Skip to content

Canton: Q3 2026

18 Nov 2026
Download PDFPDF

Canton was built for regulated markets. On Oct 28, DTCC's tokenization service went live on it, for securities at a depository that holds more than $114T. Q3's burn came from trading venues instead: Temple, Kora, and Cantex paid 56.5% of the Canton Coin (CC) burned, while wallets paid 19.9%. Kora is a private spot exchange that opened on Canton in March. The dedicated synchronizers that institutions run for their own traffic burned none.

Holders get no cash distribution. Traffic pays them only through burn, which takes CC out of supply. Trading venues and wallets pay for most of the burn now. In Q3, round mints paid apps $1.06 of rewards for each $1 burned. The institutional business adds burn once universal burn makes those dedicated synchronizers pay in CC, from an estimated Q1 2027.

Stacked quarterly shares of CC burned for Global Synchronizer traffic by payer, Q4 2025 to Q3 2026. In Q3, Temple paid 44.3%, Kora 8.0% and Cantex 4.2%, 56.5% together; Temple paid 21.6% in Q2. Yiksi paid 5.4% after buying none in Q2. C8 Wallet paid 9.7%, Console Wallet 6.6% and Send's Canton Wallet 2.8%; Loop's 0.9% sits in All other payers, so wallets paid 19.9%. All other payers made up the rest.

Key Takeaways

Traffic on Canton burns CC, while rounds mint new CC about every ten minutes and pay it out as rewards to apps, validators, and Super Validators. From Nov 5, a cap on app rewards cuts the CC that rounds mint for each dollar of burn.

  • CC burned now covers most of what the rounds mint: 68.9% in Q3, against 11.0% a year earlier. More CC burned and a shrinking mint schedule each drove about half of the rise over the year.
  • Temple, Kora, and Cantex paid 56.5% of Q3's CC burned, and app rewards returned most of what the largest buyers spent. Temple's traffic cost it 17% of what it burned, net of app and validator rewards.
  • On Q3 traffic and price, the Nov 5 cap lifts burn's share of round mints to 74.7% to 84.6%. The first five capped days ran at 77.9%, against 72.9% had the cap held nothing back.
  • Institutions become the next source of burn once universal burn starts, estimated for Q1 2027. DTCC made Canton the one public network in its Oct 28 tokenization launch.
  • Market cap ran 8.0x annualized burn in dollars on Sep 30, near the top of its range since April. At 44.6 GB a day, burn matches round mints and the multiple falls to about 5.1x.
Quarterly line from Q3 2024 to Q3 2026. Burn covers under 1% of round mints through Q4 2024, 11.0% in Q3 2025 and 68.9% in Q3 2026, rising every quarter since early 2025.

Burn's share of round mints has risen in every quarter since early 2025. The step from Q2 to Q3 came from a lower CC price, not more traffic.

Who Pays Whom on the Global Synchronizer

Like Ethereum, Canton is a public network whose fees are paid in its own coin. Ethereum shows every transaction to every node, while Canton shows each party only its own part of one. The Global Synchronizer is the shared network that orders those transactions. From Oct 12 to Nov 11, 76.7% of updates on it were private. That privacy is Canton's edge with regulated firms.

The nearest business is a settlement utility such as DTC, but Canton charges by the megabyte and burns the charge instead of collecting it. Validators buy traffic with CC. Rounds then mint new CC to the firms doing the work. With no distribution, buyback, or staking yield, holders' case rests on burn outrunning round mints. One vote on rewards can cut those mints, as the Nov 5 cap did.

A vote fixes the price of traffic in dollars

Traffic beyond each validator's free base rate costs $60 per MB, about a dollar for a median transaction of 16.7 KB. The 13 nodes that run the Global Synchronizer set that price by vote. It read $60.0 per MB in all 22 config votes that priced it from January through November. The payer burns CC at the network's own price, leaving the fee fixed in dollars while the coins float.

