Update cDAI Interest Rate Model

Summary

Eight days ago, we proposed decoupling the cDAI interest rate model from MCD parameters. While we received positive feedback regarding the updating of cDAI interest rate model, the community also clearly indicated a reticence to decouple from MCD at this time. As a result, we are proposing an updated cDAI interest rate model, but this model does not decouple from MCD. We intend to continue to iterate on the model over the coming months.

Updated Model

• Deployed contract
• Pull request

Interest Rate Model Updates

We are proposing the following changes to the cDAI interest rate model:

1. Change APY at 100% util

• Current: 15%
• New: 25%

2. Change the Utilization Rate that defines the "kink"

• Current: 90%
• New: 80%

3. Change gapPerBlock, the parameter that determines the "slope" prior to the kink

• Current: 2%
• New: 4% (note: with current SF / DSR, this gives an APY of 4.34% at the kink)

4. No change to Base Rate per Block (APR at 0% utilization)

• Current: 0%
• New: 0%

Rationale for the Changes

Change 1: Increase APY at 100% Utilization

When a Compound asset is 100% utilized, suppliers are unable to withdraw their capital, and new borrowers cannot take on loans. This is extremely suboptimal, and should be reflected by handsome compensation to suppliers. In our view, 15% APY is not sufficiently high to achieve rapid asset re-allocation in a 100% utilization scenario, nor does it accurately reflect the lockup that suppliers experience. As such, we propose raising this parameter to 25% APY.

Change 2: Change the Utilization Rate that Defines the "Kink"

The "kink" in the cDAI interest rate model reflects the utilization rate at which the slope of the interest rate goes from "gradual" to "steep". That is, below this utilization rate, the slope of the interest rate curve is gradual. Above this utilization rate, it is steep.

Because of this dynamic between the interest rate curves before and after the "kink", the "kink" can be thought of as the target utilization rate. Above that rate, it quickly becomes expensive to borrow (and commensurately lucrative for suppliers).

In our view, 90% Utilization should not be the target utilization. This would mean that the protocol aims to only allow 10% of capital suppliers to exit the system at any given time. In our opinion, this is insufficient. We suggest that a greater portion of the capital suppliers should be able to exit at any time, and as such propose lowering the utilization rate that defines the "kink" to 80%.

Change 3: Change APY at "Kink"

As we said in Change 3, the "kink" defines the target utilization of the market. Thus, the interest rate at the kink is the target interest rate. And in our view, 2% APY is too low to be the current "target interest rate". We approach the question of "what is the right target interest rate?" with humility. We honestly don't know. As such, we decided to approach searching for the ideal rate by doubling it, and then adjusting based on further market indications. We hope to see market indications that suggest that the curve is "in range". Should these indications materialize, changes to the parameter can become more incremental.

No Change to Base Rate per Block

In our view, it is appropriate to set the interest rate at 0% when there is no utilization of the asset. At zero utilization, the market is indicating that borrowing that asset is not attractive. Increasing the interest rate would only exacerbate the unattractiveness of borrowing the asset. As such, we left this parameter unchanged.