Issue #18 closed under the title “The Story Was Bought, the Budget Was Not.” Requests totalling 22.22 million ada expired without gathering enough votes, and resubmissions that had cut their asks by roughly 70% failed as well. From that, we wrote that the reason for rejection was not the amount alone: cutting the amount can reduce opposition, but it does not reduce silence. We closed with this line: “The next votes to be counted are our own.” The count came back at 15.
This week, both halves got an answer. The silence broke. AlphaGrowth’s 120 million ada — the proposal we flagged in #18 as unusual for the amount of explicit opposition it had drawn — was ratified in epoch 649. Support moved from 42.74% to 80.1%. Not one ada was cut from the request.
And the place where silence did not break became just as clear in the same week. The two proposals that require stake pool operator votes sit at 2.63% and 1.71% against a September 1 deadline.
Executive Summary
- 120 million ada was ratified. AlphaGrowth’s Cardano PRIME reached ratified status in epoch 649. DRep support 80.1% (154 votes, threshold 67%), voted-no 7.67% (39 votes), and not-voted down to 8.5%. That is a move from 42.74% support at the time of #18.
- What broke the silence was not a smaller ask. This proposal was not reduced at all. It stands in direct contrast to the two proposals we watched fail after deep cuts. What moved was not the amount — it was the volume of participation.
- SPO votes have not moved. The two proposals expiring September 1 (epoch 653) stand at
minPoolCost2.63% SPO support (24 yes, 1 no) and Constitutional Committee renewal 1.71% (9 yes, 1 no). Both require 51%. The “15” from #18 is now 25, but in weight terms it is still far short. - Four proposals are now live on the Constitution portal. The first CAP/CIS filings — item 5 on our watch list in #18 — have arrived: four in consultation, zero ready, zero done. One of them addresses the bundling of changes into a single proposal.
- Dijkstra brings the next operational task. Intersect’s weekly update records that Leios, which follows Dijkstra, will require SPOs to register BLS keys during its bootstrapping phase.
- The market went the other way. ADA fell −11.4% on the week ($0.201561 → $0.178548), giving back most of the +19.3% from #18.
1. Market Pulse
Morning fixed-point readings, August 8 to August 15.
| Asset | Aug 8 | Aug 15 | Week |
|---|---|---|---|
| ADA | $0.201561 | $0.178548 | −11.42% |
| BTC | $64,889 | $62,768 | −3.27% |
| ETH | $1,914.41 | $1,875.23 | −2.05% |
| NIGHT | $0.01910631 | $0.0172702 | −9.61% |
ADA was +19.3% in #18. This week it returned most of that. Against BTC at −3.27% and ETH at −2.05%, ADA’s decline sits deeper than the broad market move.
One note of caution. In #18 we wrote: “Price front-runs expectation, while the treasury requires 67% consent. They simply move at different speeds.” This week the reverse happened. The treasury passed its largest item and the price fell. Expectation and funding diverge in both directions. Using one to explain the other tends to produce the wrong reading.
2. 120 Million ada Passed
In epoch 649, Withdraw 120,000,000 ada for AlphaGrowth's Cardano PRIME reached ratified status.
When we placed this proposal first on the watch list in #18, the picture was: 42.74% support, 11.01% voted-no, expiry at epoch 649. We noted that “a proposal carrying this much explicit opposition is unusual lately, and it is also the largest amount this week.”
Here are the figures as of 09:59 JST on August 15, 2026.
| Measure | Value |
|---|---|
| DRep support (ratification basis, threshold 67%) | 80.1% (154 votes) |
| Voted no | 7.67% (39 votes) |
| AlwaysNoConfidence | 3.73% |
| Not voted | 8.5% |
| Abstain (outside the denominator) | 12 votes |
| Constitutional Committee | 6 yes, 0 no, 1 not voted |
Support moved from 42.74% to 80.1%, a shift of more than 37 points, and not-voted fell to 8.5%.
One point on how to read this. The voted-no share fell from 11.01% to 7.67%, but that does not mean opposition shrank. Thirty-nine votes were cast against. The share fell because the delegated weight on the yes side grew and changed the composition of the denominator. Reading that as “opposition declined” would misread the week. What declined was not opposition. It was silence.
