Exploring negative interest rates for AMPL during negative rebases (for increased borrowing)

Ampleforth Governance
1

Hi everyone!
With the recent success of AMPL breaking multiple records on AAVE, I would like to open the discussion on negative interest rates for AMPL on AAVE. I’ve spent quite some time thinking about it and I believe there are many beneficial points to the rates being able to go negative.

In the last couple weeks, we have consistently seen 100% utilization due to borrowers having to pay less interest on their loans than they made with holding their AMPL through the strong positive rebases. This was actually expected to happen and is still part of a functioning market.

What happens though once AMPL enters into a period of prolonged negative rebases? Wouldn’t borrowing demand dry up to not be exposed to negative rebases? At least this is what happened between August and Mid September, where AMPL was mostly having negative rebases and utilization on AAVE mostly stayed at below 10%.

Ideally in my view, similar to how positive rebase time periods lead to an increase of borrowing demand and a lending supply crunch, there should also be an increase of lending supply with cheap credit happening during negative rebase periods.

Now that we have entered negative rebase territory again, we will see how that second part is going to play out. If the market is going to repeat the behavior during August-Mid September, we will quickly see utilization fall to below 20/10% again.

While in that environment we might see some growth in lending supply from lenders who want to reduce their rebases by a little bit, I believe we could achieve a much stronger growth of the AMPL lending market if that utilization was not 10-20%, but more around 40-50%. Supply would increase at a much faster rate, ultimately leading to a quicker adoption of AMPL as a global reserve currency.

So how can we increase utilization during negative rebase periods? With negative interest rates!

My thoughts are, if there was the option for the interest to go negative below a certain utilization (I think 50% might be interesting), I think the following would happen:

  • There would be a demand from borrowers who want to profit from having to pay back less, maybe by selling AMPL into other stables or buying stuff
  • And even more interesting, there would now also be a higher demand from borrowers who plan to hold AMPL longterm. (Ex: Borrow 100 AMPL, rebase -3%, interest -1%, They have 97 AMPL now but only owe 99 AMPL effectively only having -2% rebase)
  • There would be a bigger incentive to supply AMPL by lenders during negative rebases, due to higher Utilization% coming from borrowers
  • And AAVE is still making money by the difference in negative interest rates for borrowers and lenders. (Especially when reaching the negative interest cap, giving a similar high revenue for AAVE as we've seen with the positive interest cap)

I just can only see win-win-win for everyone.

There would be a cap on how far the negative interest % can go, same as there is now with positive % rates. So during times of very strong negative rebases, lenders of AMPL can make money similar to how borrowers can do that during strong positive rebases.

I also have put some thoughts into how high/low the interest cap should be. Using a yet-to-be-determined high/low percentile would result in a different cap for negative and positive interest rates.

I’m very interested in hearing your thoughts about this idea!

2 Likes
2

Adding more of my thoughts that came up in the discord discussion.

Wouldn’t holders of AMPL just sell and escape the currency completely, if there was negative interest rates and negative rebases, instead of lending it out (and reduce sell pressure)?

The idea for negative interest rates(lenders pay borrowers) was to increase utilization%.

Let’s say there was a cap of -0.2%(or whatever) per day, but a utilization closer to 50%, I think more holders would be inclined to lend out their AMPL due to being able to save a bigger part of their negative rebases, than if interest was close to 0% and utilization at below 10/20%.

So it comes down to the question of:
How much are lenders willing to pay borrowers to be able to reduce their negative rebases by a substantial amount?

For borrowers it’s an amazing deal too. Imagine you can buy a car with a loan and you are getting paid to do so! I think just because of this dynamic(being paid to borrow) the interest rates would gravitate towards 0% anyway, while also giving a higher utilization%.

If lenders are not willing to pay borrowers a negative interest rate, borrowers would push the rate to 0% attracting again lending liquidity(50% utilization). Free markets basically!

3

I’m guessing some may find it risky to have negative interest rates on AAVE, because it is our “main” borrowing/lending pool (for now).

I think negative interest rates are interesting and should be implemented in a separate borrowing/lending pool, to see how the market reacts to negative interest rates. Maybe Benqi?

1 Like
4

Thanks a lot for writing down this idea! Especially for taking the time to write it in such a clear way :nerd_face: :100:

This is quite an interesting question and I could imagine a functional market evolving around this.

