EIP-9: Coupon Premium Curve Evolution

The current coupon premium curve, which is used to calculate the premium offered on coupons as a function of debt ratio, is not adequtly meeting the current market conditions. This proposal is a result of converations wiith eqparenthesis and attempts to address this issue. The concern was originally raised in “ideas” under the topic “Premium Pricing”.

The current curve used is f(x) = k ( (1/ (1-x)^2) - 1 ) with k set at 1/3.
The curve starts at zero, and asymptoticaly approches infinity as x is raised from 0 to 1.
This fits the stated criteria in the white paper, of offering zero premium when debt is zero, and offering ever higher premium as ratio of debt to net supply approaches 100%.

In the future we like to consider a function that includes variables in addition to debt ratio, such as outstanding number of coupons, price, volatility, cummulative distribution fucntion, and so on, to accurately reflect the price of the underlying derivative, as projects like Keep3R start providing the reliable relevent oracles on chain.

For the sake of simplicity, on the immediate term, we are only suggesting the following modifications to “steepen the curve” and accelerate contracton cycles:

  1. Change K from 1/3 to 1.
    This will raise the premium faster.
  2. Lower Max Debt Ratio from 35% to 20%
    This will cap the premium at 56%
  3. Cut the debt in half to ease any potential premium shock, upon commit.

Keep in mind this was the result of of a group discussion, and all the ideas are not mine. Thanks to those who contributed. We are open to additional suggestions.

thanks,
Patrick .

5 Likes

Draft implementation of this (EIP-9) built on top of PR #10 (which implements EIP-8 and some QoL improvements).

I left the overall form of curve and curveMean so that it is easy to change again if we want to switch back to k= 1/3.

LMK if you want a draft implementation built off something other than PR #10.

2 Likes

@patd Would you be able to provide some example scenarios of how this curve would work? I think it’d help elucidate how it will function.

1 Like

Here are the current and new premium as a function of debt ratio, with the old cap (35%) and the new new proposed cap (20%) respectively:

We are incrasing k by factor of 3 meaning premiums will increasae by factor of 3, or grow 3 times as fast.

Are there any apparent negatives here?

None are immediately coming to mind for me

We didn’t even change the curve fundaemntally, Just changed the slope. VERY LOW RISK from contract code perspective. As I mentioned there is a lot that can be done to improve the curve, but we tried to do something really simple, non controverisal and conservative here! Deferred the rest for futures revisions.

2 Likes

I want to bump this so we have a chance of getting it in before the next contraction.

1 Like

Working on the implementation now. What do we think about going with a debt cap of 15% instead, with max premium at 38.4%? I’m hesitant to raise the max premium with this EIP.

1 Like

I think the higher debt ratio and higher premium allowed the better. Allow the market forces to work. In times of great stress you need to offer the highest rewards for someone to come in and take that risk to keep the peg moving back up on a large contraction.

1 Like