DeFi Trader - Beanstalk, the credit-based stablecoin with high APY
Primer: Join David, the host of DeFi Trader, as he interviewed Publius from Beanstalk, which is a decentralized credit-based stablecoin protocol. Publius talked about what Beanstalk is all about and how BEAN, the stablecoin of Beanstalk, is different from other existing stablecoins.
Who is Publius
Publius is a pseudonym referring to the founders of Beanstalk
At this point in time, Publius is just one part of Beanstalk Farms, which is the decentralized organisation working to develop Beanstalk
How Beanstalk is started
Group of friends who knew each other in college
Around Thanksgiving of 2020, one of them complained that they were sick of their traditional finance job
They already had extensive experience in crypto and they were at a point where they knew they wanted to spend more time in it
At that time, Empty Set Dollar (ESD) was blowing up
They already had the conviction that the biggest problem facing DeFi was the shortage of stablecoins
So ESD happens to be a solution to that stablecoin shortage problem
All of them read the white paper and discussed the pros and cons of it
Wanted to work on a fork of ESD over the next 2 to 3 months to improve it over time
Quickly realised that they are going to have to work from scratch and build an entirely new codebase
A 2-3 months project ended up being closer to 9 months
They had also quit their TradFi jobs and focused on this project
Launched it without presale or advertising or VC rounds
Why is ESD interesting?
The current structure of the cryptocurrency landscape is somewhat similar to the following:
There are layer 1 protocol native assets like Ethereum and Bitcoin that are very volatile
This makes using them for sophisticated financial transactions sub-optimal
Stablecoins tend to fill this gap and they are mostly pegged to USD, hence there is an excessive amount of demand for USD on the blockchain
ESD is interesting because it is one of the few attempts at a novel approach to stablecoin protocols that do not use collaterals
Due to the need to have collaterals in order to issue dollars on the blockchain, there is an opportunity cost associated with locking up all the money in a bank account
The problem is that because you have a supply shortage, typically that will result in a high price. But because the stability maintenance models are good, the high price manifests itself in high borrowing costs
“And so when we say that Beanstalk is a credit-based stable coin protocol, what that means is Beanstalk is able to create stability of the Bean price - the Beanstalk stable coin - by attracting the lenders to lend Beans to Beanstalk anytime the price is too low.”
- Publius
Despite the obvious inefficiencies in the model for ESD, ESD was still blowing up
This is a signal to them that their thesis about the shortage of stablecoins was something worth pursuing because people are willing to invest in an experimental stablecoin protocol that was not limited by the amount of available collateral
This is why they are interested in the ESD model initially
Problems with the ESD model
The most glaring inefficiency was on the debt side
ESD has a key objective - to maintain peg stability, they have to attract lenders and loans
When structuring a credit-based protocol, the rules in place have to create an efficient market for lending. This is not the case for ESD
The 2 main problems are:
All the bonds for lending to ESD were all fungible so anyone who lent ESD at any point in the past is treated equally
The interest rate one receives for lending money to ESD was a hard-coded function of the debt level of the system at the time
These two rules combined to create a backwards incentive structure.
