Adding features to the BNT token

I agree that revenue generation is probably the main priority

but dont you feel that we need a way to incentivize organic growth of the bnt token . I dont think buyers will step in and take on the risk, even after the dust has settled, for the bnt apy %

and a reverse dutch auction is pretty different.

this idea is more akin to other blockchain insurance projects which require you to buy their tkns in order to receive insurance protection

maybe instead of some indirect route like staking we simply charge an annual premium for the ILP insurance directly based on the inputs/outputs of the new simulator for that TKN

ILP has always been given out for free… well it was given out based on the length of time from the LP…

if bancor starts charging a fee for ILP, it will cause a stink, however if we throw in additional use cases like what I suggested, then that will help to offset the downside of paying for ILP

Im trying to add additional benefits for holding and buying the bnt token, rather than it simply only being a purchase for insurance ( which they should get the funds back after a certain period of time )

yeah i hear you, i like the idea of insurance existing as a paid/premium/exclusive/qualified thing somehow

i’m ok with an initial stink if it means we end up with a sustainable and manageable product, because in the end people really like things that clearly make them money

the thing about a premium is that it is easier to calculate, we can get data on a TKN volatility and feed that into a premium

pragmatically i don’t know how to quantify a stake amount that justifies some random TKN asking for ILP on a DAO proposal, i actually think we’d just fuck up the risk analysis like we did with LMR :sweat_smile:

It’s an interesting question. If the ILP was limited to fees generates by the staked BNT (and any potential withdrawal fees) then there isn’t too much risk on the protocol, but I do think we need more fee generating features to make this attractive.

what i was thinking was something like:

  • we use data and our simulator to come up with some e.g. 90% confidence that TKN wouldn’t exceed 5% IL vs. BNT over 1 year
  • we offer ILP at e.g. 7%
  • each year we subtract 7% from the staking ledger of anyone who opts in

if we were wrong and TKN had 10% IL then we pay 3% ILP

if TKN had 5% IL as we expected then the 2% difference between the deficit that user would have seen anyway and the ILP premium is given up to help cover everyone else’s deficit

the 2% cost to the ILP holder becomes revenue for everyone else

seems like it would not be so difficult to do as a hybrid of the rewards and ILP contracts

the difficult bit would be the modelling and projections, i think it would only be possible on whitelisted TKN that have been through a specific risk analysis process, and we also have to be confident that the token price is resistant to manipulation by policy holders

the good thing is that we could hire an actuary to do this modelling, as it aligns with how insurance works everywhere else so probably a lot of their skills apply here

i’m seeing this as something that might be attractive to people who want liquidity for their token, e.g. treasury management, and want single sided staking (don’t want to sell TKN to make 2-sided) and want to be able to budget (volatility on the ILP premium would be lower than volatility in the TKN price)

i guess i’m seeing the staking option as another “move fees around” option, whereas premiums is “make people pay a subscription for the service” so it becomes a new revenue model in itself if managed well

the “profit” of staked BNT is hard to quantify, we can’t even quantify how much liquidity is “enough” or “too much” on the TKN side despite ~2 years of trying to do that, and this sounds more difficult to measure than that - Proposal: Limit on-curve liquidity to max(520 x 7 day fees, 100k BNT) - #37 by mikewcasale

remember that a 50/50 2-sided AMM doesn’t generate enough fees to cover IL

so to me that implies that “enough” BNT staked to cover IL would be more than 50/50

so to qualify for full ILP maybe you have to stake 1.5 BNT for ever 1 TKN on curve

at that point “single sided staking” is kind of just semantics and reality is that we’re doing something more like balancer with weighted pools to mitigate IL

Keep in mind that this is a very interesting feature for DAOs. They are more interested in liquidity than anything else, and if they can get that with their own TKN without needed to provide a second token, it is very appealing for them.

As an exmaple, see the strong response from the CROWN community.

yup, that’s exactly who i think would pay a premium for insurance

we can also combine the premiums with IL mitigation strategies like liquidity optimization and SFL

in my mind the premium would be paid on the full staked TKN balance regardless of how much TKN is oncurve or used for whatever revenue generation purpose within the protocol

say some TKN moons 100x in a P&D, we should be free to take 99% of that TKN off curve short term and put some of it back when the dust settles, in BNT terms we’re still providing the same trading liquidity even if most of the TKN moves off curve during the moon, so i see nothing wrong with that

Just curious in your scenario - how do we reactively/proactively move TKN on/off curve? Disregarding the actual trigger mechanism - how should we define price change thresholds or a different trigger for something like this?

When IL exceeds a certain %, then funds are moved off curve to protect the majority of users balances // will also limit the IL their position is exposed to ?

I -think- and I could be wrong here, that this would actually make the IL permanent. Though I think it would reduce potential future IL (but also recovery).

if you do it after the fact then yes

a very simple mechanism would be to have a fixed BNT amount that is the liq on curve

say we allocated 500k BNT to some TKN and it mooned 100x, and we still only have 500k BNT on the curve the whole way up, each 1 TKN that we take off curve on the way up makes subsequent IL less, and if we put the TKN back on the curve on the way back down in the same amount i would assume that means the IL is still impermanent as it’s a symmetric process

if we can imagine 1 mechanism that has the property of reducing IL while keeping it impermanent i’m sure we can imagine others also

i think the trigger mechanism is actually important, because if it’s something that is inline with the transaction and atomic then it can happen before the trade completes (e.g. how slippage works) but if you try to take liquidity off curve after the transaction as a response to it, then you run into problems where you have to be predicting the future about what will happen in the next transaction, which can’t really work as you never know if the pump will continue or dump

ILP worked flawlessly in bancor since v2.1 until this most recent fiasco. Therefor ILP did work, but it had one point of failure, and that button was pressed. Therefor bancor should be able to use the existing ILP framework, but with added extra protections to circumvent what happened during the recent events.