However, wouldn’t this lead to a circular effect in which the lenders pay the borrowers to borrow AMPLs (taking the negative rebase), which they sell to take the profit, leading to more downward pressure on AMPL?

2 Likes
5

Yes, that question arose in me too and I had to think about it for a while. At first like you, I thought about the possibility of a downwards spiral. But after thinking for longer, I found that there might going to be many more forces at play than only the pressure on price through shorting.

Introducing negative interest rates introduces a new way to generate profit through spending AMPL, either to buy goods and services or to buy other currencies like stablecoins(which effectively is shorting AMPL) and other cryptos(which is longing that crypto with AMPL as denomination).

When AMPL is in negative rebase territory, we can assume that lending activity would increase (to escape the rebase), driving interest rates and utilization down. This might lead to an increase in borrowing activity for the mentioned reasons above, driving interest rates and utilization up. So there will be forces from both sides.

For the interest to even reach the negative cap (let’s say maybe around -0.5% to -0.2% per day for now), there would have to be 0% utility and only unused lending supply, which seems to me very unlikely to happen.

But let’s assume that all borrowing in that scenario would be used for shorting, driving price and negative rebases down. Lenders keep interest rates negative (there would have to be a constant increase in lending supply), which leads to more shorting, driving price and rebase further down etc…

At one point, the price of AMPL would reach such a low price that the profits from negative interest rates wouldn’t justify the risk of getting cought in a price reversal anymore. In a scenario where AMPL was 0.70 and interest rate -0.3% per day, any shorter would have to profit from the negative interest rate for 187 days to offset the loss from AMPL going from $0.7 to $1. With a -0.5% rate it’s 112 days.

$0.7 → $1.0 = ~+42.9% (42.9% loss from shorting)
1 × 0.997 ^ 187 = ~0.57 (43% gain from borrowing)
1 × 0.995 ^ 112 = ~0.57 (43% gain from borrowing)
(My math might be off, anyone can check?)

Quite interesting, with negative interest rates, there is diminishing returns! (as opposed to compounding returns with positive rates)

Also do not forget, AMPL that has been used for shorting has to be bought back eventually(minus rates). Pushing price back higher.

What other forces might there be at play other than increased shorting activity?

First of all as mentioned in my first post, with negative interest rates, it becomes now more affordable to borrow AMPL longterm for holding.

One can save a couple of basispoints in rebases by borrowing AMPL instead of buying spot(unless you buy spot and lend it out on high utilization% in which case you save more by buying spot).

We could argue if this is a good or bad thing. I think it’s good for 2 reasons, one is that borrowed AMPL is not used for only shorting anymore and the second is that higher utilization% will lead to higher lending activity.

Which brings me to the next point of different forces at play: Higher lending activity.

What might be the implications of higher lending activity? One is that there is less supply on the market, which translates into less selling.
The second is that anyone, who is on the sideline and hesitating to get into AMPL, can now buy and lend it out to decrease their negative rebase. It makes it easier as an outsider to get into spot AMPL below the neutral price.

Ultimately, we wont know what happens until negative interest rates are implemented. We might start with a smaller cap like -0.1%/-0.2% per day and then adjust later.

2 Likes
6

Do Aave or Compound (and their respective forks) support negative interest rates or would this require custom contracts to build?

7

I asked the question in the AAVE #developers discord, if it was possible to change the Base interest to a negative value with current code or if it required rewriting it, and here is the answer that I got by “DeFSpartan”:

"I read your initial question, and that’s a very interesting idea but it wouldn’t be possible to implement without completely rewriting the LendingPool logic. The pool model works by allocating interest paid by borrowers to depositors equivalent to their share of the pool which happens. The index which tracks the total interest shares of the pool is updated here (https://github.com/aave/protocol-v2/blob/e558db406f4efdd0d076d72a7174c6d49ef21da3/contracts/protocol/libraries/logic/ReserveLogic.sol#L347) which would be broken by a negative deposit rate. Basically in the current model, depositors can only earn and borrowers can only accrue debt.

You could achieve something similar though by adding deposit/borrow incentives to AMPL and dynamically weighting them based on the rebase"

So unfortunately, that excludes AAVE for now. We could check if some of the other lending platforms would support this feature with less work required.

If anyone has more questions or scenarios what could happen if negative interest rates were implemented, keep them coming!

PS:
Changing topic name from:
Exploring negative interest rates for AMPL on AAVE
to:
Exploring negative interest rates for AMPL during negative rebases (for increased borrowing)

2 Likes