Anytime you felt like it was a good time to lend money to ESD, you were actually incentivized to wait
This is because if somebody else were to lend to ESD before you, the debt level would go up so the interest rate you receive for lending would also go up
There is also no difference between lending earlier or later because all the bonds were fungible
This created a very inefficient market for ESD debt
How is Beanstalk different from ESD
Pods, the debt of Beanstalk, are first-in-first-out, so they are not fungible. This means that the order in which you lend to Beanstalk matters
Beanstalk lets the market set the interest rate. Every season (1 hour), the weather (interest rate for lending to Beanstalk) changes by up to 3%
If you feel that it is a good time to lend to Beanstalk, you are incentivized to do so immediately because if not, someone else can come before you and thus negatively affect your returns
Waiting a few hours for the interest rate to go up marginally is not going to make a meaningful impact on one’s return as getting in as close to the front of the line as possible
Problems with existing stablecoin protocol
Overcollaterialized
Opportunity costs for putting up collateral e.g. MakerDAO with its DAI
Exposure to other protocols due to the collateral model
If the collateral is USDC, there will be exposure to a centralized operator (Circle) with a kill switch
Other decentralised protocols have their own risk profile that you will be exposed to due to the collateral model
Negative carry costs
$1 of DAI is backed up say $1.20 worth of collateral, but the price of DAI remains at $1, so there will be a need to impose some negative carrying costs in order to keep the price pegged at $1
Composability
Luna exists on its own blockchain so it makes leveraging the composable nature of Ethereum difficult to do
No upside to the growth of the protocol
If one is holding UST, it can be deposited in Anchor and look for a handout from Luna holders that are receiving all the upside from the growth of the UST supply
The tradeoffs in using a credit-based model
Since there is nothing backing the system, there are periods of increased volatility, more so than using a stablecoin with a collateralized model
In exchange for that volatility, the benefit is that the total supply of Beans is not capped by the available collateral
Hence the supply can increase infinitely in line with the demand
Anytime the price of Bean is above $1, the Bean supply can grow to meet all of that excess demand
BEANS - neutral/positive carry
Negative carry
Carrying costs are effectively how much one has to pay or get paid to hold on to an asset
High borrowing costs on collateralized stablecoins are effectively negative carrying costs
The high opportunity costs for the collateral is also negative carrying cost because you can’t use the collateral for other things or receive any interest on them
If it is lent out to a lending protocol, you also have to take on additional risks due to being exposed to the lending protocol
Neutral to Positive carry of Beans
Beans do not have any borrowing costs associated with them
Since Beanstalk will always be able to mint new Beans and meet demands, the long term borrowing costs of Beans will approach zero
Anyone can get it at $1 if they are willing to wait a little
No holding costs
Hence Bean has a neutral carry
Depositing Beans in the Silo (the Beanstalk bank) entitles one to get Stalk (the governance token of Beanstalk)
Owning Stalk entities one to a pro-rata portion of Bean seigniorage
Every time the price of Bean goes above $1, the supply increases, anyone who holds Stalk will receive a portion of the new beans minted, and hence participate in the growth of the system
There is actually positive carry associated with Beans by depositing in the Silo
Soil
Soil is the willingness of Beanstalk to borrow Beans
Anytime the price of Bean goes below $1, Beanstalk is willing to borrow Beans because there are too many Beans on the market
Beanstalk will then issue Soil to remove the Bean supply so that the price of Bean can go up to return to the peg
Every season (1 hour), Beanstalk changes the weather (interest rate) based on the price, the debt level and the change in demand for Soil over the last 2 seasons
This is done autonomously by the system
Pods
Pods (the debts of Beanstalk) do not have a fixed maturity. Instead, they have a place in a line
Anytime the supply of Beans increases, half of the newly minted Beans go to Stalk holders and the other half goes to the Pod holders on a First-In-First-Out basis
So while there is no fixed maturity for the debt, there is a fixed interest rate that is locked the moment you lend Beans to Beanstalk
You just need to wait for your turn in the line so that the Pods can be harvested to become Beans
Recently there is the launch of the Farmers Market, a decentralized exchange for Pods, where one can trade their Pods with Beans
This will provide some liquidity for those who want to redeem their Pods for Beans without waiting for their turn to come in the line
Why would the price of Bean go above $1?
Anyone can buy Beans and deposit them in the Silo to earn interest, thus it has a positive carry
It may make sense for some people to pay a slight premium over $1 to start earning interest sooner rather than to wait for Bean to fall in price
This may create demand for Beans due to the positive carry depositing the Beans in Silo
Why are people willing to lend Beans to Beanstalk?
Beanstalk had lived through fire and flames in the past and people feel that it has a track record
When Beanstalk was launched in early August with no presale, the protocol was growing slowly and steadily up to 2.5 million supply in the first month
Then it went viral on Crypto Twitter and people thought Bean is the same as ESD
The name of the game with ESD was getting in as soon as possible, so everyone must have thought this is the same dynamic at play
Everyone was buying ESD regardless of the price even though it was only worth $1
If done early enough, this is a lucrative play
But Beanstalk is not like ESD and the system is designed to discourage buying Beans when they go higher above $1
In the next few days, 25 million dollars came into Beanstalk when it is only 2.5 million in supply
When all the apes realised they couldn’t exploit the protocol, they all left. The price went from as high as $4 to as low as 24 cents
Over the next month or so, Beanstalk was able to attract millions of dollars of loans from hundreds of different wallets and created enough demand for Beans to return to the peg price of $1
After about a month, the price went back to peg, and this is an opportunity to prove that Beanstalk’s credit-based system works
In late November, another huge wave of demand came when someone put in a million dollars in a couple of hours, which kick-started another growth cycle
Almost everyone who lent to Beanstalk in the first debt cycle all made their money back and got a very nice profit
Inefficiencies in system
There are still some inefficiencies in the model
The soil parameter is set inefficiently so Beanstalk exited the growth cycle in a worse position than when it started
This led to Beanstalk issuing way too much debt during the last cycle
However, Beanstalk Farms had proposed Beanstalk Improvement Proposals (BIP) to make tweaks to the model to fix such issues
“The goal is not to have ownership trend towards a few hands or early investors or anything like that. The goal is to have a truly decentralised system that increases in decentralisation and decreases in concentration over time.”
- Publius
Why launch Beanstalk with no pre-sales?
Beanstalk is solving an essential problem that DeFi needs to solve
If Beanstalk is to become a primitive across Ethereum native DeFi in the long run, everything has to be done the right way to ensure the highest chance of success
Do not think that the whole market is going to adopt something if it is launched specifically to enrich a select few due to pre-sales
Hence the decision to launch without pre-sales gives Beanstalk the best chance of being a DeFi primitive
Incentives to minimise bank runs on Beanstalk
Going back to ESD, anytime the price goes below $1, there was no seigniorage paid to anyone, so there was an incentive to withdraw assets out, which causes the price to go down further
This created a huge negative feedback loop in ESD, causing a bank run
There are a few incentives structures to prevent a bank run for Beanstalk:
Seeds and Stalks are burnt upon withdrawal
When assets are deposited in the Silo, stalks and seeds are received every season
Seeds yield more stalks, so the longer the duration of the deposit, the higher the yield
Upon withdrawal, all the stalks and seeds are burnt. This creates an opportunity cost to withdraw and subsequently redeposit in the Silo again
The longer the asset is deposited in the Silo, the higher the opportunity costs become
Everyone has the same incentive structure, independent from the decision of all other people, to stay invested in the Silo as long as possible
Other existential risks of Beanstalk
Based on the current liquidity structure, they have a Bean:3CRV pool and a Uniswap Bean:ETH pool that takes up more than 2/3 of the liquidity
If the price of Ethereum goes to $100, that will be very bad for Beanstalk because it will be hard for a young system like Beanstalk to continue to attract demand
The well being of Beanstalk depends on the rest of the market, particularly with Ethereum
ETH at $2100 or even $1800, Beanstalk will still be fine
Roadmap
Generalizing the Silo so that assets that are yield generating from other protocols can earn interest from Beanstalk too
A proposal to include the Bean:3CRV pool in the gauge system so holders can earn both Curve and Beanstalk rewards
The UI will also be super clean and can do a lot of things in one transaction all within the interface seamlessly
All information presented above is for educational purposes only and should not be taken as investment advice. Summaries are prepared by The Reading Ape. While reasonable efforts are made to provide accurate content, any errors in interpreting and summarizing the source material are ours alone. We disclaim any liability associated with the use of our content.