Recover Basket and Support BANK during distressed times

I agree the rewards looks generous. To be fair to everyone we can introduce a concept of slippage similar to AMM, because for each ETH we’re supplying to the Basket, the basket factor is improving and hence the rewards should be less. So protocol will issue only 10% rewards for 20% of collateral supplied to basket.

Also the rewards can be locked for a longer duration may be 20 expansion auctions where the BANK are burnt and 5% rewards gets unlocked with each auction or something similar. This is required so people don’t dump BANK as soon as its unlocked. May be principal can be locked as well, but I’m not sure if that’s a good idea we can ask for suggestions.

Also people who are in BANK only pool and BANK-ETH pool are the ones who cares about BANK the most, so may be for the first time we can offer these bonds only to people in these pools in proportion to the BANK they are holding. They are also least likely to dump because this impact their existing position.

Can we brainstorm more around this idea ?

As a BANK only and BANK-ETH pool participant, I like the idea of rewarding more for holding/locking up BANK. I am willing to tie up liquidity, especially to be helpful to the protocol in the longer run. This proposal would encourage more long term participants to join/stay if there are just modest incentives to do so.

1 Like

Our current thinking on this is the following:

  • We want the following assets to be held by the Protocol:
    • BANK-ETH sLP
    • ETH (in the basket)

In theory BANK could also be added, but ultimately the protocol owns a lot of BANK, and having other assets we can buyback BANK with would be more valuable long term.

We can sell FLOAT however, as long as the backing is greater than the existing backing in the treasury. E.g.

  • 60% basket factor => up to 67% premium on FLOAT.

So instead we’d suspect that we can do the following bonds / trades:

  • BANK-ETH LP for FLOAT (vested over X days - X = 8?)
  • FLOAT-USD LP for FLOAT (vested over X days)

In addition we could auto stake this FLOAT into the FLOAT-only pool for BANK.

1 Like

is this a bank-bond idea?

1 Like

I’m not sure if I fully understand this, can you please explain by example. I would like to understand how we can improve the basket factor with this over time.


Say 1 FLOAT-USD LP token has a value of $10, we should be able to support $10 of FLOAT at 100% backing.

Part of the point of FLOAT (and what we’ve been demonstrating over the past month), is that we don’t need to have 100% collateral to be stable however. Instead we can utilise these bonds to boost the basket as long as the collateral ratio of the bond is above our current collateral ratio it is net positive.

Currently basket factor is floating around 67%; so any new FLOAT at a % collateral > 67% is a net positive for the basket.

Let’s say we were targeting 75% collateral, then we can take 1 BANK-ETH LP token right now at $10 in return for the equivalent amount of FLOAT at 75% collateral. Which would have a dollar value of $13.1 (1/0.75) = 31.5% premium.



  • Increases the basket factor faster
  • New tokens in the basket are yield earning (LPs earn fees / can be staked)
  • Acquires strategic tokens (e.g. liquidity provision, that ensures we have a minimum liquidity volume)
  • Increased liquidity and trade volume


  • Gains a substantial premium for initial investment.
  • Can acquire FLOAT without substantial slippage without waiting for an auction.
  • Can participate without requiring the technicality of Botting auctions.


  • Faster we’re back to 100% basket factor, the more BANK burns
  • Increased liquidity means easier integrations (Chainlink has minimum volume requirements for instance)
  • BANK-ETH LP purchases will require BANK to be removed from the market (BANK buy pressure)



  • FLOAT gets instantly dropped on the market; at worst this can lead to any value we gain from bonds being lost back to the process of reducing supply; if this drop is greater than the collateral ratio (i.e. the dumpers take losses) then then Protocol can actually profit from that movement.
    • Mitigation: Making the FLOAT vest over a period of time (gives a longer time for the earning potential of the asset to exceed the amount given out).
    • Mitigation: Wrap the FLOAT up more permanently in staking (already exists and ~25% APY).
  • The Protocol assets we acquire go down in value (e.g. as we’ve seen with ETH)
    • Mitigation: We can handle the sharp reductions in basket value (see ETH’s price drop) but this will put more pressure on BANK
  • Bonds could increase volatility of FLOAT as it is an additional market force (both positive / negative that needs to be balanced).


  • FLOAT price is below target price when they desire to sell, leading to a lower premium (they could always wait for the auction).
  • FLOAT target price moves downwards due to the market tracking nature of FLOAT.

