Pyth Token Phase 3

Ideas Bank
1

Summary

Phase 2 (https://forum.pyth.network/t/passed-op-pip-87-pyth-token-phase-2-pyth-strategic-reserve/2293) established the PYTH Strategic Reserve, funded by protocol revenues, to purchase PYTH on the open market. However, monthly OIS (Oracle Integrity Staking) emissions (~7M PYTH per month) still exceed these monthly purchases (~2–3M PYTH), creating net sell pressure.

This proposal introduces Phase 3: allow OIS rewards to sunset naturally when the current reward pool depletes in early March 2026 (estimated March 5th). No new allocation would be made to OIS, but we urge the DAO to consider retaining economic incentives for publishers, including a requirement to stake as a condition to providing data.

Please note that governance staking would remain unchanged — PYTH holders can continue staking for voting as they always have.

With OIS emissions ending and protocol revenues growing, the network moves toward sustainable tokenomics where purchases made by the DAO exceed emissions.

Metric Current (until Feb 2026) Post-Phase 3 (Q2 2026+)
Monthly OIS emissions ~7M PYTH 0 PYTH
Monthly PYTH purchases ~2–3M PYTH ~5–15M PYTH (growing)
Net monthly flow (OIS vs purchases) OIS exceeds purchases by ~4–5M Purchases exceed OIS

Note: Post-Phase 3 purchases based on projected DAO net proceeds ramping from ~$250K (Q2 2026) to ~$758K (Q4 2026), translating to ~5M–15M PYTH/month at current prices.

Background

  1. Phase 2 Recap

OP-PIP-87 (https://forum.pyth.network/t/passed-op-pip-87-pyth-token-phase-2-pyth-strategic-reserve/2293) (passed in December 2025) directed the Pyth DAO to:

  • Allocate 1/3 of Treasury monthly to purchase PYTH on the open market

  • Accumulate purchased tokens in the Strategic Reserve (https://forum.pyth.network/c/pyth-reserves/12)

  • Task the Pythian Council with optimizing on-chain pricing for Pyth Core, Entropy and Express Relay

The mechanism is now operational. Recent purchases:

Month PYTH Purchased Report
December 2025 2,157,087 PYTH Link
January 2026 2,160,000 PYTH Link
  1. Current Revenue Sources
  • Pyth Pro (Off-chain Subscriptions)

Per CO-PIP-9 (https://forum.pyth.network/t/passed-co-pip-9-proposal-to-assign-douro-labs-as-administrator-and-initial-distributor-for-subscriptions-to-pyth-data-off-chain/2220), Douro Labs distributes 60% of Pyth Pro revenue to the DAO.

Period Total Revenue DAO Share (60%)
Sep–Dec 2025 $262,222 $157,333
December 2025 $90,408 $54,245
January 2026 $122,833 $73,700
Total $475,463 $285,278
  • Price Feed Listings

One-time payments for new price feed listings via the Price Feed Council (https://forum.pyth.network/c/price-feed-council/6) have generated over $300,000 since 2025. Example: HAEDAL/USD (https://forum.pyth.network/t/price-feed-request-haedal-protocol-haedal/2120/4) was listed for a $20,000 payment.

  • On-Chain Protocol Fees

Fee collection is now being activated across products:

Product Status Proposal
Express Relay Live (Feb 9, 2026) OP-PIP-92
Core (Price Feeds) Pending OP-PIP-93 (this week)
Entropy Pending OP-PIP-94 (this week)

Prior to fee activation, contracts accumulated ~$100K (Core), ~$150K (Entropy), and ~$110K (Express Relay) as documented in the Phase 2 discussion (https://forum.pyth.network/t/pyth-token-phase-2/2284).

