# Separate Positive and Negative Curves and their Parameters **URL:** \ **Category:** Ampleforth Governance\ **Created:** [May 6, 2025, 4:03pm UTC](https://forum.ampleforth.org/t/separate-positive-and-negative-curves-and-their-parameters/842 "2025-05-06T16:03:36Z")\ **Posts on this page:** 10\ **Page:** 1
**Author:** ![Naguib](https://yyz2.discourse-cdn.com/flex030/user_avatar/forum.ampleforth.org/naguib/32/96_2.png) [@Naguib](https://forum.ampleforth.org/u/Naguib)\ **Post date:** [May 6, 2025, 4:03pm UTC](https://forum.ampleforth.org/t/separate-positive-and-negative-curves-and-their-parameters/842/1 "2025-05-06T16:03:36Z")
**Summary:** This proposal introduces an update to the rebase curve logic that enables separate responsiveness parameters for positive and negative rebases. It also proposes restoring the deviation threshold to 5%, reversing the reduction to 2.5% introduced in[Proposal #32](https://www.tally.xyz/gov/ampleforth/proposal/32?govId=eip155:1:0x8a994C6F55Be1fD2B4d0dc3B8f8F7D4E3a2dA8F1). The proposed update, implemented in[PR #297](https://github.com/ampleforth/ampleforth-contracts/pull/297), allows independent control over the steepness of the positive and negative sigmoid curves—unlocking more balanced and purposeful rebase behavior. **Background & Context:** - [AIP-5](https://aips.ampleforth.org/AIPs/aip-5) introduced the sigmoid rebase curve, replacing the previous linear model with a more symmetric and bounded mechanism. - [Proposal #32](https://forum.ampleforth.org/t/steepen-rebase-curve-to-tighten-price-distribution/772) adjusted the curve’s growth parameters to steepen the sigmoid, with the goal of tightening price distribution and accelerating convergence to equilibrium. The steepening successfully achieved its primary objective: it virtually eliminated long periods of consecutive negative rebases during demand contractions, allowing the supply to correct more decisively and quickly reach equilibrium. However, because the same steepness was applied to both positive and negative rebases, it introduced a new issue: - Positive rebases became overly aggressive, often overshooting demand. - This oversupply led to unnecessary negative rebases shortly afterward, adding noise and volatility. - In parallel, the deviation threshold was lowered to ±2.5%, which has proven too sensitive—causing the system to react to insignificant price movements. **What this proposal is:** - Enables asymmetric sigmoid parameters — allowing different steepness for positive and negative rebases. - Preserves a sharp negative rebase curve — useful for rapid correction during demand drops. - Smooths the positive rebase curve — reducing overreactions to short-term price moves. - Restores the deviation threshold to ±5% — filtering out noise and reducing false rebase triggers. ![rebase_curve_2025_april](https://canada1.discourse-cdn.com/flex030/uploads/governance1/original/1X/fabcf1508f3474380d2e820682e83df0d0b988b4.png) ---
**Author:** ![Togenkyo](https://yyz2.discourse-cdn.com/flex030/user_avatar/forum.ampleforth.org/togenkyo/32/102_2.png) [@Togenkyo](https://forum.ampleforth.org/u/Togenkyo)\ **Post date:** [May 8, 2025, 8:50am UTC](https://forum.ampleforth.org/t/separate-positive-and-negative-curves-and-their-parameters/842/2 "2025-05-08T08:50:02Z")
I’m in favor of widening the neutral zone, but not sure about the steepening of the negative rebase curve part. It’s difficult to see on the chart, but it looks to me as if we are jumping from 0% rebase to -5% rebase within a tiny price movement of 0.01$ from 0.95$ - 0.94$. This seems very sudden to me and might create volatility around 0.95c by users trying to avoid a -5% rebase and creating a selfulfilling dynamic. I know there is VWAP, but still. Could you maybe provide us a table showing at what price levels each % of the negative rebase is happening? ---
**Author:** ![Richard\_Meng](https://avatars.discourse-cdn.com/v4/letter/r/258eb7/32.png) [@Richard\_Meng](https://forum.ampleforth.org/u/Richard_Meng)\ **Post date:** [May 8, 2025, 2:49pm UTC](https://forum.ampleforth.org/t/separate-positive-and-negative-curves-and-their-parameters/842/3 "2025-05-08T14:49:20Z")
> [@Naguib](#): > > Positive rebases became overly aggressive, often overshooting demand Could you please elaborate on this one? How we know its over shooting demand? Thank you. ---
