POLL: Initial Forth-Bond Parameters for HourGlass Launch

Ampleforth Governance
1

Hi all,

As you may be aware, we are launching the HourGlass protocol next Monday, October 31st. HourGlass is a permissionless convertible bond market that enables the FORTH-token to be used as collateral to borrow stablecoins – at a fixed-rate, and liquidation-free.

Purpose

The purpose of this post is for the team to get a sense of the initial bond parameters (rate, duration, etc.) that the community is most aligned towards for the purpose of the initial launch.

NOTE: The act of ‘Listing’ an HourGlass bond does not bind the DAO, it merely gauges market (lending) demand. The DAO’s treasury assets are only moved once they choose to ‘Activate’ the bond. So, the outcome of this poll and subsequent signal vote does NOT bind the DAO to utilize its treasury assets in any particular way. HourGlass is completely permissionless, so anyone is free to deploy a bond with different parameters if they choose, but we are hoping with discussions related to this post to get some consensus before launching Monday.

Background

If you are new to HourGlass, please see the links below:

Rationale

As a refresher, there are several upcoming expenses that the DAO can expect in the coming weeks. These include the following:

  • Deepening SPOT/USDx Liquidity Pool (most significant)
  • Providing liquidity for an AMPL/SPOT Pool on ElasticSwap
  • Creating a btnFORTH-USDC ElasticSwap pool with zero-IL
  • Other Projects on the MIRO board

After discussions with the Ampleforth team, it seems like the best use of an HourGlass bond is to borrow stablecoins against the FORTH token to deepen one of the liquidity pools (SPOT/USDC or SPOT/USDT). The bond could also be paid back (partially or fully) using the fees earned from that same pool. Given the market downturn, we believe debt-based financing makes more sense than selling FORTH tokens on the open market at current prices. Because HourGlass bonds are liquidation-free, this ensures the FORTH DAO retains ownership while seeding much needed liquidity early on.

Implementation

The following are the key parameters and their significance:

Parameter Example Significance
Interest Rate 5% Since these are zero-coupon bonds (meaning they make 0 interest payments), the interest rate is the overall borrowing cost the DAO incurs if they activate the bond, hold the USDx, and reclaim collateral upon maturity.
Tranche Ratio 20/80 Tranche ratios impact the risk profile of the A-tranche
Penalty % (optional) 2.5% The % of Z-tranche that’s forfeited to Bond-holders in the event that the DAO defaults. This can be used to incentivize lending demand
Maturity Term 1 year The amount of time the DAO has to repay the loan
Stablecoin USDC The repayment token

Below are three bond configurations we came up with. If there is one option that gains general consensus, we can deploy that bond to concentrate liquidity. The DAO can then view the lending demand (in the form of deposits), and choose whether to activate that bond. If over the next few days, no consensus is reached, HourGlass can deploy multiple bonds and gauge demand that way.

  • The rates below factor in market research such as borrowing rates on Aave, those offered in TradFi, and those seen in CeDeFi (such as Maple Finance). They are matched with appropriate tranching ratios so that the risk profiles are comparable
    • 5% interest (APR) w. 20/80 Tranche Ratio
    • 7.5% interest (APR) w. 25/75 Tranche Ratio
    • 10% interest (APR) w. 30/70 Tranche Ratio
  • Penalty: 2.5%
    • Since the majority of the DAO activities identified are LPing for pools with zero or low-IL, we don’t think there’s a significant risk of default. Indexing on a higher penalty % should help incite more lending demand
  • Stablecoin: USDT
    • The DAO will need both USDT and USDC for Liquidity Pools. We do not have a strong opinion on which the HourGlass bond should be used for, but chose USDT due to the recent censorship questions with USDC
  • Maturity: 1 year
    • Based on discussions on the timeline of SPOT launch, we think 6 months or 1 year is enough time for the DAO to start earning revenue on SPOT minting/redeeming, trading fees etc.

Note on Amount/Volume: There is no strict minimum volume required because the DAO can choose whether to activate the bond whether its $1k, $10k, or $500k. There is likely a minimum volume to make it worth the gas required to activate the bond, but this can be discussed after observing the initial demand. We think goals for the initial bond could be between raising $10k - $150k.