  1. OIS (Oracle Integrity Staking)

Oracle Integrity Staking (https://forum.pyth.network/t/passed-co-pip-4-upgrade-staking-program-in-mainnet-to-add-staking-for-data-integrity/1743) was launched in late 2024 following the Community Integrity Pool (CIP) design (https://forum.pyth.network/t/community-proposal-for-the-design-of-the-community-integrity-pool-cip/1571) proposed by CMS and the community. The program enables PYTH stakers to delegate to data publishers, with slashing mechanisms to penalize poor data quality.

The Pyth Data Association funded the initial reward pool to bootstrap the program’s practical implementation. Current spending: ~1.8M PYTH/week (~7M PYTH/month).

The reward allocation is nearly exhausted. The program will naturally conclude in early March 2026 absent re-funding.

Proposal

→ Do Not Renew OIS

When the current OIS reward allocation depletes (projected early March 2026), do not allocate additional tokens to this program. This is a Pythian Council operational action that:

  • Ceases new reward emissions once current funding is exhausted

  • Allows existing stakers to withdraw at their convenience

  • Preserves staking infrastructure for governance voting

  • Does not affect publisher operations or data quality

Rationale

OIS has never been triggered. Since launch, not a single slashing event has occurred. The DAO has distributed close to 100,000,000 PYTH in staking rewards to publishers and non-publisher DAO members without the slashing mechanism ever being invoked. While this may indicate high publisher quality, it raises the question of whether the current implementation of OIS is the optimal use of DAO resources.

Pyth’s primary security is structural. The network’s data integrity derives from:

  • 100+ first-party data publishers (exchanges, market makers, trading firms)

  • Aggregation algorithms that filter outlier prices

  • Confidence intervals that communicate price uncertainty

  • On-chain verification of cryptographic signatures

Ending OIS as originally designed enables sustainable token economics.

Current OIS emissions of ~7M PYTH/month exceed the ~2–3M PYTH in monthly purchases. Eliminating OIS emissions to all stakers (publishers and non-publishers) while growing protocol revenues—projected to ramp from ~$250K/month (March 2026) to ~$758K/month (December 2026)—allows purchases to exceed this emission source.

Implementation

No contract upgrades required. The Pythian Council has the constitutional authority to manage OIS pool capacity and reward rates.

Action: If this idea receives community support and passes as a CO-PIP, the Pythian Council will pass an OP-PIP to set the OIS reward rate to 0 before the current funding depletes (target: before March 5, 2026).

Staker Impact: Existing stakers can withdraw tokens at their convenience. No lockup changes. Governance staking functionality remains available.

Future Considerations

I would suggest to keep the following points out of scope for this proposal but definitely merit future community discussion:

  1. Mandatory Publisher Staking — Requiring data publishers to stake PYTH (subject to some slashing risk) as a condition to providing data could create stronger alignment between data quality and economic incentive.

  2. Unlock Schedule Adjustments — The current vesting schedule includes significant unlock events (May 2026/2027). Modifications could further support sustainable token economics. These are substantial changes warranting dedicated analysis and separate proposals.

Discussion Points

  1. Staking behavior: Do community members anticipate that sunsetting OIS rewards will lead to significant unstaking, or will stakers remain for governance participation?

  2. Publisher participation: Are there concerns about publisher engagement without OIS delegation incentives?

4 Likes
2

I think the market impact of removing OIS rewards deserves more focused discussion alongside the long-term tokenomics case — especially given how large the staked base is relative to trading flows.

According to @bats4 Dune dashboard, ~1B+ PYTH are currently staked, which represents a very significant pool of tokens relative to circulating supply.

If OIS rewards sunset and even a modest percentage of these tokens were unstaked and sold over a short period of time, the resulting supply entering the market could be meaningful. The key question isn’t whether emissions are being reduced long term — it’s how the market absorbs potential short-term flows if incentives abruptly drop to zero.

We should also be realistic about behavioral incentives. I do not believe a significant proportion of stakers are primarily participating for governance. While this is only my interpretation, I believe OIS was introduced in part to align with the broader market expectation that staked tokens generate yield. If the reward component is removed entirely, we should carefully consider how many participants will remain staked purely for voting rights.