3. What Broke the Silence Was Not the Amount
In #18 we wrote:
Cutting the amount can reduce opposition, but it does not reduce silence.
This week’s result does not contradict that. If anything it confirms it from the other side.
Scalus cut its ask from 8.5 million ada to 2.46 million, a 71.0% reduction, and still expired without enough votes. dOSPO/OMF cut from 12 million to 4.09 million, a 65.9% reduction, and expired as well. AlphaGrowth, meanwhile, gathered 80.1% on 120 million ada without cutting a single ada.
Across those three, the size of the request and whether votes arrived do not line up. Two point four six million did not clear; one hundred and twenty million did. What can be concluded is limited, but at minimum “it failed because it was too large” does not hold for this week’s examples.
So what was different? We do not have that answer. What the voting record shows is that more people voted. What moved them sits outside the record. We could speculate here, but that would be composition rather than observation. For this week we will leave the facts as facts.
The one thing that is certain is that clearing the threshold required participation, not persuasion. Sixty-seven percent cannot be reached by having more supporters alone, because delegated weight that does not vote does not count toward the yes side in the ratification calculation. Thirty-nine votes remained against, and it still reached 80.1%. That says a proposal can pass without reducing opposition, provided participation rises.
4. But SPO Votes Have Not Moved
This is the part of the week that matters most to anyone running a stake pool.
Two proposals expire on September 1 (epoch 653), and both carry a 51% SPO requirement. Here is where they stand.
| Proposal | DRep support | SPO support | SPO votes |
|---|---|---|---|
Reduce minPoolCost to 75 ada + Plutus memory limits (Part 2) |
32.43% (68 votes) | 2.63% | 24 yes, 1 no |
| Update Constitutional Committee 2026 | 30.12% | 1.71% | 9 yes, 1 no |
In #18 we wrote that “15 pools have voted.” For minPoolCost that number is now 25 — ten more. But the requirement is measured in delegated weight, not vote count, so 25 pools amount to 2.63%. The distance to 51% has barely closed.
The second item may matter more in practice. If the Constitutional Committee renewal does not pass, the committee falls below its minimum size. Intersect’s weekly update states that ratification is needed to prevent the committee dropping under the five-member requirement. SPO support there is 1.71%, from 10 pools.
The fixed-fee proposal would change the flat amount deducted from block rewards before distribution to delegators, from 170 ada to 75 ada. The smaller a pool’s total rewards, the larger a share that flat amount represents. In other words, the operators who feel this vote most directly are the smaller pools.
There are still more than two weeks to the deadline, so falling short today is not a failure. But if one week produced ten pools, the path to 51% in the remaining two weeks is not natural accumulation.
5. Four Proposals Arrived on the Constitution Portal
In #18 we listed this as a watch item: “The first CAP / CIS — whether actual proposals appear on the Constitutional Amendment Portal.”
They have. As of August 15, 2026, the portal shows four in consultation, zero ready, and zero done.
| Type | No. | Title |
|---|---|---|
| CAP | #1 | Require Milestone-Based Treasury Disbursement and Return of Undistributed Funds |
| CAP | #2 | Remove Article II section 7.4 |
| CAP | #3 | Evolve the Net Change Limit into a Strategic Spending Plan |
| CIS | #4 | The Bundling Gap in Constitutional Amendment |
A CAP proposes a change to the text of the Constitution itself. A CIS is a suggestion for improvement that does not carry a textual change, serving as an entry point for discussion.
Three of these four touch directly on the problems we have been following for the past three weeks.
CAP #1 would put treasury spending on a milestone basis and require the return of undistributed funds. What we have watched until now is the side where requests expire without votes. This proposal addresses the side after approval.
CAP #3 would develop the Net Change Limit into a strategic spending plan. As noted in #18, the information action Net Change Limit: Cardano Treasury (Epochs 613-713) expired in epoch 647 even though it carried no financial request. The very action that would set a bound on how much the treasury may net-decrease over a period could not gather votes. That question now re-enters from the constitutional side.
Then there is CIS #4, titled “The Bundling Gap in Constitutional Amendment.” It was filed today and already carries discussion on the portal.
