Hubble Uranus Vaults

Proposal: Build high-yield Sol, ETH and BTC vaults by automating already-existing and soon-to-be existing features of the Hubble infrastructure into a single product.

Overview: The Hubble borrowing apparatus provides the tools necessary to build unique yield opportunities that have not previously been available on the Solana Network. However many of these opportunities may not be readily apparent to an average investor, who might instead see Hubble as a standard lending protocol akin to Solend or Mango, albeit with fewer borrowing options. Making the value proposition of Hubble apparent and tearing down barriers to entry are critical to the long term success of the protocol.

The existing and soon-to-be existing infrastructure of Hubble can be sequenced in a variety of ways to produce high yields with relatively low risk to the investor. By packaging these strategies into a single product and automating the underlying process, Hubble can establish itself as both a leader in innovation within the Solana DeFi ecosystem and as a high-yield investment platform.

Details: An investor deposits collateral (Sol/BTC/ETH) into the vault (possibly called the Uranus Vault becauseā€¦.space humor.) The vault deposits these tokens into a yield-bearing strategy, as described in the ā€œcollateral yieldā€ portion of the Hubble Gitbook, and takes out a 50% LTV USDH loan against the collateral. The USDH is deposited into the Stability Pool, where it earns HBB rewards.

At this point, there are two paths that could be taken. In the Long Orbit Vault, the HBB rewards are staked to earn USDH rewards. The USDH rewards are then used to purchase the L1 token initially deposited and added to the investorā€™s collateral. In the Short Orbit Vault, the HBB rewards are automatically converted to the L1 token initially deposited, bypassing the HBB staking portion of the cycle, and added to the investorā€™s collateral.

By continuously adding to the investorā€™s collateral, the LTV of the loan continuously decreases. To further protect against the risk of liquidation, the vault automatically deleverages positions as they approach 75% LTV by taking the USDH deposited into the stability pool and reducing the open loan to 50% LTV. The end result is a vault that produces high quality and high quantity returns at relatively low risk.

Additional Considerations: Because this strategy is both novel and high yield, it would be justified for the protocol to charge a premium for access and/or performance beyond the loan origination fee. In the case of the Long Orbit Vault, the vault adds funds to the stability pool and automatically stakes HBB, which reduces sell pressure on the HBB token. Because both of these actions are directly beneficial to the protocol, fees should generally be lower in this vault. The loan origination fee, gas fees, and a low premium would be adequate. In the Short Orbit Vault, the vault is again adding funds to the stability pool. However, in this case, there is now automated sell pressure against the HBB token. For that reason, fees should generally be higher and a significant portion of the returns should be passed on to staked HBB tokens.

There are several different ways by which this could be accomplished. For example, the vault could convert the HBB tokens to L1s at a 1.5-2x premium over the spot price of HBB with the difference being split between the protocol and staked HBB holders. This would ensure that the protocol is collecting a continuous premium for service and adding value back to the HBB tokens. A more elegant solution would be to charge a dynamic fee that takes into account the total HBB rewards rate and ratio of staked tokens to Short Orbit tokens. However, the math is this case is less straightforward and beyond the scope of the proposal.

Even with the fee structure in place, thereā€™s still an obvious concern about the effects of automating sell pressure against the HBB token. If we accept that a certain amount of sell pressure will always exist irrespective of automation, by isolating much of that pressure into a single vault product, it gives the protocol loose control and allows for better revenue capture on HBB sales. In addition, if the vault products are successful, it should result in a significant uptick in loan origination fees, which again means more yield for (and therefore more value in) staked HBB.

7 Likes

this is very cool. I think this idea excites me personally as a defi user, but also executing this successfully creates one helluva USP for Hubble which we all want. The Long Orbit - Short Orbit Vaults sound really cool, the idea that you automatically receive more of your L1 deposit AND your debt is automatically getting paid off with your in-house USDH rewards. I made some notes/thoughts I had below ( though they probably sound clearer in my head)

  • One question i have is , what is the incentive to use the long orbit vault compared to short orbit? seems like short orbit would be a go-to as it simplifies the process, then again maybe long orbit gives more flexibility with your USDH-HBB rewards? Just trying to understand this a little better

what would the offered APY/selling prop be here to outside investors/defi users? high yield does sound attractive just curious if thereā€™s a ballpark on what users could expect to earn here

The dynamic fee sounds intriguing but I agree, think itā€™d be way harder to explain/sell to people. Maybe an option to allow users to choose between the fees? Maybe Iā€™m digging too deep here as Iā€™m not even sure if this would be possible logistically

Really cool proposal here, think we can do something special here with Hubble especially once yield strategies come into play. Would love to see this play out and what Marius and our devs might think

Also, Uranus vault. Classic :pinched_fingers:

1 Like

A very interesting idea with obvious appeal as an investor, but I think there are some challenges to address.