Because the staked base is large, I would be very interested in seeing any modeling that accounts for:

  • Unstaking behavior assumptions — what fraction of stakers might exit once rewards stop?

  • Liquidity depth across Solana DEXs and major CEX order books — how much volume can realistically be absorbed without significant slippage?

  • Timing & velocity of flows — gradual trickle vs. clustered exits around key milestones (e.g., reward depletion)?

Additionally, if we are making a structural change of this magnitude to emissions, it may be worth discussing whether it should be accompanied by commitments around future unlock schedules. In other ecosystems, when incentive programs are reduced or removed, proposals are often paired with adjustments to future emissions or team/token unlock allocations to reinforce alignment and mitigate perceived supply overhang.

I fully understand and appreciate the long-term sustainability rationale. However, incorporating explicit analysis of short-term supply dynamics — and potentially aligning unlock policy alongside emission reductions — would give the community greater confidence in the overall transition.

9 Likes
3

I agree with everything mentioned above by @Frozenmind .

In my opinion, I believe the unstaking due to OIS rewards being sunset will be quite large amongst those who are still active in the space.

Unfortunately I think the governance narrative is quite tired these days and doesn’t really hold much weight. Attention has now shifted to tokens with tokenomics geared towards supporting token price and not being bound by unstaking lock ups. I could also see an announcement like this being made at near on ATL prices disgruntling a lot of the community and potentially creating fud on the timeline.

If the OIS rewards are sunset I think it would be best to be announced alongside future plans for Pyths tokenomics or alterations to unlock schedules.

9 Likes
4

Dear @zenyas,

Thanks for your post.

While I agree that the current Strategic Reserve buyback alone is not sufficient to offset OIS emissions (~7M PYTH/month vs. ~2–3M purchased), I would suggest considering a more adaptive approach rather than fully sunsetting emissions.

One possible path is to dynamically tie OIS emissions to DAO buybacks. For example, the level of OIS rewards could scale with protocol revenue and the amount of PYTH repurchased on the open market.

This would better align incentives across the ecosystem: if the network generates stronger revenue and supports the token via buybacks, higher staking rewards can be sustainably funded. Conversely, in lower-revenue periods, emissions would naturally compress. In my view, this creates a healthier feedback loop between growth, token support, and participant incentives.

Concerns:

  1. Retail staking incentives: I believe a significant portion of retail participants are staking in OIS mainly for the rewards. Without meaningful rewards, many may choose to unstake rather than remain purely for governance participation, especially if they are not deeply involved in the DAO, causing a potential massive dump pressure on $PYTH token.

  2. Publisher incentives: Similarly, if publishers no longer receive PYTH emissions via OIS, it is unclear what the long-term economic incentive is for providing high-quality data. If the expectation is that publishers continue contributing without direct rewards, it may be helpful to articulate the alternative incentive structure more explicitly.

  3. Token unlock overhang: More broadly, the structural sell pressure may not come solely from OIS emissions. The large and predictable annual unlocks remain a key consideration. With an estimated ~20% increase in circulating supply around May 2026, it would be helpful to understand whether there are any plans to revisit PYTH tokenomics holistically.

Overall, I support the goal of moving toward sustainable token economics, but I believe a demand-linked emissions model may achieve similar objectives while preserving incentive alignment across stakers and publishers.

From SCP with Love

8 Likes
5

This looks like the trend of the market these days. As we mature, and realize the incentives model, as it is now, stake to get passive rewards, isn’t helping the tokens price.
I agree something has to give. Removing it completely and not having something to go in its place would be reckless and create a spiral in the tokens price.

I think ee should discuss the tokenomics model and then Implement something to replace the existing system.

Because if you just want to value a Governance Token, that has almost no other use, then we are talking about memecoin prices and not blue Chip ones. Also, there will be a lot of unhappy and underwater holders.

