Game theory analysis
Consider the situation where the price of vBNT is close to 1 BNT, and the vBNT/BNT pool is reasonably deep. Under this situation, there is no excuse to not own a BNT stake.
Erin is a cryptocurrency enthusiast with 10,000 USDC in her wallet, and is actively seeking a new investment opportunity. She does not yet own BNT, and there is another project that has won her attention. From a purely theoretical standpoint, Erin has a strong incentive to buy BNT, and stake it, no matter what her ultimate investment decision is. To illustrate this, we can consider the extreme case.
Erin purchases as much BNT as she is able, and stakes it in a whitelisted pool to receive an equal value in vBNT. Then, she immediately sells the vBNT for USDC to immediately recover her initial investment, save for gas fees and the small vBNT/BNT price discrepancy. She is now back to where she started with 10,000 USDC (minus a little), and immediately invests in the other project.
Erin’s BNT stake on Bancor will be accruing fees, regardless of the fact that she has sold her vBNT. As time progresses, the amount of BNT she can withdraw per unit of vBNT can only increase. This means that at any point in the future, if Erin wishes to purchase any amount of BNT less than her staked amount, she can do so at a discount; Erin’s investment in the other project is irrelevant.
For the sake of the illustration assume that Erin continues with her investment strategy for three years, and does not return to Bancor to withdraw her stake at all during that time. In this period, the price movements of BNT are also irrelevant - the only important consideration is that the amount of BNT associated with her stake has increased, which is all but guaranteed.
Assume a modest 15% APY over these three years, resulting in roughly a 50% increase in the total number of BNT associated with Erin’s stake. Even in the worst case scenario, every vBNT Erin can get her hands on is worth 1.5 BNT upon withdrawal of any portion of her initial stake after this 3 year period. If the price of BNT becomes the single most valuable asset in the universe, and it is completely infeasible for Erin to access her entire stake - it makes no difference. She is still in the coveted position of being able to purchase BNT at a minimum 33% discount, at her leisure, until she has withdrawn her full stake. This is also true if BNT becomes relatively worthless - she still gets to make a 33% profit on every vBNT to which she has access.
In this scenario it is also important to remember that Erin has two options for obtaining vBNT: 1) She can purchase it directly from the vBNT pool, or 2) she can buy BNT and stake it to get vBNT. Either way, the vBNT she gets can be used to withdraw her 3 year-old stake.
Consider the alternative; Erin decides not to purchase BNT. The markets play out in an identical fashion, and she arrives at the same point three years later, but without an existing BNT stake. She is now unable to make the free, immediate 33% profit on staked BNT, and without any obvious advantages.
Even assuming Erin has no interest in speculating on BNT, the most risk-averse choice is to buy it and stake it, provided the generated vBNT can be sold near the BNT price at the time of deposit.
This illustration glossed over some potentially important details. For example, the ‘closeness’ of vBNT sale price to BNT at the time of deposit is critical, and the gas costs associated with setting up the deposit were also ignored. It is also assumed that the vBNT liquidity pool is sufficiently deep that a large swap can be performed with minimal slippage. However, even taking these cautionary notes under consideration, it requires a special level of certainty in alternative investment options to dispute this deduction.
Conclusion
A deep source of vBNT liquidity can fundamentally change the mechanics of serving as a BNT liquidity provider. It offers BNT stake owners leverage and hedging capabilities, and the ability to stake vBNT for additional yield-earning. The limited volatility of vBNT relative to BNT could attract speculators to the platform, and help to engage an important, underserved day trader demographic on our exchange. If the pool is able to achieve sufficient depth, it becomes trivial to demonstrate that there is no good argument to not own a BNT stake, with only modest assumptions. The new features of the vBNT token are already lying in wait of a pool deep enough to support them - regardless of the outcome of the on-chain voting decision of this BIP.
The on-chain voting for this proposal determines the outcome of the vBNT burning scheme. If successful, a variable proportion of swap fees on whitelisted pools will be used to purchase vBNT via the vBNT/BNT pool, and burn it. As outlined above, this creates a positive feedback loop where BNT tokens are attracted to the protocol and staked, and may result in a situation where additional BNT staking is required to unlock existing, mature stakes. In any case, the amount of vBNT available will never suffice to unlock all BNT staked on the protocol ever again, creating a runaway increase in TVL over time.
Together, these features should strengthen the unique value proposition for Bancor, and further distinguish it from other decentralized exchanges.
Economic bug bounty
This proposal has been carefully considered, and composed with input from the core development team, and its economic consultants. All reasonable efforts have been made to anticipate the economic consequences of the Bancor Vortex, and the discussion points included in this document attempt to capture our views as thoroughly as possible. We are convinced that the proposed implementation withstands scrutiny, and we invite criticism from the crypto community. Therefore, we are announcing an economic bug bounty.
A compelling argument for the presence of an economic bug that renders the proposal unfit for implementation in its current form, will be eligible for a bounty of 5,000 BNT.