What it addresses is the treatment of bundling multiple changes into a single proposal. Recall the point we made in issue #17: the minPoolCost reduction was bundled with the Plutus memory limit increase (Part 2). minPoolCost is not among the security-relevant parameters in CIP-1694, but maxBlockExecutionUnits is. As a result of the bundling, a 51% SPO consent requirement entered the ratification conditions.
So the very origin of the requirement that currently sits at 2.63% is now itself on the agenda as a matter of constitutional amendment.
By discussion volume the portal shows CAP #3 leading with 9, then CIS #4 with 5, and CAP #1 and CAP #2 with 2 each.
The portal works by signing in with any Cardano wallet; no account registration is required. You start by reading the Constitution and highlighting the text you want to change.
6. What Dijkstra Brings to SPOs Next
From Intersect’s weekly update #124 (August 14), here are the parts that bear directly on operations.
Dijkstra’s scope and target dates are unchanged. What goes into the first phase was settled with the presentation to the Technical Steering Committee on July 29, and the work has moved to translating that scope into an ecosystem-wide timeline.
Three dates are stated:
- Amaru node mainnet block production is targeted for November 2026
- Phase 1 (Nested Transactions and Linear Leios) to mainnet by the end of 2026
- Phase 2 (Peras) activated via an intra-era hard fork in Q2 2027
And here is the line operators should note. After Dijkstra, Leios enters a bootstrapping phase, and SPOs will be required to register BLS keys there.
No operational procedure has been published yet. But unlike a vote, this is not the kind of thing where a sufficient fraction moving by the deadline is enough. To keep producing blocks, each pool works through it individually. The Hard Fork Working Group meets Tuesdays at 15:00 UTC.
The same update reports that MLabs completed three items: Cardano.nix Milestone 5, Plutarch Milestone 5, and Milestone 6 of the research toward tooling for elliptic curves. The 1.16 million ada for MLabs whose execution we confirmed in issue #17 (ratified in epoch 645, enacted in 646) corresponds to this body of delivery.
7. The Committee Has Voted Against the Name of Protocol 12
One more vote of a different character is in motion.
Name the Protocol Version 12 hard fork "von Bergen" is an information action expiring at epoch 651 (August 22). Being an information action, it carries no financial request.
The figures: DRep support 52.49% (98 votes), voted-no 0.0% (1 vote), not voted 43.13%, abstain 18 votes. And the Constitutional Committee stands at 0 yes, 2 no, 4 not voted.
Information actions carry a 100% threshold, which in practice means they do not pass without unanimity. So a proposal that only assigns a name currently carries two committee votes against it. The reasons cannot be read from the vote itself.
8. Midnight Watch
What could be confirmed around Midnight this week.
A statement that it has entered a beta phase came on August 14 from Charles Hoskinson, referring to new capability. This is reporting of a statement; an official Midnight document announcing it could not be confirmed this week.
VIA went live on both the Cardano and Midnight mainnets. Per VIA Labs’ announcement, this was August 13.
USDM became deployable on Midnight, with Hoskinson responding to it. Again, this is reporting of a statement.
NIGHT closed the week at −9.61% ($0.01910631 → $0.0172702), in the same direction as ADA’s −11.42%.
In #18 we noted that the MIP (Midnight Improvement Proposal) process had been published. A first proposal following that process could not be confirmed this week.
9. Risk Dimensions
Four points worth holding from this week’s material.
1. “The silence broke” is a single observation. The 80.1% is a fact, but one case cannot establish whether it marks a change in the treasury overall or something specific to this proposal. The two SPO-side items did not move in the same week, so at minimum this is not “participation rose across the ecosystem.”
2. Thirty-nine no votes remain. The opposition share fell because support rose, but the opposition itself did not disappear. Since the approved amount is the largest so far, the execution process will be watched more widely than in earlier cases.
3. SPO support at 2.63% is not on pace with the calendar. Ten pools were added in a week to reach 2.63%. That rate does not arrive at 51%. If it does not, minPoolCost stays as is and the Constitutional Committee’s minimum-size problem remains.
4. Price and treasury pointed opposite ways this week. ADA fell 11.4% while the treasury passed its largest item. In #18 a rise coincided with failure; this week a decline coincided with success. For two weeks running, the two have not pointed the same way.