**Author:** ![Amtaha](https://avatars.discourse-cdn.com/v4/letter/a/87869e/32.png) [@Amtaha](https://forum.ampleforth.org/u/Amtaha)\ **Post date:** [May 10, 2025, 6:27pm UTC](https://forum.ampleforth.org/t/separate-positive-and-negative-curves-and-their-parameters/842/4 "2025-05-10T18:27:03Z")
I don’t think we need to steepen the negative rebase. Since the last reconfiguration of the rebase ample has corrected quickly when in negative territory. I support expanding the neutral bounds, but the possible benefits of steepening the negative rebase don’t seem necessary in opinion. ---
**Author:** ![pz8245566](https://avatars.discourse-cdn.com/v4/letter/p/c89c15/32.png) [@pz8245566](https://forum.ampleforth.org/u/pz8245566)\ **Post date:** [May 13, 2025, 4:14am UTC](https://forum.ampleforth.org/t/separate-positive-and-negative-curves-and-their-parameters/842/5 "2025-05-13T04:14:05Z")
It is not a wise choice to blunt/dull AMPL’s market perception capabilities. ---
**Author:** ![Pascal](https://avatars.discourse-cdn.com/v4/letter/p/e9a140/32.png) [@Pascal](https://forum.ampleforth.org/u/Pascal)\ **Post date:** [May 19, 2025, 12:49am UTC](https://forum.ampleforth.org/t/separate-positive-and-negative-curves-and-their-parameters/842/6 "2025-05-19T00:49:59Z")
I mulled over this for the last week and I think I’m board with this proposal. The 2.5% was way too sensitive to random market fluctations so I’m on board with the 5% deviation threshold. Giving Ampl more wiggle room before a negative rebase may also help enable flash minting well before negative rebase, which with the 2.5% deviation threshold it seemed that flash minting almost always occured with a negative rebase. I would love to have the protection of a larger deviaton ratio of 5% and addtionally Spot flash minting which should help to reduce the sheer amount of negative rebases. I also think if the old 2.5% deviation ratio also introduced ‘fake’ positive rebases I can see that as being problematic. Because eventually the market always catches up. So yeah on board with this proposal. ---
**Author:** ![Naguib](https://yyz2.discourse-cdn.com/flex030/user_avatar/forum.ampleforth.org/naguib/32/96_2.png) [@Naguib](https://forum.ampleforth.org/u/Naguib)\ **Post date:** [July 9, 2025, 10:50am UTC](https://forum.ampleforth.org/t/separate-positive-and-negative-curves-and-their-parameters/842/7 "2025-07-09T10:50:20Z")
Hi all, This proposal took a backseat for a while, apolgize for the delay in responses. It seems that the change of the deviation threshold made it artificially look like the negative rebase curve becomes steeper, it simply starts at a later point as you can see the two curves are overlapping outside of the deviation buffer. > [@Togenkyo](#): > > I’m in favor of widening the neutral zone, but not sure about the steepening of the negative rebase curve part. @Togenkyo The proposal doesn’t including steeping negative rebase curve, the curves are almost identical. > [@Richard\_Meng](#): > > > [@Naguib](#): > > > > Positive rebases became overly aggressive, often overshooting demand > > Could you please elaborate on this one? How we know its over shooting demand? Thank you. We know from the negative rebase happening right after positive rebase to compensate for the overshooting. ---
**Author:** ![Naguib](https://yyz2.discourse-cdn.com/flex030/user_avatar/forum.ampleforth.org/naguib/32/96_2.png) [@Naguib](https://forum.ampleforth.org/u/Naguib)\ **Post date:** [July 9, 2025, 10:51am UTC](https://forum.ampleforth.org/t/separate-positive-and-negative-curves-and-their-parameters/842/8 "2025-07-09T10:51:17Z")
Agree, there is no steepening in this proposal. ---
**Author:** ![Naguib](https://yyz2.discourse-cdn.com/flex030/user_avatar/forum.ampleforth.org/naguib/32/96_2.png) [@Naguib](https://forum.ampleforth.org/u/Naguib)\ **Post date:** [July 22, 2025, 4:27pm UTC](https://forum.ampleforth.org/t/separate-positive-and-negative-curves-and-their-parameters/842/9 "2025-07-22T16:27:49Z")
Snapshot vote started: [https://signal.ampleforth.org/#/proposal/0x7d30b65f2bb09099db88fb47d0ac5781058383a7b4b1d76760d8567d0c256fa1](https://signal.ampleforth.org/#/proposal/0x7d30b65f2bb09099db88fb47d0ac5781058383a7b4b1d76760d8567d0c256fa1) ---
**Author:** ![Naguib](https://yyz2.discourse-cdn.com/flex030/user_avatar/forum.ampleforth.org/naguib/32/96_2.png) [@Naguib](https://forum.ampleforth.org/u/Naguib)\ **Post date:** [August 10, 2025, 12:36pm UTC](https://forum.ampleforth.org/t/separate-positive-and-negative-curves-and-their-parameters/842/10 "2025-08-10T12:36:40Z")
On Chain vote! > **[Tally | Ampleforth | Separate Positive and Negative Curves and their Parameters](https://www.tally.xyz/gov/ampleforth/proposal/52)** > > Separate Positive and Negative Curves and their Parameters