There are certainly trade-offs involved and this should be a limited program when started to see the effectiveness, but in theory it should be able to boost the basket far quicker than the 1-2% per auction movement.

For understanding:
Let’s say we were targeting 75% collateral, then we can take 1 BANK-ETH LP token right now at $10 in return for the equivalent amount of FLOAT at 75% collateral. Which would have a dollar value of $13.1 (1/0.75) = 31.5% premium.

  1. If we target 80% collateral premium will be 1/0.8 = 25% and 1.11% premium if we target 90%.
  2. So we have to go step by step, first bond offering with 75% collateral target, second offering is 80% collateral target and so on.
  3. At 75% target the protocol will issue $13.33 FLOAT for $10 FLOAT-USD or BANK-ETH LP
  4. The protocol will only benefit from the difference b/w current Basket Factor and Target Basket Factor, in this example 75% - 67%. Hence would require a lot of floats to be issued, which is actually a good thing and will also address liquidity concerns.

I can think of below risks:

  1. If overall percentage of FLOAT issued using LP is greater than the minted using ETH, protocol may end up having less liquidity to support the auctions.
  2. In a death spiral kind of scenario protocol may end up holding LP with only FLOAT and BANK. Hence we need to ensure there is a balance between the ETH collateral and LP collateral in the basket. Basically ETH collateral will act as floor in worst case scenario.

I also think basket factor should be considered in 2 parts:

  1. Underlying non protocol assets (ETH in Basket & LP + USDC in LP) - We should have a target ratio for this, above which we can burn BANK otherwise we restock the basket to this target ratio, let’s say 80%. This should also be used to decide the expansion/contraction scenarios.
  2. Total assets, including BANK and FLOAT - This provides a general health of the protocol

We should also think about having multiple offerings, with different maturity and fixed amount of FLOAT to be issued, for example:
X FLOAT @75% target (33.3 % premium) for 6 months
Y FLOAT @80% target (25% premium) for 5 months
Z FLOAT @85% target (17.6% premium) for 4 months
Amount of FLOAT to be issued should be decided based on the target ratio (where we want to burn Bank)

Overall I like the idea, it address concerns regarding volume, liquidity, float supply , restocks the basket and support BANK.

Yes this is a great idea. Bonds would be an additional supporting force for float. We just need to play around with the parameters so that its slowly introduced into the system. Seems to be working out well for ohm so far.

This is an interesting distinction, and I can certainly see the motivation. The issue then becomes that LP tokens (being 50% protocol tokens i.e. BANK / FLOAT would not be counting towards the basket factor). So 100$ of LP on the open market only counts as 50$ towards the basket factor. So we will end up reducing the basket factor (target ratio) with these assets.

What do people think about these timelines, It’d be great to have them locked up for that long but unsure of peoples appetite for that.

Thanks for starting the thread @nps and :clap: to everyone else for their contributions

To be really clear though, is the point of this thread not to support the price of BANK by selling assets held by aligned stakeholders, such as FLOAT-USD V3 LP and BANK-ETH sLP, for Bank?

If that’s what everyone really means, I agree with the general approach. Otherwise, I’m not sure there is enough of an incentive to sell Float for a future premium. For example, how much ETH (as a proxy for the value of the basket) needs to rise for Float to rise by 33%? Increasing the basket factor is beneficial for Bank, so I can still understand why this is a consideration. I just wonder whether the initial focus should be on Bank, particularly given that (according to this dashboard - Dune Analytics) there are a lot more holders of Bank (4,630) compared to Float (350).

Some other thoughts:

  1. What is an ideal cap on this scheme? Ie. how much BANK can be issued in this manner?
  2. Is it a risk to accept LP tokens for longer than 1-3 months given that liquidity can be fleeting and can move on to other DEXs over longer periods? Eg. Maybe Curve becomes the primary venue for Float, and somewhere other than Sushi for Bank/ETH
  3. What other collateral types might the protocol want to accept in the future?
1 Like

I think that was @nps original intention behind the thread so:

  • 100$ of FLOAT-USD V3 LP can be exchange for $100 of (minted) BANK + some premium.

This would certainly support the price in the short term (as that premium gets evaluated to market demand) and is the most direct method of doing it. (I do still think it would work, but evaluating premiums and volumes is going to be tricky.)