Can we not learn from what recently happened and find a way so that this scenario is no longer repeatable?

this combination led towards the recent events

  1. weak market - cant do anything about this

  2. bnt was already weaker than the rest of the market - bnt was previously being handed out like hotcakes to all the big TKN pools which led to much selling pressure. This has already been turned off so we no longer have this problem. However I argue that because there was not enough incentive to hodl and stake bnt, this contributed towards the selling pressure. People would stake tkns, receive bnt then sell. From here we learned that even with a high bnt stake apy%, this alone is not enough to maintain strong organic growth of the token. I believe bnt needs additional features added to the token that will incentivize hodling. People will hodl the bnt tkn if it either makes them $$ & or if it helps to protect their $$ this is an area where we can try to mitigate our problems

  3. ILP was given out to massive wallets whom then immediately sold the ILP BNT for their respective tkns - this is an area which we can try to mitigate our problems

  4. bnt was shorted on various exchanges like ftx - nothing we can do about this


so if we

  1. find a way to incentivize long term organic growth by adding features to the bnt token
  2. control ILP in a way in which a mass exodus cannot lead to catastrophe

then we will achieve

  1. a more resilient BNT as there is buy pressure from outside the protocol which can support the price
  2. in future events where ILP mints excessive amounts of BNT, the sell pressure of BNT is more controlled

add features that achieve this, then I believe bancor will be steps closer to solving the issue of IL


features ?

  1. the new BNT features should avoid inflationary pressures, however, I would argue that if there was an incentive to add +BNT apy ONLY to those who stake BNT, these BNT lp’s are more likely to re-stake their new bnt rather than sell it down, as it would negatively impact their existing holdings. All other pools should not receive any BNT tokens for lp’ing. like the old system, a high bnt apy will help attract new users and buyers of bnt. however in the old system, tkn lp’s probably would have sold off their bnt earnings for more tkn. with this change, bancor would not have that problem

  2. in order to have ILP - the LP would need to stake a # of bnt in order to qualify for ILP. this amount would be based off the size of the tkn position. since the price of bnt flucuates , and Im hoping for a $$ increase, then the amount of bnt needed to qualify for ILP should be a % rather than a # from the tkn position. When the price of a tkn or bnt goes up or down, this will affect the amount of ILP coverage the user has on their position. With this being said, if a large whale chooses to pull out from their position, and this then in turn affects the deficit in the pool, then perhaps this can also affect the ILP % coverage they have when they pull out.

  3. if we buy and hold bnt… we all want to make money from it. anyone who purchases bnt and takes on the risk should be entitled to some potential profit. Even if the protocol itself holds bnt, it should also be eligiable to these profits. lets think about where these profits can come from

ILP didn’t work flawlessly though, it was being tracked on a cash basis (dune showed mints/burns only) without showing the surplus/deficit of liabilities building up in the protocol invisibly

the deficits existed in v2.1 and were simply calculated and displayed by the v3 migration

everything i’m saying is aligned with what you’re saying isn’t it?

“control ILP” => you have to calculate it to control it, and as i mentioned above, you can’t simultaneously market “single sided staking” and also “ILP if you stake enough BNT to cover it IL”

all i’m saying is do all the same calculations denominated in TKN instead of BNT so we can have ILP and also still have “single sided staking”

1 Like

yep what youre saying is alligned, im just writing down my thoughts and brainstorming as I do

I editied my post above and added this

  1. in order to have ILP - the LP would need to stake a # of bnt in order to qualify for ILP. this amount would be based off the size of the tkn position. since the price of bnt flucuates , and Im hoping for a $$ increase, then the amount of bnt needed to qualify for ILP should be a % rather than a # from the tkn position. When the price of a tkn or bnt goes up or down, this will affect the amount of ILP coverage the user has on their position. Therefor as the user maintains their position at times they may need to buy additional bnt in order to maintain full coverage. With this being said, if a large whale chooses to pull out from their position, and this then in turn affects the deficit in the pool, then perhaps this can also affect the ILP % coverage they have when they pull out. ← since the blockchain knows when an event like this occurs, when the whale removes their position, can the protocol not then do the math and see what happens to the potential deficit then dictate ILP based off that?

tl:dr - so ILP is based off of the lp needing to stake x amount of BNT tokens, but also ILP coverage is also affected by the possible deficits that may occur when a user pulls out

ah so, in b3 everyone has the same deficit and surplus all the time for a given TKN

so whether you are a whale or a minnow, the % deficit doesn’t change when you withdraw

this is a key difference from v2.1 where everyone has unique positions

do you feel that perhaps this idea of socialized IL may need to be revisited?

We can now say that after this was initiated, it didnt pass with flying colors

purely from the perspective of offering a credible insurance product, having a single number to manage risk and premiums against seems better actually

we can simply take TKN off people’s staking ledger faster than deficit accrues to that TKN, boom, insurance product done

otherwise we have to do all these things like you say:

  • assess individual risk profiles
  • ask people to regularly top up their balance, so it becomes actively managed rather than passive
  • treat whales and minnows differently
  • rethink the “single sided staking” value prop

probably other stuff too