Peg Stability Module

For USDH to become a ubiquitous stablecoin on Solana and beyond, its peg will need to remain immensely resilient. This post will present an upcoming implementation of Hubble’s Peg Stability Module (PSM); what its purpose will be, and how it will work.

Please see addendum regarding UST and USDC decision shift.

Introduction

When Hubble was launched, the USDH peg was maintained via arbitrage as part of the Redemption Mechanism. This model would have made USDH redeemable for collateral assets on the protocol when its market price was below peg. Moving forward, redemptions will be removed, and a PSM will soon be implemented, which will be a 1:1 minting/burning mechanism with UST.

This will enable risk-free, zero-fee arbitrage between USDH and UST on Hubble that also builds a treasury for the protocol. The PSM will mean that USDH could potentially be backed by a large % of UST (DAI is backed by ~30% of USDC due to the Maker PSM).

A PSM has thus far proven a more sustainable model than the Redemption Mechanism, as is displayed in the MakerDAO vs Liquity dynamic.

Why Remove Redemptions

The Redemption Mechanism, much like Recovery Mode, was inspired by the Liquity model. The Redemption Mechanism was intended to lift the USDH price when it fell below peg. Users would have been able to buy below-peg USDH from the market, and redeem it for $1 of crypto from Hubble. This would have raised USDH demand and reduced the amount of USDH on the market, thus lifting USDH price back to peg.

This feature made USDH attractive as a stablecoin that could be arbitraged for crypto assets at a low fee. However, in this model, Hubble wouldn’t have had a treasury, thus redemptions would have taken place against user collateral.

From the blog (by @DudeBro):

“Looking at data from one year since launch, Liquity’s users experienced more than twice as many redemptions (1,266) as liquidations (532), and redemptions were much less easy for users to prevent.”

“Looking closer at the data, around 17% of redemptions were taken from troves (accounts) holding a collateral ratio above 150% (below 66.6% LTV). For example, one trove had a collateral ratio of 189% when it forfeited 130 ETH on May 20th, 2021.”


Redemptions would have incurred a net-zero loss, as a user’s USDH debt would have decreased concurrently with the collateral they forfeited in the redemption, but the user’s exposure to assets they deposited would have been reduced without their express consent.

This model does not align with the UX-focus of Hubble. Instead, the approach will be to implement the PSM much like MakerDAO did with DAI.

DAI and the PSM

Massive thanks to @belami for researching the stats, figures and sources for this section.

MakerDAO introduced the PSM on December 24, 2020. It had an incredibly beneficial impact on DAI peg stability:

It also rapidly increased the demand for DAI. Before the end of 2020, the DAI supply was at ~1 billion with a monthly trading volume of ~4 billion. After the PSM implementation, DAI supply increased by 150% to ~2.5B within three months, with a monthly average trading volume of ~9 billion.

To say it another way, after the PSM was introduced, there was a 150% increase in the total DAI supply since the beginning of MakerDAO in just one fiscal quarter.

DAI Supply

The monthly DAI volume tripled in one month after the PSM implementation, from ~4B in December 2020 to ~12B in January 2021:

The initial implementation of Maker’s PSM included a 0.1% fee for swapping USDC to DAI and vice versa. However, once this fee was removed at the end of 2021, the PSM volume increased by a sizeable margin, since the PSM’s fees now competed lower than Curve:

Why We Want to Implement the PSM

The primary reason for implementing the PSM will be to strengthen the USDH peg.

The PSM will provide a 1:1 exchange between USDH and UST. This means that:

If USDH is above peg:

  • You can deposit UST in the PSM
  • Mint USDH 1:1
  • Swap USDH on the market for gain

This arbitrage establishes a clear price floor for USDH, and it builds up a reserve of UST on Hubble. A UST reserve will be useful because:

If USDH is below peg:

  • USDH can be burned in return for UST, raising USDH’s peg
  • This only works if a UST reserve has been established.

The PSM allows for arbitrage opportunities with zero slippage. The implication of this has the potential to extend beyond simply being an arbitrage mechanism; it can become a zero-risk, decentralized stablecoin swap for Solana DeFi at large.

Proposed PSM Approach

USDH - UST

The PSM will initially be implemented with UST as the token paired with USDH.

NB: There will be room to onboard other stablecoins in additional PSMs.

Zero Fees

There will be zero fees for PSM swaps. From the DAI charts, it is evident that the fee removal had a positive impact on PSM usage. Furthermore, having zero swap fees aligns with our goal of being a capital efficient protocol.

