A better Alchemix on Terra that uses PoS staking yields

Hi all,

Just a quick thought about Alchemix and how we might be able to do it better on Terra.

Right now afaik, Alchemix lets you take out a self-repaying loan using DAI as collateral to borrow alUSD at <50% LTV. Behind the scenes, Alchemix deposits the DAI in Yearn vaults to earn yields, which are then used to pay back your loan.

Personally I have a hard time seeing why you would want to collateralize your stablecoins to borrow less stablecoins. But putting that aside, I think it would be interesting if we could do the same thing with bLUNA and UST.

In essence, I am proposing a protocol for self-repaying UST loans collateralized by bLUNA (or other supported PoS staking assets down the line), where the yield generated by the bLUNA would be used to pay down the loan. Basically this would be like Anchor except that the yields from bAssets are used to repay loans instead of being diverted to UST depositors.

Any thoughts?

I have a hard time understanding why Alchemix is useful. As for the bLuna collat to pay back stablecoin loans, the resultant stablecoin would need to be its own stablecoin, not UST to achieve capital efficiency. Just want to make sure that’s been considered.

I think Alchemix it’s a very good idea. The example often given is if you want to buy a car, you can put your stable on Alchemix take a loan, and you will have a “free” car because the have been payed upfrond and the loan will be repayed itself.

1 Like

Is Alchemix an undercollateralized loan?

Hi Do, basically Alchemix allows you to deposit DAI and then take out 50% of that in alUSD which you can actually convert back to USDC or USDT and use for whatever purpose you want. Then Alchemix puts your DAI to work to earn yields and it pays back the loan for you - they say for around 18-24 months your loan should fully pay itself. I think this is useful if you have a more immediate need for an item as opposed to the current setup of Anchor Protocol which is to provide 20% Savings APY. In Alchemix case, you want to purchase a car for $10,000, so you put in $20,000 into Alchemix, and it allows you to borrow up to $10,000 of your deposit and then they do the work for you so your loan pays by itself eventually. I think the use case here is if you already want to spend the 20% in future interest as opposed to waiting for it. Not sure how this changes Anchor’s positioning if something like this is built on it.

@TerranJay I think your example exactly illustrates why Alchemix doesn’t make too much sense to me.

If you already have 20k in stablecoins, why would you not just take 10k out and buy your car with it as opposed to locking up all 20k for a 10k loan?

This is why I think using a yield-bearing PoS asset like bLUNA as collateral would make more sense for the idea of a self-servicing loan (e.g. if you’re bullish on LUNA long term but need cash immediately).

@dokwon can you pls explain your point about capital efficiency?

If you have 20k in stablecoins and you need 10k for a car, wouldn’t you prefer having 20k in a yield farm instead of only 10k? Your car will be paid for faster with 20k deposited.

Seigniorage of each alUSD = $1 if the peg holds up. It’s basically DAI with less liquidity and more tail risk. The expectation of its 1:1 redeemability should theoretically be enough to support a range close to the peg. It’s sort of a callable zero coupon bond where alUSD is the residue.

DeFi bond/options markets are inevitable. It would certainly be interesting to integrate PoS yields somehow. Addition of bUST (stripped UST bond) to Anchor would essentially enable the same functionality as Alchemix, afaict.

The “self-repaying loans” narrative is quite compelling imo. Traditional savings accounts just aren’t that sexy, no matter the yield. Self-repaying loans are more goal oriented. If i can get a car “for free”, and the time frame and capital is easy to comprehend, it sounds like a good deal.

The car example isn’t great, since if I need X to buy something, and I have 2X in a liquid asset, other than tax reasons or an interest rate arb (say I get paid 8% vs paying the auto lender 4%) its not clear why I would borrow against that liquid asset instead of just selling it to fund the purchase.

This does raise the point that a stablecoin that natively earns yield would be very compelling

If you have a liquid savings account, you don’t need to sell any of it to buy the car. You can “take out a free loan” (issue a zero coupon bond), and continue earning yield on every penny in your savings account.

Alternatively, you can borrow against your liquid fixed income asset and reinvest it to earn a higher yield on your savings. It’s not unusual for fixed income hedge funds to have 20x leverage on their bond portfolios.

Thanks for explaining, it makes sense to me now. I do wonder if most people would opt for this option, which as you said would pay for the car faster, as opposed to keeping 10k liquid by just buying the car without a loan, especially if it’s as @TerranJay said and it takes 18-24 months to repay the loan. I can definitely see a use case for this but I wonder how much demand there would be.

I’m guessing most of the initial demand would come from yield farmers. alUSD is currently trading at ~0.9 DAI, indicating ~1.8x “free” leverage possible with this construction. 20% yield would become 36%.

