TAP-31: Adjust RBS Capacity

nicnombre

Following the earlier RFC, this TAP proposes the adjustment of RBS wall capacities, with three 'For' options -- 70% reduction, 80% reduction, or 90% reduction -- and an 'Against' option.

The purpose of this proposal is to bring RBS to a state in which it operates as it is meant to: providing challenging, but not insurmountable, liquidity in a range. The current parameters, which offer 10% of supply on the upper wall and 1% daily emittance in the upper cushions, is too extreme to accomplish the aforementioned goal. It also provides a large source of liquidity which could be abused by a malicious actor.

I feel this change is urgent, and I am trying to push a governance timeline that reflects that. Considering more harm can be done by a too-funded RBS system versus a not-funded-enough system, I suggest a liberal capacity change now that may be reverted or augmented in the days and weeks that follow. Keep in mind that, though RBS launched over a year ago, this is the first time the upper range has been tested. I see this as similar to a launch day problem, though the process of fixing it is more tenuous than it might have been day one.

To recap, vote options would be:

Reduce RBS walls by 70%
Reduce RBS walls by 80%
Reduce RBS walls by 90%
Other/Do not reduce RBS walls

dr00

I appreciate the urgency here. Could you explain some of the possible outcomes given the different options vs leaving things as-is? I think most holders will want to understand the various scenarios to make a more informed opinion when voting. Thanks!

nicnombre

dr00 yeah. as is its doubtful that the cushion is really utilised. Range can rise with the MA (currently ~$0.35 from this) but a wall of 10% supply most likely deters anyone from engaging with it. To me at least, the range should provide a check on the market, forcing a demonstration of volume and consensus in some direction if it wants to reprice, but 10% supply as that demonstration is unreasonable.

I think the risk presented by too large of a reduction is a wall breaks and the range becomes irrelevant for some period of time but as someone who would like to see a healthier premium develop before there are real emissions this is acceptable to me. I think personal opinions on that should drive peoples votes.

indig0

Would urge you to make only 3 choices, removing the middle option (70%).

abipup

FYI I'm gonna say "reserve factor" as a proxy for "both ask and bid factors" since imo they should be equal, even though they're technically two system variables. This factor is a % of treasury value dedicated to wall capacities and is currently set to 0.1 (or 10%).

I would propose to also shave down the choices to make here, and instead request a 50% reduction to 0.05 reserve factor, 80% reduction to 0.02 reserve factor (your initial proposal from the RFC), or 0% reduction to leave reserve factor at 0.1.

I really don't agree with the rush from a security standpoint; as far as I can see, a malicious actor has no path to a successful attack using the upper wall. Current price action has no bearing on this attack's probability of occurrence or success.

That being said, I believe there's a mismatch between total system liquidity, and RBS capacity. We have quite low liquidity and quite high RBS capacity. Effectively this causes us to bounce around the range without breaking through, leading to unpredictable emissions at a high scale. I'm most favorable to a 50% reserve factor reduction to balance this discrepancy out. I think 80% goes too far and would need immediate readjusting, in the case we'd go to a decent premium.

I did some simulating based on @rusowsky's code and have some illustrative charts below. I only use this sim to informatively compare bullish and bearish scenarios. The system is too complex in practice to have a precise picture of reality, so take them with that grain of salt. I would expect price to be more directional than the random price walk used in the sim; I'd expect more breakouts of cushions and walls than what's portrayed in the graphs below.

Bullish scenarios:

https://imgur.com/a/RJeWOsG

https://imgur.com/a/hTPd8SY

https://imgur.com/a/zsFgpqj

Clearly we spend way too much time between cushions with 0.1 reserve ratio. It's indeed too conservative. With a 0.02 ratio I worry about too intense of pullbacks being allowed as there's otherwise very little system liquidity to support a premium. I like 0.05 as a happy medium. It still allows breakouts but is stronger defending pullbacks in the bullish scenario.

Bearish scenarios:

https://imgur.com/a/J8xF7UI

https://imgur.com/a/73VcZgr

https://imgur.com/a/tZsj9yN

The bearish scenario essentially keeps us around backing, regardless of RBS setup. With a 0.1 reserve ratio we just bounce between cushions forever with little gain. A 0.02 reserve ratio gives substantial room for activity, where 0.05 allows less movement.

Personal conclusion

I believe either 0.05 (50% reduction) or 0.02 (80% reduction) would be okay. I personally like 0.05 better because it is sufficiently lenient in bullish scenarios, but maintains decent liquidity at premium vs 0.02. In my opinion 0.02 skews the liquidity balance too far in the other direction but still believe it's a better setup than current.

abipup

Also very sad that the forum refuses to display the pictures I post on it…

hodlmao

Personally in favor of 90% reduction, at current prices it's still quite significant $2.5M and with higher OHM price in future, it will be even bigger amount.
Current setup with 25M dollarinos is not really adequate to our recent liquidity deployments.
I would also streamline the voting process and reduce the allowed options, this is too much imho.

unbanksy33

Agree that high capacity deters price discovery. Reduce RBS walls by 90%.

z_33

In favor of reducing capacity. Only suggestion is to make this a ranked-choice vote on snapshot if possible.

cc: @abipup

Relwyn

[unknown]

Super appreciate you taking the time to run the sims and sourcing the evidence. Helps make a more informed decision. I'm leaning 80%.