Diagram. Apps and users pay validators fees. Validators buy Global Synchronizer traffic at $60 per MB by burning CC; the 1,369M CC burned in Q3 went to no party. Rounds minted 1,986M CC under a curve capped at 10B a year: 1,443M to featured apps, 271M to validators and 271M to Super Validators.

Hedera is the nearest design: it also prices fees in dollars under a council of large firms, but it pays those fees to its nodes and treasury instead of burning them. Tron also sets fee rates by vote and burns its fees, but it prices them in TRX, not dollars. A bank's operations desk gets a settlement cost it can budget. For a holder, a lower CC price burns more coins per transaction while each coin is worth less.

A lower CC price burns more coins for the same traffic

Burn in dollars is traffic times the voted price. CC burned is those dollars over the CC price. The two can move apart; in Q3 they moved in opposite directions.

CC burned of 1.369B splits into burn dollars of $157.5M over a CC price of $0.115. Burn dollars split into traffic of 2.63M MB at a voted $60 per MB.

Rounds mint CC about every ten minutes under a curve capped at 10B CC a year in this phase. Each round pays apps, validators, and Super Validators. Each app claims a share of the app reward pool in proportion to the burn it reports. Apps as a group claim no more than that pool. More burn therefore mints more CC until apps claim the whole app pool. Rewards nobody claims collect in the unclaimed pool. In Q3 round mints paid $1.06 per $1 burned to apps, $0.20 to validators, and $0.20 to Super Validators.

App rewards work like an exchange's maker rebate paid in its own shares. The app brings the flow while the network burns the fee.

Dedicated traffic is the next source of burn

A dedicated synchronizer is a second lane on Canton's MainNet, beside the Global Synchronizer: same assets, its own operator. Its traffic "doesn't burn anything"; its operators have paid for it through license fees to Digital Asset. Hong Kong's HK$20B of digital green bonds ran on HSBC Orion with their own dedicated synchronizer.

Diagram. Before universal burn, Global Synchronizer traffic burns CC at $60 per MB and earns rewards, while dedicated synchronizers burn no CC, earn no rewards and pay license fees to Digital Asset. Under universal burn, estimated for Q1 2027, the Global Synchronizer keeps $60 per MB with no discounts. Dedicated synchronizers then burn CC at the list price less voted discounts, with operators reporting their traffic.

Universal burn targets the split between trading traffic that pays in CC and institutional traffic that pays Digital Asset instead. CIP-0129 passed on Oct 23 with all 13 nodes in favor. It will make every production synchronizer burn CC for its traffic, starting from the Global Synchronizer's list price.

The Global Synchronizer keeps that price at every volume. Dedicated synchronizers will get a volume curve and commitment discounts. The 13 nodes vote interim discounts case by case until a second CIP sets those terms. That CIP also sizes the CC collateral an operator posts for a commitment of a year or more. Digital Asset will move each existing operator from license fees to burn, while new operators get a permissionless route. For holders, that adds burn on traffic that pays nothing into CC now.

Network app rewards ran above burn, yet each of the six largest traffic buyers earned less in app rewards than it burned. Apps that bought almost no traffic took 34.2% of app rewards. Another 7.3% went to apps whose traffic purchases are not on record, CBTC among them. Most apps in both groups are assets and bridges. Other parties buy the traffic that pays them.

A general app is one whose transactions move no featured asset and earn the base reward weight. Before the cap, a general app could earn up to $1.50 of app rewards for each $1 its traffic burned. Under that ceiling, traffic bought only for rewards could pay for itself while the app reward pool had room. In those rounds, holders funded the difference through new supply. From Nov 5 a general app earns less than it burns, which turns traffic bought only for rewards into a loss.

Top buyers get back under 1 CC of app rewards per CC burned

A featured app earns from the app reward pool each round. Featured App Guidance v2.0b took effect on Nov 5 and asks a basic featured app to lock 5M CC. Every app earns 1.0 reward weight per dollar its traffic burns. A composed transaction moves a featured asset, such as a stablecoin or bridged bitcoin, inside a featured wallet, venue, or bridge. It earns 1.10 reward weight with one such asset and 1.20 with two or more. That added weight goes to the assets. Guidance v2.0b bars an asset-moving app from paying its users a net reward for their activity.