Would this ā€œvaultā€ exist on the current Vaults page? This strategy is essentially automating most of the functionality on Hubble. It wouldnā€™t really be a vault so much as a bot that utilizes vaults in its loop. If it exists on the Vaults page, there would need to be a way for the user to point it towards an existing loan to run the strategy. Itā€™s critical that users can still manually adjust their loans during periods of market volatility. 50% LTV is not terribly secure depending on the assets in the loan. Will the user be able to change which loan the ā€œvaultā€ points to, in this scenario? How will the newly-purchased collateral be chosen for multi-asset loans, all towards one token or equally split? Not impossible questions to answer, just things to consider.

I also strongly dislike even the option of the Short Orbit vault. I do not think Hubble should encourage HBB dumping under any circumstance. People are going to do it, sure, but I dislike the idea of the protocol automating something that actively harms its investors. Emissions dumping is behavior that all DeFi protocols struggle with and great effort is spent trying to mitigate the damage. You can see this with Solend now experimenting with SLND call options as rewards rather than just native token emissions. I know Marinade has floated similar ideas for MNDE. With the Short Orbit vault, Hubble would be automating activity that most protocols are trying to discourage. There are some investors who will never get behind a token that automatically dumps itself, regardless of the fees generated. I really donā€™t think we should go out of our way to discourage HBB adoption.

Overall the idea is certainly interesting, but I wonder if a simpler HBB Vault/USDH Vault loop might be a more effective use of dev resources in the short term. This vault seems like it could require a lot of dev work for something that doesnā€™t necessarily result in a better yield for the user. Automatically buying new collateral at the current spot price means that you may may be leaving money on the table by not manually waiting for a better entry. The users who perform this strategy manually have the potential to generate better returns, and thatā€™s without taking the new fees into consideration. I do understand that some users prefer to DCA regardless of market conditions, though. This strategy would likely be very appealing to them.

1 Like

thanks for your proposal, appreciate that everyone wants to help improve hubble and make it one of the best projects on solana!

after briefly reading through, i think what your trying to say is very similar to Neptune Protocol on Terra, so i would appreciate if the devs could also check that out so they can get an idea of how it works and maybe if we could implement the similarity into Hubble

https://articles.nept.finance/neptunes-leveraged-yield-farming-how-it-works-4c476d626349

Thanks

1 Like

The Short Orbit would accumulate exclusively L1 tokens through the yield-bearing collateral and the sale of HBB tokens. Due to the selling pressure generated against the HBB token, this vault would carry higher fees to pass value through to staked HBB holders. The Long Orbit is effectively an automation of the Hubble mechanics already in place. This vault would accumulate L1 tokens via the yield-bearing collateral and HBB tokens via the stability pool. Over time, the staked HBB tokens would generate USDH yield, which would be used to generate additional L1 tokens. Fees would be lower, as the mechanics to execute this vault strategy already exist and the loop stakes HBB rather than selling it. So the TL/DR is that the Short Orbit generates exclusively L1s but carries higher fees. The Long Orbit generates fewer L1s, a lot of HBB, and carries lower fees.

Itā€™s hard to pin concrete numbers to the phrase ā€œhigh yieldā€ in this context due to the unpredictable nature of the crypto market. However, if we hold current prices and rewards rates constant, we can formulate some rough numbers, which almost certainly wouldnā€™t hold up but nonetheless are useful to understand the potential value of the strategy.

For the Short Orbit Vault, the vault would generate an 18% return on a 50% LTV loan (so a 9% return on the initial deposit.) If we attach a 33% fee to that return, weā€™re left with a 6% return in L1 tokens. This 6% would be added to the yield generated by whatever yield-bearing collateral strategy the initial deposit was placed into (maybe 1% for BTC, 3-4% for ETH and 7-8% for Sol.) The final result is a low risk vault generating 7% APY on BTC, 9-10% APY on ETH or 13-14% on Sol.

