Add factors for reducing non locked supplies APR

Introduce a new factor, NonlockedSupplyAPRFactor, for each asset, for reducing non locked supply APR by:

new non locked supply APR = actual non locked supply APR * NonLockedSupplyAPRFactor / 1000.

Developer team must implement a modification in the smart contract code in order to apply this correction on the equation that computes non locked supplies APR, and also adjust computation of borrowing APR accordingly, which will be reduced as less borrowing interests will be needed, in order to conserve the mathematical equality between total interests paid by borrowers and total rewards paid to supplies and vires stakers.

Set initial value of this factor to 20 for all assets.

Notice that:

  • The initial values of this factor in the proposal reduces non locked supplies APRs to the 20% of the actual values, for all assets. This factor can be adjusted later at any time, via voting.
  • Locked supplies APRs must not be modified by the implementation of this proposal, regardless the lock period.
  • Locked vires APR must not be modified by the implementation of this proposal.
  • LP tokens need to have the same APR as non locked supplies, so they are affected in same way as non locked supplies with this proposal. Otherwise it would be easy to bypass the rate cut, by exporting non locked supplies to LPs. If you have LPs and want max interest rate, you will need to import to a locked supply. LP tokens are not locked supplies, regardless high utilization may temporally make them unredeemable.
  • The proposal alone doesn’t have any effect until the team implements the logic that applies the new factor introduced.
  • The proposal that introduced the locked supplies was on same direction, And initially worked, but very soon the difference between locked and non locked supplies became small. So incentives to lock supplies became low. In addition, that proposal didn’t solve the problem of non locked deposits with high interest rates (described below).

The aims of the proposal are:

  • Avoid the systemic liquidity issues generated by high APR of non locked supplies. This is a major flaw in the vires protocol.
  • Makes more justice in the actual difference between locked and non locked supplies rates. Actually you get a very small extra rate for locking your deposits, so there is small incentive to lock. Most locked funds were created when difference was still big, but as locked supplies increased, this difference made very small.
  • Non locked supplies interest rates have currently all the advantages: high interest rates and withdraw at any time (provided there are liquidity). If you want availability, you have to get less interest, If you want more interest rates, then lock.
  • Reduces withdraw pressure via two paths: by encouraging locking of supplies, now very discouraged due to very small difference between locked and non locked supplies, and by reducing the accumulation of debt. Less competition for withdraw = more available to withdraw for those who prefer not to lock funds. So, the proposal reduces interest rates of non locked supplies, but improves withdraw availability on high utilization situations.
  • There are people that complain, with reason, that you can borrow while providing collateral with high interest rate (looping), thus reducing a lot the effective borrowing rate and yielding to borrowing abuse and liquidity unavailability. So, this proposal increases the difference between borrowing rate and supply rate, as you can borrow only by providing collateral via non locked supplies. The result are two well differentiated roles: investors that lock supplies for providing liquidity in exchange of high interest rates, and borrowers that supply at low interest for collateralize their borrowings.
  • In addition, after the correction, interest rates will reflect better the temporal preference of suppliers: those who press more for withdraw get less interest rates. Those who want more interest rates must lock.
  • And the platform will be healthier, as the non locked supplies will not generate unnecesary withdraw pressure nor contribute to systemic liquidity unavailability via high interests that has to be readily paid.
  • All this without affecting the long term investments incentives. Interest rates don’t change for locked supplies, even for a 3 month lock period.

Some may ask: why not to increase protocol share of an asset, that is, the portion of interests paid by borrowers that are paid to locked vires?

  • First, because this should be a fixed parameter of the system. Otherwise the same people with voting power can easily vote for extracting rewards from others, thus encouraging bad governance incentives to the system. This proposal, on the other side, only reduces non locked supplies interests, but doesn’t increase locked supplies ones with it.
  • Second, because it does not help for solving the problem of non locked high interest rates. It only changes the destination of the paid interests from non locked supplies to vires stakers, which also receive rewards in a non locked fashion.
  • One may answer that, as borrowing interests under this proposal are reduced, it will encourage more borrowing and so utilization will grow back and problem will raise again. However, we must not forget that under high utilization (right now, above 90%), borrowings are denied. So this effect is really prevented by current system of adaptative limits (with parameters that can be adjusted further if necessary)

Some math with current data:

  • The current, before adjustment, borrowing rates for USDC and USDT are 40%.
  • With 250.9M USDT borrowed and 291.7M USDT borrowed, the total daily interest paid by borrowers are 275K USDT and 320K USDC.
  • 10% of each goes to vires stakers revenues: 27.5K daily USDT and 32.0K daily USDC
  • Remaining 90% goes to suppliers: 247.5K daily USDT for USDT suppliers and 288K daily USDC for USDC suppliers.
  • With a total supply of 252.3M USDT, each 100 USDT supplied receive 247.5 / 252300 * 100 * 365 = 35.8 USDT annualy (35.8% APR)
  • With a total supply of 291.9 USDC, each 100 USDC supplied receive 288 / 291900 * 100 * 365 = 36.0 USDC annualy (36.0% APR)
  • There are 136M USDT locked supplies and 116.3 USDT non locked supplies. So, locked USDT supplies receive a total of 133.4K USDT daily and non locked USDT supplies receive 114.1K USDT daily.
  • There are 68.7 USDC locked supplies and 223.2 USDC non locked supplies. So, locked USDC supplies receive a total of 67.8K USDC daily and non locked USDC supplies receive 220.1K USDC daily.

So, with proposed corection of non locked interest rates:

  • Non locked USDT supplies receive 20% of 114.1K USDT daily = 22.8K. APR for non locked USDT supplies becomes accordingly, 7.2%
  • Non locked USDC supplies receive 20% of 220.1K USDC daily = 44.0K. APR for non locked USDC supplies becomes accordingly, 7.2%
  • This implies a total reduction of USDT suppliers interests of 91.3K USDT daily, and total reduction of USDC suppliers interests of 176.1M USDC daily.
  • This reduction need to be cancelled on the borrowers payment side, as total interest received by supplies = total interest paid from borrowers.
  • So, total USDT daily interest paid by borrowers go from 275K USDT to 183.7, and borrowers USDT APR after adjustment becomes 26.7% from 40%.
  • And, total USDC daily interest paid by borrowers go from 320K USDC to 143.9, and borrowers USDC APR after adjustment becomes 18% from 40%.
  • Total borrowing payments of USDT and USDC combined goes from 595K daily to 327.6K daily.

These numbers will tend to partially recover back as part of non locked supplies go to reinforce locked supplies, but in long term will reflect the temporal preferences of suppliers and make the platform healthier, by avoiding the systemic liquidity issues generated by high APR of non locked supplies.

1 Like