Apps set these weights themselves when they report each transaction's burn, until the network takes over that measurement in February 2027. The Featured App Accountability Committee checks the reward weight each app reports against the burn it paid and can pause without further warning any app that reports more than 103% of the weight its burn earns.

Treemap of 847.1M CC of app rewards from Aug 8 to Sep 30, 2026, by app. Trading venues 37.1%: Temple 222.7M, Kora 58.7M, Cantex 16.3M, Tradecraft 16.5M as other venues. Assets and bridges 27.3%: USDC bridge 62.0M, CBTC 59.3M, Ember 59.2M, cETH 50.4M. Wallets 9.6%: C8 Wallet 49.3M, and 32.2M across others, led by Console Wallet 19.9M. Other apps: Yiksi 45.7M and 174.8M across the rest.

Over the 54 days from Aug 8 to Sep 30, network app rewards came to 847.1M CC. Temple took 26.3% of those rewards, ahead of the USDC bridge, CBTC, Ember, Kora, and cETH. Cantor8 ranks second as an organization, led by C8 Wallet. Trading venues lead on traffic they buy, and settlement assets on traffic other parties buy.

Dumbbell of six organizations. Each earned less in app rewards than it bought in traffic: Temple, Kora, Cantor8, Yiksi, PixelPlex and Cantex.

Temple, Kora, and Cantex run trading venues, while Cantor8 and PixelPlex run the C8 and Console wallets. Yiksi joined them as a top buyer in Q3 with 5.4% of the CC burned, after buying none in Q2. It runs a regulated crypto-fiat exchange for the Horn of Africa. Those six each earned 0.63 to 0.89 CC of app rewards per CC of traffic bought. Across the network, apps earned 1.04 CC of rewards per CC burned over the same 54 days, close to the quarter's rate measured in dollars. A buyer that runs its own validator also collects validator rewards on top.

App and validator rewards on a general app stop at $0.99 per $1 burned

Canton changed the asset-weight formula twice. Asset weight counted in dollars of burn, at most $1 per asset in each transaction. That limit pushed apps to tune settlement costs around it and worked against batch settlement. A batch of 25 settlements that moved two assets in one transaction earned rewards on $2 of asset weight, while the same 25 sent one at a time earned on $2 each, $50 in all. On Sep 18 the Tokenomics Committee approved removing the limit, on the Featured App Accountability Committee's advice. On Nov 5, Guidance v2.0b replaced the dollar measure itself with fixed composed-transaction weights, which lowered asset rewards.

Two panes. App rewards run from $1.00 to $1.25 against a new $0.79 cap. Validator rewards fall to $0.20 by May and stay there.

The reward cap followed in a config vote that closed Nov 4 with no node against and took effect with Guidance v2.0b on Nov 5. It cut the most an app earns from $1.50 to $0.79 per unit of reward weight. With validator rewards already capped, a general app's traffic now returns at most $0.99 of app and validator rewards per $1 burned, against $1.70 before.

In Q3 app and validator rewards together ran at $1.26 per $1 burned, above every new ceiling. At 1.0 weight, the cap would have cut Q3 app rewards by 25.4%.

CIP-0104 goes further and pays each app on the traffic its transactions actually burn, as the network measures it rather than as the apps label it. Its first activation vote failed on Aug 17, 7 against and 2 for. The amended CIP passed Oct 27, runs a dry run in December, and switches in February 2027, subject to Tokenomics Committee approval. From then on, rewards to apps with no recorded traffic depend on the weight the network itself assigns.

The featured app list turns over by vote: on May 20, CIP-0116 tied featured status to CC locked per party. The Tokenomics Committee then moved to pause 45 apps under that rule in late June.

Paired bars by quarter. Grants run 19, 74, 63, 63 and 40. Revocations run 2, 9, 20, 83 and 26. In Q3, 34 grants and 17 revocations reached nine accepted.