Using a more capital efficient construction, perhaps something like Liquity (LUSD), 5x leverage shouldn’t be a problem. This is roughly what Alpaca Finance currently is offering, so it seems feasible.

I guess it’s really about whether you’d rather have the APY up front at the cost of locking your capital, right?

For example, at the time of writing, the crypto market just crashed, offering a timely opportunity to do some degen long leverage play on Bitcoin. I could use $20,000 of my own money, but if my risky play fails, I lose my actual money.

But say I had $100,000 UST I didn’t need to touch for a year. What if I could lock it in Anchor for a year and get the 20% yield up front? I get a “free” $20,000 UST right now on the back of the crash I can do degen shit with and if I lose it, that’s no hair off my chest - I just have to wait a year to get access back to my $100,000 when the APY pays off the loan. And if it works, I can payoff the rest of the loan to access my funds.

This seems to make sense to me from a use case point of view, although I don’t know if it’s feasible from the backend perspective of the protocol - you might have to fudge some numbers or something.

1 Like

Just reread and I would be very happy for this to come to Terra Ecosystem. At times like this I wish I could program…

2 Likes

Hi folks I am very new to Terra (been mostly hanging out in Ethereum and Polkadot projects) but I have deep dived into Anchor and Terra recently and have been impressed.

Just wanted to say I have been involved with the Alchemix project from the fair launch, have used it extensively (self repaying loans) and have found the Alchemix community and developers very switched on. They’re good developers with a sophisticated understanding of financial services and DeFi.

Given Terra and Anchor have are fine tuned towards offering a consistent 20% savings rate, the opportunity to offer a similar vehicle to me is a bit of a no brainer.

One of the Alchemix community members 0xdef1 did a great explainer of how Alchemix works fundamentally here: dyor.fi

I’m by no means an expert but I am a user and can say it does work as intended and is an absolute innovation in composability on DeFi built on Curve and Yearn.

The key to understanding how it works is by the Vaults mandating all “advance yield loans” are denominated in the same unit of value.

alUSD the synthetic USD

DAI the algorithmic stablecoin pegged to USD

The ability to swap alUSD via the $500m Curve pool alUSD-USDT-USDC-DAI

And the yield earning instrument via Yearn DAI V2 Vault also being USD denominated.

All of this means there is zero risk of liquidation of advance yield loans as long as the peg holds and obviously smart contracts are sound.

The only “cost” is opportunity cost (having to overcollateralise your position) and time. The higher the yield rate from Yearn the faster it pays down the global debt. The lower the rate the longer it takes.

Given Anchor has a fixed 20% rate it makes it very attractive to use a version of Alchemix on Terra in my opinion.

Happy to respond to any questions or point interested developers in the direction of the Alchemix core devs.

5 Likes

relatively new to terra, but signed up for an account here to chime in - i’m definitely interested in an alchemix fork on terra, because it’d be a great lego in the overall eco system.

The protocol has already been explained a few times, but here’s another angle - when you have a yield-bearing asset, e.g. aUST, it can be assumed you’ll get X UST of interest over the next year. Through alchemix, you can essentially get an advance on that interest. Assuming X UST is less than 50% of your deposited collateral value, you can take out that loan, which will be reduced over time as your collateral gains interest.

For example, if I deposit $1k in anchor, after one year that would be worth ~$1.2k. I can then deposit the aUST in alchemix, pull out $100 up front, and after that year in anchor, my deposit is worth $1100 and my alchemix loan is paid off. Either way, at the EOY i have $1200, alchemix just allows me to pull out interest at the beginning of the year instead of waiting until the end.

I can see this being useful in the following situation: Imagine an upgrade to the anchor protocol that automatically covered your deposit through ozone. One way to implement this would be to spend some of the original deposit on purchasing the insurance premium. Another method could be to deposit all of the UST in anchor, and take a loan (advance on the future interest) through alchemix to pay for the insurance premium. I would guess that with the second method, you’d end up slightly ahead at the end of the year.

Thoughts?

1 Like

Greetings,

This is being released on the Terra ecosystem January 17th with Obverse.

Currently available in testnet: obverse.money

For those of you wondering why this technology is useful, read the Abstract in the whitepaper.

White Paper:

@brew @Ilo_Muoto

I’m glad to see you and others here are in support of this technology coming to Terra. Im one of the developers behind Obverse so if you review the white paper and have any questions or ideas for improvement, please let me know.

Zenith

1 Like

Kinetic.money is launching soon on Terra fam.

We’re actually building something similar to Alchemix. Will be posting soon about it

1 Like