Independent Vault

The PSM will be its own vault, It will have no effect on the other vaults and LTVs of the platform. The swap between USDH and UST needs to be as simple as possible.

If we are to onboard other stablecoins in a PSM model, each will have its own PSM wherein they are paired with USDH.

Philosophical question

Why should you mint one stable with another stable? Arbitrage. The main offering of USDH is leverage or a predictable borrowing experience to users, meaning they can take out some USDH and easily predict how much they will pay back in a year’s time, for example. USDH is being minted out of collateral and is being accepted more and more in Solana, but a high enough demand of USDH could put it above peg which could impact user’s experience when wanting to repay their loan. To do that, we allow 1:1 minting such that above-peg arbitrage is possible.

Risks

The primary risk of a PSM is that USDH is backed by the another stablecoin. In this case, USDH is backed by UST. There is a fair amount of debate surrounding UST and the risks surrounding it, but UST may be the best option if USDH is to remain fully decentralized. However, this is not set in stone, and the contributors will continuously evaluating our options and weighing the risks of each.

Another factor is what would happen if the UST in PSM is depleted, and USDH drops below peg. The solution to this will be Stability Fees, which is beyond the scope of this discussion and will be discussed as its own topic in the near future.

Timeline

The PSM will be implemented within the next two weeks, and it will be essential that the team and the community work together to drive adoption and utilization.

Addendum

Note that this proposal was written before the recent death spiral of UST and LUNA. In light of this, a decision has been made to opt for USDC as the primary PSM token alongside USDH.
Despite the centralization risk of accounts being frozen, USDC has proven itself a reliable stablecoin. In addition, USDC is deeply embedded in Solana DeFi, and thus its integration into the PSM will make a strong case for functionality and adoption.

4 Likes

fabulous! looking forward to the PSM

Adding UST as a collateral would be mutually beneficial to both ecosystems.

We could mint a portion of UST into USDH when used as collateral and swap it back to UST (using the PSM) when the user withdraws.

Has anyone sought out a grant?
Can we host an AMA between Terra and Hubble to help the community understand the technology, and mutual goals?

1 Like

This all sounds great, I didn’t realize the PSM would utilize specific stables and I agree that UST is the best choice. UXD may be another to explore in the future but I am not terribly familiar with it.

I don’t have much to add here, the proof is in the results with DAI and I think it’s the right move. I am most interested to learn more about the Stability Fee, so looking forward to that discussion when the time comes.

What are our plans to build up the initial UST reserves?

the UST reserves will be built as people are arbitraging USDH being above peg. Right now you can sell USDH at 0.5% premium so you can just deposit 1000 UST, get 1000 USDH and sell it and get 1005 UST. then keep looping until USDH loses the premium.

So the reserve will keep building as long as there is a premium on the peg, and it will be drained when there’s a discount on the peg.

If the discount continues & ust reserve is drained then we raise stability fees to drive demand back to usdh.

How can USDH ever become the #1 stablecoin if it has to rely on another stablecoin to keep its peg?

Why not just reduce the minting fee to 0.1% to keep it closer pegged to $1?

1 Like

I thought this would be the case, I just checked the fees on Crema also which it says 0.01% so it all checks out. Thank for clarifying @y2kappa, do you think that liquidity in the pools between UST-USDH could be a barrier at some point?

Part of the plan regarding UST-USDH LPs is to set up a Saber pool between the two tokens that has dual incentives. The plan with that is to negate the risk of insufficient liquidity between the two tokens. However, even if there is not enough USDH-UST direct swaps, CLMMs like Crema’s internal swap (Lite), will facilitate low-slippage swaps, which can then be routed on-platform from USDH-USDC to USDC-UST.

But, as I said, the ideal scenaria, and the one we will push to accomplish, is to have a direct USDH-UST at all times.

1 Like

Not many reasons to believe that the reason the USDH supply is due to high initial minting fee, however the contrubutors will discuss this, as well as opening multiple types of vaults to see which the market prefers.

Also, UST will only be used to create arbitrage opportunities or allow people to mint without taking directional risk.

1 Like

The answer to that is that every stablecoin has to rely on another asset. Even USDC and USDT have to have a reserve backing.

Using a decentralized asset that is closely pegged to USD seems like a good way to secure the peg long term.

In my opinion USDT could also be added to backing assets and work in the same way as USDC