10. Next Week’s Watch List
minPoolCostSPO votes — whether they move from 2.63%. Deadline epoch 653 (September 1).- Constitutional Committee renewal — whether SPO support moves from 1.71%, and whether it resolves before the committee drops below minimum size.
- The two committee votes against
von Bergen— whether reasons are given. Deadline epoch 651 (August 22). - Discussion on CIS #4 — whether the treatment of bundling advances to concrete textual language.
- AlphaGrowth execution — the epoch in which the 120 million ada is actually disbursed, and the shape of reporting that follows.
- Leios BLS keys — whether a registration procedure is published for SPOs.
Closing
At the end of #18 we wrote: “The next votes to be counted are our own.” The count came back at 15.
This week the numbers split in two directions.
One is that the silence broke. The 120 million ada request we had described as carrying unusually heavy opposition gathered 80.1% without a single ada being cut, and not-voted fell to 8.5%. The reading we offered in #18 — that cutting the amount does not reduce silence — is confirmed from its reverse side. It passed without being cut, so what was binding was never the amount.
The other is that SPO votes have not moved. Twenty-five pools have voted, amounting to 2.63%. On the Constitutional Committee renewal it is ten pools and 1.71%. Both expire September 1, and one of them bears directly on the committee’s membership. The fixed-fee floor weighs most on pools with the smallest rewards, and the vote that sets that floor currently sits at 2.63%.
Meanwhile the Constitution portal has begun to fill with proposals addressing that structure itself: tying treasury spending to milestones, growing the net-change bound into a plan, and questioning the gap in how bundled amendments are handled. That last one was filed today, and it speaks directly to the point we have followed since issue #17 — that a 51% SPO requirement entered the conditions as a result of bundling.
We now know that silence can break. A single week produced a 37-point move.
What remains is whether it breaks on the SPO side too. The deadline is September 1.
Sources
- Intersect MBO — Intersect weekly update #124 (2026-08-14; Dijkstra scope and timeline, Amaru November target, SPO BLS key registration during Leios bootstrapping, vote status of the two September 1 items, three MLabs milestones): https://www.intersectmbo.org/news/intersect-weekly-update-124-aug-14-2026
- Intersect MBO — The Constitutional Amendment Portal: giving Cardano’s Constitution a way to grow (2026-08-07; CAP / CIS mechanics and lifecycle): https://www.intersectmbo.org/news/the-constitutional-amendment-portal-giving-cardanos-constitution-a-way-to-grow
- Intersect Constitutional Amendment Portal (as of 2026-08-15: four in consultation, titles and discussion counts for CAP #1–#3 and CIS #4): https://cap.intersectmbo.org
- Intersect MBO — Constitutional Amendment Portal update (2026-08-13): https://x.com/IntersectMBO/status/2087811588447224211
- Cardano governance actions (AlphaGrowth ratified,
minPoolCost, Constitutional Committee renewal,von Bergen): https://gov.tools/governance_actions - Koios REST API —
proposal_list/proposal_votes/proposal_voting_summary(proposal status, itemised and aggregated votes, SPO vote counts): https://api.koios.rest/api/v1 - AdaStat (DRep support on a delegated-weight basis and the ratification denominator): https://adastat.net
- VIA Labs — VIA live on both Cardano and Midnight mainnets (2026-08-13): https://x.com/VIA_Labs/status/2088335112727576667
- CoinGecko (ADA / BTC / ETH / NIGHT values for August 8 and August 15): https://www.coingecko.com
Transparency Note
Some material surfaced this week but did not enter the body. The reasons are recorded here.
- Wanchain follow-up: listed on #18’s watch list, but no new announcement after the deadline could be confirmed this week. We do not write an absence of confirmation as a follow-up.
- Midnight beta phase and USDM deployment: both confirmable only as reporting of statements, not as official Midnight documentation. The body marks them as such.
- A third submission from the two rejected proposals (Scalus, dOSPO/OMF): listed on #18’s watch list, but no third filing appears in this week’s proposal list.
- Why AlphaGrowth support moved 37 points in a week: the voting record shows that votes increased, but what caused it lies outside the record. We present the facts without speculating.
- Why two Constitutional Committee members voted against
von Bergen: not readable from the vote itself, so only the fact is stated.