We’d like to understand the community’s opinion towards this initial configuration with the poll below. As a reminder, HourGlass is completely permissionless, so anyone is free to deploy a bond with different parameters if they choose. This initial deployment by the team is NOT binding the DAO to any commitment.

If you differ on any of the parameters, please feel free to comment below on alternative parameters.

Poll

Select all the options that you agree with:

  • The configurations are all acceptable
  • Needs different interest rate and/or tranche ratio
  • Needs different default penalty %
  • Needs different stablecoin for repayment
  • Needs different maturity

0 voters

Select the bond you are most likely to buy (or expect lenders to buy):

  • 5% interest (APR) w. 20/80 Tranche Ratio
  • 7.5% interest (APR) w. 25/75 Tranche Ratio
  • 10% interest (APR) w. 30/70 Tranche Ratio

0 voters

3 Likes
2

Great work with the proposal. I’ve had a chance to explore the app as well, and it will be good to go shortly after a few kinks have been ironed out.

I think we can afford to take some risk here and lock in a 10% APR with 30/70 ratio, but that is based on an expected market reversal (which very well could fail as well). So that being said, perhaps a combination of 5% AP with 20/80 Tranche ratio and 10% APR with 30/70 Tranche ratio could be reasonable.

1 Like
3

I’m proud to see the HourGlass team launch a more complex and useful protocol than what it had a year ago.

Disclosure: I am speaking as a Forth DAO member, not as an HourGlass member. Although I am a co-founder and an advisor, I have never received any income or part of the grant. The 3rd developer who took my place is @BearMauler.

I think it’s helpful to spell out a few risks as I understand them (and HourGlass should evaluate and document risks to educate and protect users). Bond holders bear more risks as bond issuers have more rights. These risks should be rewarded by a properly configured bond:

  1. This is also a risk with ButtonZero bonds: bond holders are exposed to FORTH price after maturity. Holders earn an initial 10% interest in terms of senior or A-Tranche claims on buttonFORTH. The interest is fixed because initially 1 A-Tranche = $0.91 and the bond contract protects against a decrease in buttonFORTH value up to the 30 ratio, so likely 1 A-Tranche = $1 at maturity. After maturity, the bond contract sends buttonFORTH in a waterfall sequence to the senior A-Tranche contract, and any remaining to the junior Z-Tranche contract. A-Tranche is now valued in terms of buttonFORTH, which is fully exposed to FORTH price changes. A-Tranche is no longer protected against a decrease in buttonFORTH value. If FORTH price drops 50% between maturity and redemption, then these holders redeem just $0.50 (at -45% interest).

  2. Back to HourGlass bonds: if holders do not want to redeem at -45% interest as in the previous scenario, or simply do not want to be paid in illiquid DAO tokens, these holders may see their interest rate go to 0% (at least this is still better than -45% interest). If FORTH price drops, it makes sense for issuers to wait for FORTH to rise back to its original price before repaying. As long as there are A-Tranches remaining (as long as there are holders who have not redeemed), issuers can wait a long time before repaying.

  3. Holders may not be compensated by the 2.5% Z-Tranche penalty. If FORTH price drops > 70% between activation and maturity, the underlying ButtonZero bond contract does not send any buttonFORTH to the Z-Tranche contract. After maturity, 1 Z-Tranche = $0 regardless of FORTH price movements. The penalty of 2.5% is also worth $0. Even a high penalty of 50% does not mitigate this risk.

It makes it easier for DAOs to configure their bonds if HourGlass valuates its bonds like how TradFi valuates their callable/convertible bonds.

The risk of default is high if the DAO wants to be a long term LP and thinks the next round of bonds won’t raise as much as it needs. I believe the penalty should be 20% to reduce default risk and attract bond holders that want to be paid in USD.

I think there is a chance FORTH can go below the 30 ratio in 1 year. I suggest managing this risk by reducing the maturity from 1 year to 2 months and keeping the interest the same at 10% APR (does the interest input to the contract change?). This lets the DAO set attractive prices for short term market conditions and will probably attract more liquidity. Although liquidity varies from bond to bond, the DAO can keep the liquidity whenever it likes by defaulting.