The rule binds incumbents too: a July vote moved to pause Tradeweb's treasury party. Nexode came back in September after review, and on Oct 29 the Tokenomics Committee recommended revoking Grofty Wallet, Yapper Agent, and Ibex for abusing free traffic. Of the 137 apps the lock rule tracks, 38 fell short on Nov 11. Of those, 20 still held featured status.

Super Validators

Thirteen nodes run the Global Synchronizer and cast its votes, one each; none of the 13 voted against the cap. Nearly every organization that receives Super Validator rewards locks most of what it's earned.

Super Validators are the organizations that hold seat weight. Seat weight sets each one's share of the tranche that rounds pay Super Validators. It's held in three seat types. An operator seat pays the party that runs its node. A hosted seat pays a party other than the one that runs the node. That party can belong to the same organization. Tradeweb, the Canton Foundation, and C7 run nodes and take their own rewards this way. An escrow seat holds weight for a partner that still owes milestones. Escrow seats carry many of the institutional names and held the most seat weight on Nov 5, yet their share mints nothing each round.

Escrow seats held the most seat weight on Nov 5

An escrow seat works like a performance-vesting grant: the seat holds weight while its rewards go to the unclaimed pool. A vote releases CC once milestones clear. The network pays the partner in CC, not cash, for the business it brings to Canton.

Treemap of 404.45 units of seat weight on Nov 5, 2026. Escrow seats 43.9%: DTCC, Nasdaq, Visa, LSEG, Marex, a16z crypto and USDT0 at 10 each, Apollo 7, other escrow seats 100.7. Hosted seats 38.6%: Broadridge, the Canton Foundation and Tradeweb at 10, Chainlink 9, Canton Strategic 8.4, C7 7.4, other hosted seats 101.4. Operator seats 17.4%: Digital Asset 31.9, Cumberland 27.2, SV Nodeops 10.3.

Escrow seats outweigh operator seats more than twofold, with DTCC, Nasdaq, Visa, LSEG, Marex, a16z crypto, USDT0, Apollo, and Franklin Templeton among their 41 beneficiaries. Hosted seats pay Broadridge, Tradeweb, the Canton Foundation, Chainlink, Canton Strategic, C7, and Circle. Six Super Validator seats won approval in Q3: TreasurySpring, Shinhan Asset Management, a16z crypto, LSEG, Paxos, and Marex. Three took weight after the quarter as escrow seats: Marex on Oct 7, LSEG on Oct 10, and a16z on Oct 28.

Nomura won a seat at seat weight 10 on Nov 10 under CIP-0130. Paxos's weight passed on Nov 11, while a Goldman Sachs seat waits in an open draft.

Before escrow seats take their share, the Super Validator tranche runs about 466M CC a quarter. Escrow seats held about 44% of seat weight on Nov 5. Votes on Nov 16 and 17 cut escrow seats to about 41% of seat weight. Rounds now send the escrow seats' share of the tranche, about 189M CC a quarter, to the unclaimed pool. That is about twice what votes minted from the pool in Q3.

Two founding Super Validators show how far the curve has stepped down. Tradeweb booked $1.4M for Super Validator and validator work in the September quarter, against $2.3M a year earlier, while Broadridge booked $1.4M against $3.7M. Scheduled changes in the minting curve and in Super Validator rewards drove most of that fall. More seat weight now shares the Super Validator tranche too. A CC price about twice its year-ago level kept the fall in dollars smaller than the fall in CC.

Bullet rows of Canton node revenue in the September quarter: Tradeweb $1.4M against $2.3M a year earlier, Broadridge $1.4M against $3.7M.

Both filers kept their seats while revenue per seat fell by more than a third. A seat's dollar income now leans on the CC price, since more seat weight shares a tranche that stays flat in CC between curve steps.

42 of 44 reward recipients held the top lock tier on Nov 5

To keep full seat weight under CIP-0105, a Super Validator locks 70% of lifetime rewards, a bar that steps down to 55% over three years. The requirement itself ends 30 days after the next issuance step down, forecast for late summer 2029.