The reason we started considering FLOAT as the “reward token” was because the basket factor means we have a very clear value proposition for why we have a discount, and a number to relate that discount to.

The thing to highlight here is we expect the FLOAT price to be constant - you gain from the additional FLOAT you receive as opposed to the FLOAT price increasing

The focus is ultimately still to provide more utility to the protocol - and hence support BANK holders (more assets in the vault, including BANK-ETH means we have more trading volume and more integrations).

Putting some numbers on it:

  • The basket contains 2037 ETH @ $2.1k = $4.31M
  • The float market cap is $6.49M at target price.
    • Hence the basket factor of 66%.
  • All else being equal (FLOAT demand is 0, the basket doesn’t earn yield, we gain no ETH through auctions, target price stays constant), we’d need an ETH price of $3.19k to hit 100% basket factor. Or equivalently ~1,000 ETH.

For FLOAT, the aim is to have the value of the assets outpace the reward of FLOAT / excess demand to support the additional FLOAT added.
As for a cap on the scheme I’m not entirely sure - I think we need one perhaps 5% as a starting value that we can ramp up / down as we see how quickly the assets sell for. This can also be balanced against the liquidity reward reduction (removing liquidity rewards and moving them the the bond system should be an equivalent incentive except we have the LP tokens for good rather than temporarily).

This is more easily handled - the basket bag system, and the upgrade-ability of the basket means we can transition any assets to any other assets.

Long term this plays into the “basket of goods” argument, so anything that is deemed to be an acceptable measure for the crypto market to maintain purchasing power.

1 Like

As for a cap on the scheme I’m not entirely sure - I think we need one perhaps 5% as a starting value that we can ramp up / down as we see how quickly the assets sell for. This can also be balanced against the liquidity reward reduction (removing liquidity rewards and moving them the the bond system should be an equivalent incentive except we have the LP tokens for good rather than temporarily).

Yeah I think following a similar approach to Maker here with % caps on each collateral type makes sense and 5% seems like a good starting figure which can be scaled up if the demand is there. Additional caps could be set for other DEX venues such as Curve if those pools if they came to take volume share.

1 Like

Thanks @paulm - this is all very clear

The proposal makes a lot of sense. It should work well with the expectations for much better liquidity for Float in the coming weeks and months too.

While I liked the idea of issuing FLOAT for the LP tokens, I don’t believe this is a sustainable way of implementing the idea of Bonds. It may help in the short term with some liquidity concerns but there are associated risks and it’s difficult for simple de-fi users to understand and participate. For a newbie project like FLOAT its also too risky to hold its own assets as it can lead to a death spiral kind of scenario.

So here is what I think, how we can implement this idea going forward.

Basic Rules

  • Float bonds will be issued as zero coupon bonds (discount bonds)
  • Bonds will have linear maturity and discount rate
  • Bonds will have an attribution factor, i.e. percentage of bond price which will go towards the basket. Higher attribution factor will lead to lower bond yield and less bonds issued by protocol, and a lower attribution factor will lead to higher bond yield and more bonds issued by the protocol. The factor can be adjusted based on market condition.
  • If basket factor is below 100% FLOAT bonds will be issued for WETH or USDC (similar to expansion scenario in auctions)

Best to explain using some examples: lets say attribution factor is 50%, FLOAT target price is 1.5 and time factor is 365 days

  1. If Basket factor is 60% - FLOAT bonds will be issued with a discount of 20% for $1.2 [(100% - 60%) * 0.5] with a maturity of 145 days [365*40%]. Net positive for the Basket is the difference of 20%. For BOND holder YTM (yield to maturity) is 25% in 145 days.
  2. If Basket factor is 90% - Float bonds will be issued at a discount of 5% for $1.425 [(100%-90%) * 0.5] with a maturity of 36.5 days [365*10%]. Net positive for the Basket is the difference of 5%. For BOND holder YTM is 5.26% in 36.5 days.
  3. If Basket factor is 99% - Float bonds will be issued at a discount of 0.5% for $1.4925 [(100%-99%) * 0.5 ] with a maturity of 3.65 days [365*1%]. Net positive for the Basket is the difference of 0.5%. For BOND holder YTM is 0.5025% in 3.6 days.