The Long Orbit Vault would generate an 18% return in HBB on a 50% LTV loan. The HBB would then be staked to generate an additional 50% return in USDH as it accumulates. The math here becomes a little bit tricky but for the sake of simplicity, weā€™ll say that itā€™s a 25% return on that 18% backloaded into the latter half of the year. The vault would generate 9% in HBB and 2.25% in L1s plus the yield generated by the yield-bearing collateral. The final results would be 3% BTC, 5-6% ETH or 9-10% Sol plus 9% HBB.

1 Like

First to address the elephant in the roomā€¦ the name is amazing

Reading through the proposal and the replies it seems like what we are trying to achieve is an automated strategy for our Hubble activity that increases the overall yield that we receive from our activities, which iā€™m a fan of. However, I think this falls onto a spectrum of what the user is expecting and how ā€˜degenā€™ they are for lack of a better word.

I feel that if we make the strategies overly complex or rely solely on dumping HBB emissions to generate yield, we will attract a certain type of user who is defi savy and it looking for the highest yield. These people tend not to be the sticky capital we are looking for as they move from opportunity to opportunity looking for the next juicy lemon to squeeze dry (I know this is not always the case). For me this is not really what we want for the long term viability of Hubble as we have to keep up with the latest strategy or shiny new thing to draw people back in.

To this end I think something simple such as a funds entered into a Stability Vault (formerly stability pool) which produces HBB rewards which are then auto staked to earn USDH with those rewards users can have three options either 1. repay their debt 2. stake back into the stability pool to build their position further thus compounding rewards over time 3. acquire L1s at the market price. A simple system like should (iā€™m not a dev as a warning) be fairly easy to implement and very easy for users to understand and plan with effectively.

The plus points of this system that all value created by the Hubble protocol is accrued to the Hubble protocol / users and retained by the gradual acquiring and compounding of HBB and USDH or repaying debt to free up collateral (which if people get comfortable with could encourage further lending). When you consider the type of user that would use this system they would tend to be long term investors looking to optimise their holdings and not have to think too much about it making them likely to stick around.

As @BlockchainFixesThis mentioned the acquiring of L1s is advantageous, the stability pool provides an automated function for this in terms of liquidation rewards which you make either a 9% gain if you sell immediately or get a discount to market price on desirable tokens via automated DCA, perhaps we can even have these rewards auto compounded to the users collateral position lowering their LTVs in times of volatility. Although these rewards are infrequent as they rely on market conditions as the user base of Hubble grows then liquidations should increase in tandem as not everyone is a responsible borrower, this also could be amplified by the lowering of LTV to 75%. Taking this into consideration liquidation rewards could start to play a more significant factor in peoples assessment of their investment returns and usefulness of Hubble.

This all in theory would make Hubble a great place to park your cash and earn some extra yield while you wait for the moon. If we build a reputation for Hubble as a place that people never want to leave (or have to) then word should spread and entice more people to come join in increasing user base and protocol revenue which is ultimately what will benefit every long term Hubble user the most.

1 Like

What youā€™ve laid out is entirely reasonable. My entire motivation behind writing the proposal and subsequently advocating for it is to automate the strategies that myself and others are already employing (whether that be the Short Orbit, Long Orbit or one of the various alterations described.) In the process, weā€™d truly be making our Hubble investments a one or two click, set-it-and-forget-it investment. That is my ultimate goal. Wellā€¦that and being the man who brought something called ā€œthe Uranus Vaultā€ into existence. You could carve that onto my tombstone someday.

I love strategic leverage as much as anyone but, almost by definition, those positions shouldnā€™t be kept open indefinitely. They need to be closely monitored and adjusted as market conditions dictate. Theyā€™re funā€¦but theyā€™re also time consuming and stressful. Theyā€™re not for everyone, not even for everyone far enough out on the risk curve to be investing in DeFi on a beta chain. If we can automate the process, saving time and mitigating risk, we can make Hubble more accessible and less intimidating for the average user. That feels like a path to sustainable, long-term growth.

2 Likes

Apart from automating the HBB staking, USDH harvest & staking, I am not sure these are long term high yield strategies. Firstly, itā€™s because we hope HBB rewards turn to 0 at some point (if we can achieve market based liquidations - btw this doesnā€™t mean the SP will turn to 0 rewards, but HBB only).

