In almost 3 months since the launch of v2, we have seen active usage of the protocol on the Ninjas side (Credit Account users). At the same time, it is noticeable that the Ninjas are engaged in shifting the capital of passive LPs to more profitable farms. So, at the time of launch, FRAX pools were in the greatest demand, then interest shifted to LUSD Convex pool, and then, respectively, to sUSD and GUSD Convex (for USDC and DAI pools).
You can see raw SQL here.
At the same time, over the past 3 months, we have seen an increase in farming rates. Below are the current figures for the most popular farms in Gearbox:
|Farm opportunity||TVL, $||APY|
This results in receiving less profits in LP side while the Ninjas earn 30%+.
At the same time, it can’t be said that everything is going smoothly on the side of the Ninjas: we see that utilization is constantly 80%+:
Accordingly, every time a large LP withdraws funds (1M withdrawal = ~3% utilization ratio), this leads to a sharp increase in the utilization ratio and an increase in interest rates:
This accordingly leads to degradation of the Ninjas user experience: if you farm with x5 leverage, and the borrow rate is up to 10%+, even if this rates are kept only 10-20 hours you need to farm the next week to cover losses incurred due to the expensive cost of borrowing.
Thus, the current interest raet curve causes the pool utilization to be “too high” and at the same time, Ninjas suffer from the borrow rate fluctuations, and the LPs are underpaid. An increase of r1 parameter (see here math behind it and current values of curve parameters) will lead to a new equilibrium point - the demand for credit from the Ninjas will decrease, respectively, the utilization will be below 85% and when withdrawing LP funds, the rate will not soar so sharply.
Do r1 parameter of Interest rate curve for USDC and DAI (and upcoming FRAX) pools as 2.5%.
This means that at optimal utilization ratio 85%, the lending rate will be 3.75% - this will reduce the demand from the Ninjas, since their profitability will decrease (for example, for sUSD convex pool it will be ~16% instead of 25% at x7 leverage).
It should also be noted that a decrease in utilization leads to a decrease in the pool’s Supply APY (because more capital stays passively in the pool), but an increase of r1 is likely to offset this effect:
I think that one-time big changes of interest rate curve can have a negative perception by Ninjas - so we can increase it in two steps:
Change it to 2%. If utilization falls below 80%, then keep it at is
If utilization is still higher 80% for the next 3 days after changing r1, change it to 2.5%.
There are 3 options to be discussed:
- Keep current r1 parameters of Interest Rate Curve for USDC/DAI/FRAX pools
- One-time change r1 for DAI/USDC/FRAX pools to 2.5%
- Two-step change of r1 parameter for DAI/USDC/FRAX pool (3-day lag between, if 1st step is enough to move Utilization Ratio to below 80%, keep r1 as 2%).