7 Likes
6

I agree with the comments above. Removing incentives entirely could create overwhelming selling pressure and would almost certainly impact the price of the token. Imho, this is a risky move, especially considering that we may be heading into a bear market sooner rather than later and - unfortunately - governance alone is not enough to sustain the token’s performance. It might be more prudent to introduce amendments to the tokenomics first before proceeding with this change

10 Likes
7

very much agree with everything said above by @Frozenmind @lowkeigh and @scp

So far I do understand the good reason behind - the initial staking reward pool is sunsetting and it needs a replenishment. Tug of war - buying back the tokens and at the same time giving more for the rewards does not make any good at all

But at the same time we are talking, as mentioned above, about 20% of the total supply being staked

I guess there is a certain point reached when the utility of the coin should be discussed further:
For example there is a discount on Pyth Pro package if you are staking XX amount of PYTH
Burning some Pyth with the revenues
Decreasing the circulating supply and so on

I do fear that just cancelling OIS rewards will not solve the case at all

6 Likes
8

Great feedback @Planck ,

I think we need to first address the “giant in the game”: the token unlock & tokenomics. While revising the tokenomics, we can then consider how Oracle Integrity Staking (OIS) fits into the system. OIS and its reward structure didn’t exist when the original tokenomics were designed.

To make the number more clear:

  • Monthly buyback (current): 2–3M PYTH
  • Monthly OIS emissions: 7M PYTH
  • Token unlock in May 2026: 2.13B PYTH (i.e., 2,130M PYTH)

Addressing the issue of a few million in buybacks versus OIS staking (just around 2% of the upcoming unlock), while simultaneously ignoring the 2.13B $PYTH unlock, doesn’t really solve the supply-demand equation.

I think I’ve covered everything I wanted to say, and I hope to see more community feedback. I’ll put my pen down here.

From SCP with Love

2 Likes
9

Aggressive utility programs. Multiple parallel programs with the same objective, increase token utility and demand.

I’m just glad someone was listening about the discount idea lol.

3 Likes
10

great offer, why not integrate tariff payment into Pyth Pro, using Pyth tokens

For example, if you pay in Pyth tokens, you get a discount of about 20% at the exchange rate to the dollar.

11

Overall, I support the original idea. And while i do agree with most of the comments made by you all here, I am also in favor of splitting changes rather than bundling them together.

  • About the unstaking risk — I think the fear is overblown.

Before OIS launched, there was already ~1B PYTH staked in governance. Today that number is ~1.4B. So governance staking grew by about 400M tokens over the OIS period. And of the 1B currently staked in OIS, most of them were staking before OIS even existed.

And again, stakers who have remained committed through market cycles are clearly here for long-term protocol alignment: an 8% APY isn’t what drives their conviction in my opinion.

  • About the elephant in the room (token unlocks) — you’re right, and I believe we should address it in a follow-up proposal.

As zenyas stated in his Future Considerations, two unlocks remain: May 2026 and May 2027, each about 21% of total supply (~2.1B tokens).

I agree we should consider converting annual cliff unlocks to monthly linear vesting (or something similar) — this brings smoother distribution, less “event risk” narrative, and generally easier for the market to absorb. Delaying the next unlock by say 6 months is also technically possible and worth considering to further build alignment.

That said, one thing I’d like to emphasize: the more changes bundled into a single proposal, the less likely it is to pass as I only need to disagree with 1 point (out of many) to vote against and block the changes. So I’d support tackling unlock reform as a follow-up proposal.

  • About token utility — “PYTH would become just a governance token”

I think this misses the point. Neither OIS nor governance staking drives token demand. I believe that what drives demand is Pyth’s actual protocol performance — overall adoption, revenue from data fees, Pyth Pro, Express Relay, usage, etc.