Liquidity – Liquidity remains the biggest risk, what if the bonds matures but there is no liquidity to sell the matured FLOAT
Mitigation – Locked FLOAT can be used to provide 1 side Liquidity on Uniswap v3 in FLOAT-USDC pool to earn additional yield, protocol can offer this as an option to bond holders.

Protocol can decide to use 50% (or X%) of attribution factor to provide liquidity. The amount can be converted to USDC and can be used to provide the liquidity on the other side on Uniswap v3 FLOAT-USDC pool. The USDC in pool should also count towards the basket factor. This will also provide some diversification in case of sudden price drop for ETH.

The liquidity will grow organically as more bonds are issued.

In future a liquidity manager can be created to manage the liquidity on Uniswap v3 for all locked FLOAT & USDC but this can be a separate topic for a later release.


  1. Basket can be refilled very quickly by offering lucrative yields.
  2. These lucrative yields can bring more people in the ecosystem and eventually more adoption for the protocol. Best way to get more people involved is to offer the better risk adjusted returns. Hopefully bonds can help with that.
  3. FLOAT supply becomes elastic as more FLOAT can always be issued either using bonds or auctions. When basket is less than 100% more FLOAT can be issued using Bonds and when basket is more than 100% more FLOAT can be issued using Auctions.
  4. Less sell pressure on BANK from auctions because of increased Basket factor
  5. Basket does not contain any LP tokens and contains core assets only. So it reduces the risk of death spiral kind of scenario. If needed LP token can always be purchased using ETH at a later stage when protocol matures. Other assets can also be purchased for diversification, also the ETH can be staked.
  6. Diversification of basket by adding USDC and putting locked FLOAT to use. USDC will be helpful when there is sudden price drop for ETH
  7. Organic growth of liquidity for USDC-FLOAT pool

Interface -
There should be a front end interface to mint bonds using either USDC or ETH (like the FLOAT mint interface) where both USDC and ETH users should be able to buy bonds seamlessly. Behind the scene protocol can convert stable coins to ETH or vice versa and add liquidity on Uniswap v3.
This will allow stable coin holders to lock-in a fixed yield.

FLOAT issuance -
FLOAT can be issued either using a window system or using an AMM system. May be a window system to start with and then an AMM type of system for future.

Window system – We can offer bonds based on target basket factor, start with 70%, 75%, 80% etc. Amount of Bonds to be issued can be calculated based on basket factor and attribution rate.

AMM system – In the long run an AMM type system will be better where for each bond issued increases the basket factor and hence reduces the yield and time to maturity.

Other features which can be included -

  • If basket factor is above 100%, Bonds can be purchased for WETH+BANK (confidence expansion scenario in auctions). This would require a call premium for people to surrender their bonds and forgo the pending yield on the bonds. WETH can be used for accrued price and call premium, the differential in BANK can be burnt. FLOAT price will be adjusted upwards in such cases. Overall the logic for adjustment of FLOAT price should be harmonized with auctions logic.
  • If there is a lot of demand for bonds, then bonds can be offered/purchased using dutch auctions similar to FLOAT.

wow, totally agree with this! very detail answer and suggestion.

This all looks good - I like the formalisation of the attribution factor and agree that having non-core assets is very valuable and more resilient.

My only concern is diversifying the basket to include USDC - while this will provide a lot of easy stability and protect against ETH price movements it does attack one of the most unique points of FLOAT.

We’re not only stable because someone else’s stablecoin is doing all the heavy lifting. This kind of move I generally see as a short term play that means you can never become the dominant stablecoin - you’re always relying on someone else’s token that must be as big or bigger than yours.

Centralised stablecoins are the algostables’ drug of choice - it makes your life so much easier because you outsource all the stabilisation work to someone else, but the more you rely on them the more your stability becomes hooked on theirs and you can’t go back without a long painful separation.

Completely agree with you on this @paulm , I would be very much against diversifying the basket to include USDC.

In an ideal world I would also avoid holding USDC or any other pegged stablecoins in the basket, but as long as we’re holding only the amount required to provide the liquidity in the pool we should be “ok”. In long run when we’re listed on CEX’s and there exists multiple DEX’s pool, USDC can be converted back to ETH or other strategic assests. Also it addresses liqidity concerns in short term with limited risk.

With regard to this would suggest staying away from USDC, as it’s a potential legal attack vector. I think there will be a market for crypto-native sable currency that is not tainted by $.