off curve reserves and will no longer compensate for ILs in BNT. BNT is an IL yard stick which changes with BNT price, so the BNT should follow the prominent TKN prices to accrue the least IL.