As for accepting PYTH as payment for Pro (and potentially offering discounts), that’s a DAO-authorized distributor business decision (i.e. Douro Labs or who will be appointed in the future by the DAO). As far as I understand, there’s nothing preventing them from offering such a package. @zenyas might be able to share if such is offered to prospects, and if so, what is the share of PYTH deals vs cash (or stablecoins).

12

The original proposal is based on the premise that OIS emissions is the cause of net sell pressure, and that stopping OIS emissions is a move towards sustainable tokenomics.

Since the proposal is premised on that main consideration, the issue of tokenomics/reducing sell pressure becomes the key factor that has to be talked about, as much as we want to separate the issues.

Based on a brief on-chain analysis, the majority of the OIS rewards were redirected back to protocol-owned wallets, so the actual amount of PYTH entering the open market (and contributing to sell pressure) is even smaller.

In contrast, the token unlocks to categories such as Publisher Rewards (22%) and Ecosystem Growth (52%) far outweigh the OIS emissions as @scp pointed out.

For comparison, the current publisher rewards unlock schedule equates to ~51M PYTH per month (2.15B divided by 42 months). And if we just zoom in on this unlock category, publisher engagement clearly becomes the central issue. I am guessing that a large volume of token selling comes from publishers (more detailed on-chain analysis can be done to prove this). So the question becomes - why are publishers unwilling to hold, stake, and build a vested interest in the future of Pyth? More has to be done to better engage and align interests of publishers - with or without OIS. Tokenomics could very well be one of the key issues.

Then, there’s also the token unlock category for Ecosystem Growth, which probably contributes to a certain amount of sell pressure as well. Given its massive size, the DAO and community would probably benefit from some broad reporting/transparency on actual spending from this bucket, providing reassurance that this is not the main cause of tokens flooding the market.

Looking at the responses from the community so far, it seems like the sentiment is generally hesitant about the idea of sunsetting OIS emissions. So if one of the purposes of this discussion is to gauge community sentiment towards the idea, it looks like we have our answer.

But from a logical and mathematical perspective, I actually agree with @KemarTiti
that the probability of mass unstaking simply due to sunsetting of OIS is overblown.

5 Likes
13

The simple solution to all the issues discussed seems to be a direct token burning of 50% for each upcoming unlock.

Put up the proposal and let’s get voting.

4 Likes
14

Dear @KemarTiti,

I do not disagree with the option of sunsetting OIS. However, I believe we need to reorder the issues we are trying to solve.

First, we should revisit the tokenomics and define a more sustainable long-term model. Based on that revised model, then we conclude that sunsetting OIS is part of the path toward sustainability.

SCP with Love

2 Likes
15

I’m going to have to say I disagree with this train of thought. Whether warranted or not there was a lot of hype around staking Pyth in the hopes of getting airdrops from eco partners which would have been a big driver early days for some stakers.

So yes, while initially governance may have attracted that initial stake I would bet a larger chunk than just the 400m that came on board over that period have stayed for the 8% apy.

I have a follow up proposal to my original one which I think may be a nice middle ground and will keep the community/retail token holders happy in the interim.

With the suggestion of token unlocks and emissions on that front being addressed, why don’t we cut OIS yield from publishers as they receive rewards regardless and need to maintain their minimum stake whilst keeping the yield in place for regular holders.

With the rate Pyth Pro submissions are coming in at, I imagine this would come close to covering any shortfalls that would have been there from the original 7M Pyth figure.

This way if we do decide to completely sunset OIS rewards we could remove the community portion alongside adjustments to token unlocks which I imagine would be generally viewed more favourably as a win-win situation.

5 Likes
16

That’s a very solid analysis.

Definitely would benefit of analyzing where the selling pressure is coming from.

I think that both publisher having more skin in the game and significant change to the unlock schedule needs to be addressed over the next two months and avoid the May 19 unlock to happen in its current form.

3 Likes
17

Yeah :100: smth need to be done before may agree

Im a staker so i support bigger ApY / stable rwrds or smth in that direction and not a favour of removing the staking option.

1 Like