However you are on the right track about what we want to launch next using Hubble as a primitive. The simplest vault is the following: deposit BTC, mint USDH at 40% LTV, and allocate it in the highest APY LP from Saber and compound those gains. This strategy will look like a high btc yield strategy. If you get a 20% APY on saber then you get 20% on 40% ā†’ 8% yield on BTC (which you canā€™t find anywhere else), meaning we can attract quite a few people.

There are several optimizations to be made here which Iā€™m very excited about. Because the user will get an attractive, automated strategy, we will use maybe 20% of the minted value to provide USDH-USD* LP to ensure that leveraging doesnā€™t break the peg and that it keeps USDH tightly pegged, and in exchange for serving the greater good, we provide the automation for free almost.

2 Likes

Also, I would be very keen on hearing some more names for these vaults. Work is about to start on these soon.

Thanks for the reply! This was the feedback I was ultimately looking for. Our vaults are structured differently but weā€™re driving at the same point. Iā€™d be honored if your new product took on the name ā€œThe Uranus Vault.ā€

I agree here, I find that we should first focus on the vault loop, thatā€™s a major use case for Hubble. New defi users want LESS clicks when doing yield strategies, an automated vault solves this issue and makes the protocol extremely easy to use and scalable.

We want to grow the value of both hubble tokens and the platform, this grows the platform at the expense of degrading the value of the token, solutions need both otherwise in the long run we create a great platform for short term use, that isnā€™t reflected in token price in the long term. Looking forward, I even look at the price action of the token to determine the love of the platform, its easy to see that many defi projects donā€™t get the love for the native token compared to the use on the platform, which creates disconnections between users and actual utility. I believe this questions the real long-long term utility and strategy of the project.

To take this idea step further I think it could be interesting to explore collateral native yields where by the deposited collateral is supplied to lending protocols.

Take mSOL and Apricot as a case study we currently have 84,654.97 mSOL held as collateral worth $9,500,954.50, if this collateral was then deposited into Apricot it would earn the below rewards on an annual basis.

Deposit APR: 0.25% = $23,752.38 = 211.05 mSOL
APT APR: 0.75% = $71,257.15 = 2,225,497.93 APT
MNDE APR: 0.87% = $82,658.30 = 321,216.41 MNDE

These rewards are then distributed to the users based on their percentage of the mSOL held in collateral, minus a 10% facilitation fee ($17,766.28 extra revenue) which is either added to the Hubble treasury or distributed to HBB stakers.

Sorry if my math is off anywhere iā€™ve not had my coffee yet, but I think you can see the point iā€™m trying to make. This would need to be considered by the community to see if having an added layer of risk (protocol risk with Apricot) would be worth the rewards. Also we would have to discuss with them if they would be happy for us to have this much capital held with them. But just an idea, happy to hear your feedback.

Yield-bearing-collateral is going to add a nice new dimension to the protocol. My guess would be that they start with something safe/easy like P2P lending on Solend, Tulip, Apricot, etc. The biggest challenge with this scenario is that as Hubble reaches scale, it will be bringing so many Sol/ETH/BTC tokens onto a lending platform that utilization rates will get rekt, causing rewards rates to plummet. Some of this can be avoided by spreading tokens across multiple platforms or using ones that have leverage farming operations (like Francium) due to their much higher utilization rates.

What would be super cool (but not at all realistic) would be to eventually create a separately run Solana fund where tokens could be deposited. Within this fund, you could have diversified exposure to all Solana yield-bearing apparatuses across the ecosystem. Portions of the funds could be allocated to each of the liquid staked derivatives (Marinade, Lido, Socean, Parrot), lent on P2P lending sites, put into exotic LPs, used in stakedSol/Sol loops, deposited into covered call vaults, or whatever else comes along. If managed correctly, a fund like this could probably generate 10-15% APY (in Sol) consistently. It could either be controlled exclusively by Hubble or spun off into its own entity entirelyā€¦ :thinking:

2 Likes

Great Info as always!

Quick question, i have $RAY lent on Solfarm, so i get tuRAY in my walletā€¦ would it be possible to use that as collateral on Hubble? or would that have problems in Solfarm

Not sure if rehypothecating usersā€™ assets without their direct consent is the best. What has been talked about/planned is allowing individual users to opt in to strategies like this with their own collateral. And reap the rewards themselves.

Wondering now if there would be benefits to having a single Hubble account that handles this type of yield strategy or having an account made for each user. Have to look at this from a security point of view, too!

Thanks for the reference! I might discover a thing or two - will check it out :slightly_smiling_face: