# Maker - Fixed Rate Collateral
Fixed rate primitives as collateral on Maker introduce a strong arbitrage market with AAVE and Compound, bringing in liquidity to Maker with a significant, predictable value capture. Loan conversion lets Maker set the borrow rates on other platforms and internalize a large amount of the interest paid there, allowing Maker to extend it's role as the reserve bank of DeFi.

Currently there are inefficiencies between very similar rates that should be arbitraged closer together, but the market mechanisms to do so don’t exist.

_{Aave’s rates on DAI, source: https://daistats.com/#/collateral}

Onboarding fixed rate collateral into Maker would cause these rates to further converge and push large waves of liquidity between Compound and AAVE via Maker. In particular the USDC collateral type allows value capture on the large reserve of USDC backing DAI and lets Maker play the market defining role not just on DAI borrow but also USDC borrows.

Lets dive in, but first a quick background.

## Background
### How Element Fixed Rates Work
Element enables fixed rates by splitting the a variable cash flow from the capital that creates it. These are represented as ERC20s, one for principal and one for yield. Buying the locked Principal at a discount guarantees yield when the principal can be unlocked:

1 Fixed Rate USDC 3% APR (ptUSDC) = 1 USDC on maturity at 1 year.
1 Fixed Rate USDC 3% APR (ptUSDC) = .97 USDC on day 1

### Element Fixed Rate Protocol Risk

Since Element simply splits variable rates from principal, the fixed rate positions are fully backed at all times and the smart contracts to support the split are very simple. There's no internal lending, liquidations, or other economic risk factors. Holders of fixed rates are exposed to a small Element smart contract risk and the risk of the yield source itself, such as Compound or Aave. For this reason pricing of the risk of the rates is quite similar to Compound or Aave deposit risk.

### Element Fixed Rate Borrowing Adaptor

Obtaining fixed rates by buying collateral at a discount is a format of risk transfer, but who wants to take the risk on end of deal?

In the current Element deployment speculators can earn a very high rate of return speculating on future interest rates of Yearn. They sell discounted collateral to get interest rate leverage.

In the new Compound and Aave yield sources a new buyer of the risk position emerges. Borrowers on Compound and Aave are natural buyers of the future yield flow, they are yield short and combined with yield long they can get an interest rate hedged position. For example

* I borrow 1 USDC against 1 ETH
* I obtain the rights to interest on 1 dollar in Compound for the next year and I pay 0.02 cents based on current market rate
* Interest rates spike and the paid interest is 0.05 for my loan
* My yield claim pays 0.045 [because of supply borrow differential]
* My cost is 0.02+0.05 - 0.045 so I pay 0.025 in total
* 0.02 to the depositor and 0.005 to Compound for liquidation management

In this way the Element protocol enables market powered fixed rate lending on Compound and Aave.

### Fixed Rate Collateral vs Interest Rates

When fixed rate collateral is added to a borrowing system it has the net affect of discounting the borrow rates. For example if a stablecoin collateral is borrowable against for less than its fixed rate earned the effective borrowing interest rate becomes negative:

https://twitter.com/wjvill/status/1527680960866705408
This tweet simulates this process on a mainnet fork with data including slippage, fees, etc.. on Fiat, a Maker analogue which already accepts fixed rates.

Basically, it is a simple leverage system where DAI is minted with fixed rate USDC (ptUSDC) as collateral.

If stability fee is 1%, fixed rate is 3%, the leveraging cycle can continue until the fixed rate reaches 1%, since the leveraging cycle produces downward pressure on the rate.

1. Buy fixed rate USDC, ptUSDC (3% APR)
2. Use ptUSDC to mint DAI at 1% stability fee
3. Repeat

In the example from the thread the negative interest rate is a very strong downward peg pressure, to the point of causing a persistent Fiat stablecoin depeg through Curve, despite being fully collateralized. With DAI, this is not an issue.

## Leveraging into the lending side, Yield Token Compounding (YTC)
Selling your fixed rate USDC, ptUSDC, at an interest rate lower than the variable rate allows the user to effectively leverage into supply side APY for AAVE/Compound.

If fixed rates go for 1% APR but the variable lending rate is at 3% APY. You can use Element to leverage into that 2% differential. This causes sell pressure on ptUSDC, driving the fixed rate up. However, this also drives massive amounts of liquidity into the supply or lending side. This causes the variable borrow rate on AAVE/Compound to decrease, since more liquidity is entering the lending side on leverage than the borrow side. This effect ultimately causes the variable borrow APY to converge on Maker's stability fee.

## Maker Loan Conversion

Combining the ideas of the Compound/Aave fixed rate borrow and negative interest rate arbitrage loop from bonds as stablecoin collateral, we propose that Maker can use Element to convert lending on Compound/Aave to Maker lending.

The Maker currently has billions of USDC collateral for DAI, which earns low interest rates. If the Maker system used fixed rate USDC bonds yielding from Compound/Aave it could earn much more from the USDC collateral while maintaining downward peg pressure. Moreover as the net 'buyer' of bonds Maker would be the counterparty for the fixed borrowing enabled by Element.

Consider a user Alice who wants an Aave loan at a 1 year fixed rate for a Maker unsupported collateral. She takes out a 1 million USDC borrow and buys 1 million of yield tokens from the Aave Element market. This drives fixed one year bonds from 2% to 2.5% because of the principal sales. An arbitrage then buys the 1 million of bonds using high leverage on Maker at a 2% interest rate.

In this scenario, an external liquation market (Aave) handles the liquidation and backstop of Alice's collateral, but Maker net earns 2% while backing the year long fixed rate. Alice's USDC loan is effectively converted to a 1 year DAI loan yielding for Maker at double the rate of current USDC-A vault.

**Example**:
Let's connect everything above:

1. Compound has a borrowing rate of 3% variable APY, and a lending rate of 2.5% APY on USDC
2. Fixed Rate USDC, ptUSDC is at 2.5% APR, based on the fixed rate borrowing adaptor
3. Maker has a stability fee of 1% APY
4. Use maker as leverage to buy Fixed Rate USDC repeatedly
5. Fixed Rates are now at 1% APR due to leveraging cycle and buy pressure through Maker
6. Borrow USDC from compound using Element's fixed rate adaptor at 1% APR. Fixed rate APR on borrowing is less than the current 3% variable APY due to Maker's leverage cycle.
7. Yield Token Compound (YTC) or leverage into the lending side, 2.5% variable vs. 1% fixed on Compound repeatedly using Element.
8. Fixed rate APR increases again, to say 2.5% APR, matching Compound's variable lending rate. However, leveraged liquidity flooded into the lending side, causing Compound's variable lending rate to decrease and ultimately begin to converge on Maker's stability fee.

This cycle continues between extremely liquid markets (AAVE, Compound) and Maker. Maker acts as the intermediary, driving a convergance on the rates.

## Value Capture

The upsides to Maker of onboarding fixed rate collateral:

1. Increase protocol earning from the USDC that backs DAI while preserving downward peg pressure.
2. Long term increase protocol earning from other non-stablecoins, such as fixed rate ETH offerings.
3. Use the large USDC reserves to become the market rate setter not just for DAI but also for USDC.
4. Conversion of external currently non Maker users of USDC on Compound/Aave to net users of Maker.
5. Enable highly scalable fixed rate loans that do not require active management [liquidations, oracles ect] from Maker.

## Implementation
A few ways to introduce fixed rates as collateral:

- A Vault with Oracle that uses a price oracle from a fixed rate amm
- Multiple debt vaults approved for single ERC20 Element Bond Series and each references an oracle set by time weighted price oracle from the Element AMMs.
- Pros:
- Similar architecture and rollout to other assets
- Cons:
- Limited Scale
- Inaccurate representation of the risk, bonds are much closer in risk to Aave direct deposit module than volatile assets.
- Multiple votes for multiple assets, and possibly irrational yield curves
- Vaults with Oracle hard-coded interest formula
- Multiple debt vaults approved for single ERC20 Element bond Series with interest rates hard coded by governance. Bonds are priced according to some very high interest rate [10% or more for USDC].
- Pros:
- Very Scalable
- Cons:
- No liquidations possible, similar to USDC-A
- Weird price behavior possible at debt ceiling.
- Multiple votes for multiple assets
- PSM for fixed rates
- A PSM is implemented that buys any Element bond at a preset interest rate model which qoutes price based on time to expiry.
- Pros:
- One PSM could price multiple assets, so only one vote
- Maker sets interest rates with much more possible granularity
- Highly scalable
- Cons:
- Non-standard implementation
- More pricing input needed.

**Liquidations/Emergency Shutdown**
The Principal Token matures equal in value. The only risk is insolvency before maturity, assuming this lending is isolated (e.g. only borrow Dai against Dai PT). Here a “liquidator of last resort” could step in to backstop, if they’re willing to hold onto the insolvent position until maturity. They also get a fixed rate APR along the way. Batch auctions are the simplest solution and Maker itself can act as the backstop with minted DAI.