4

No disclosures to report on my end. ForthDAO member. AMPLer and AMPL NFT holder.

If bonds will be short-term, then we may further consider selling A-tranches at a lower price ($0.80) to attract larger volumes of liquidity from market participants, contingent upon the upper bound for the number of FORTH tokens bonded.

25% interest rates are outstanding in a bear market.

5

Looks like we both have some wild numbers. $0.80 is a 25% yield in 2 months, 150% APR may be too high. My suggestion of 10% APR may be too low too, this puts yield at 1.6% and the bond at $0.98, too close to $1 for the lender to choose to buy at market price instead (but if we configure it right we can sell this to whales as the bond has no slippage).

I guess a $0.95 bond that yields 5.26% at 63% APR is a good in between, but I have no data to back this up.

6

Thanks @coldpress & @Bhau for some discussion! Love to see these type of points brought up, as there isn’t a single answer to problems like this.

To address the 3rd point about the penalty: it’s true that holder’s “may not be compensated by the 2.5%”. However, this is only true when the Z-tranche is worthless (i.e. price drop of % equal to Z-tranche ratio). The purpose of the penalty is not to eliminate risk altogether, but actually to mitigate impact to the lender of the 1st risk that was pointed out. (i.e. the lender being exposed to FORTH price after maturity). The penalty incentivizes the DAO/borrower to pay back (in situations where the Z-tranche is not worthless). This is also why the penalty need not be something large like 50%. If the desire is to protect lenders against a price drop, it makes more sense for the DAO to choose a larger A-Tranche ratio and pay more interest to compensate.

To address the 2nd point, lenders are only able to able to redeem repaid stablecoins before maturity, and must wait until maturity if they are to redeem A-tranches. This is by design in order to ensure the DAO has the first opportunity to retain it’s ownership of the FORTH token. Thanks again @coldpress for helping to point out these intricacies!

For the maturity date, it’s important to note that the price of FORTH over the course of the year does not matter to the lender. It’s only the price at/after maturity. We think with the current market uncertainty, it makes more sense to issue a longer term bond - however, if there is an appetite for a shorter term bond, we can certainly help to list one.

Based on the poll results at the time of this post, we’ve gone ahead and listed the following bonds, but the HourGlass team can help to list a different configuration if there’s enough support.

1 Like
7

I’m trying to understand the rationale behind using APR terminology. If the duration is only for two months, why scale it out to per annum? (25% x 6 = 150%). To those unfamiliar with TradFi, seeing this “big” number of 150% might serve as a deterrent.

Anyhow in the absence of data, better to proceed cautiously.

8

Thanks for the detailed reply.

Wow, I didn’t expect that. This belongs in the HourGlass docs.

My 3rd point was not really about a high penalty. I mentioned the high penalty only to highlight that 2.5% penalty and 50% penalty are both worth 0 in that scenario, I did not mean for the penalty to protect against a price drop.

I agree, a 10/90 bond protects bond holders against price drops. It makes it more likely that the A-Tranche is worth more and that the Z-Tranche is worth something. Both are good for bond holders if the DAO defaults.

I still think if bond holders want to be paid in USD by maturity, then bonds must have a high penalty to incentivize the DAO to repay in USD to avoid the penalty.

Let me rephrase my 2nd point because it was not addressed: this scenario is after maturity, the DAO defaults, and bond holders don’t want A-Tranches, they want to be paid in USD no matter what. The DAO can wait forever to repay USD. These bond holders may have to wait a long time to get their USD. Over a long time, the interest rate goes to 0%.

You have to be careful advertising fixed interest rates and emphasizing USD returns, and even more careful if you do both. Neither of these are guaranteed if the DAO defaults. If bonds holders get the idea they will earn “stablecoin + interest” (as you say in your tweets and docs), you should at least incentivize repayment of USD. This is why I suggested a high penalty here:

The longer the maturity, the harder it is to predict FORTH’s price at maturity.