Bullet rows for all 44 organizations and the four largest, each against the 70% Tier 1 line of CC earned.

On Oct 4, the 44 organizations that receive Super Validator rewards had earned 25.7B CC, about 64% of supply. On Nov 5, 42 of them held the top lock tier. Until CIP-0127 ships, a lock is a balance in a disclosed wallet that its owner commits to hold. Listed filers back that commitment in their filings, with Tradeweb and Broadridge each reporting 1.3B CC restricted from transfer at Sep 30. CIP-0127 passed on Oct 23 with no target date, moving enforcement from the Canton Foundation to the network itself.

All disclosed lock wallets held 18.35B CC on Nov 5, 17.6B of it among the top 100 holders. The total read 17.00B on Nov 11, after SV Nodeops moved 1.35B CC to a new party it planned to name as its final lock party.

Thirteen nodes decide, one vote each

Eleven organizations hold the 13 nodes, since Digital Asset and Cumberland run two each. Each node casts one vote whatever seat weight it hosts. CIP-0129 lets a 2/3 vote change any of its pricing parameters. Twelve of the 13 cast votes in Q3. C7's node cast none.

Grid of ten organizations by five action types, as a share of votes cast. Mint votes read 100%, while config and weight votes split most.

Mint votes were unanimous among nodes that voted, with the Tokenomics Committee checking each milestone before the vote. Config changes and seat-weight votes split them most: a node can back a partner's milestone and still resist a change to the rules. Daml is Canton's smart-contract language. A Daml models upgrade failed on Jul 13 with 8 nodes against and passed a week later. On Sep 18, nodes held TreasurySpring to the seat weight of 4 that CIP-0122 set.

Above the nodes, the Canton Foundation's new framework of Oct 22 sets a maximum of 25 board seats across five stakeholder groups, with four-year terms and staggered rotation.

The Unclaimed Pool

Votes and the Development Fund mint CC outside the rounds. Only votes draw on the unclaimed pool. Rewards nobody claims collect in that pool, along with every escrow seat's share. They wait there until the 13 nodes mint them by vote.

Votes minted 96.2M CC in Q3

The unclaimed pool's largest use was a one-off: CIP-0067 sent historical unclaimed rewards to the Global Synchronizer Foundation. Nodes accepted 1.62B CC for it on Nov 5, 2025, 12 in favor and 1 not voting. Since CIP-0067 excludes future rewards and milestone rewards, Q3's mints rest on different terms. Under escrow terms such as CIP-0124's, a partner proves the milestone and its calculation, the Tokenomics Committee agrees, and two thirds of nodes assign the CC.

Ranked bars of the 96.2M CC that Super Validator votes minted in Q3 2026, by share of that total: Zenith 26.4%, LayerZero 20.4%, Chainlink 20.0%, Wormhole 15.7%, Canton Strategic 6.7%, Hex Trust 6.3%, Blockdaemon 3.1% and QCP 1.5%. The four largest took 82.5%.

Zenith's execution layer took the largest mint, 25.4M CC. Bridges, oracles, and custody partners took most of the rest, while Canton Strategic Holdings took 6.4M. For Wormhole, two separate entities confirmed the value before the vote. Vote mints follow from Canton's design: the unclaimed pool keeps filling while escrow seats hold that much seat weight.

Mints outside the rounds explain the extra supply

Votes and the Development Fund minted 163.5M CC outside the rounds in Q3, 21.1% of Q3's live supply rise. Round mints net of burn made up the rest, adding CC every round.

Waterfall. Round mints of 1,985.7M less burn of 1,369.0M give 616.7M. Mints by vote add 96.2M and the Development Fund adds 67.35M. The four lines sum to 780.2M, and a residual of -4.6M brings them to the live rise of 775.6M.

In Q3 the Development Fund paid 67.3M CC of that total to grantees, minting each payment when it went out.

Two stepped lines rise together across the quarter. One is live supply less round mints less burn, reaching 159M CC. The other is the running sum of mints by vote and Development Fund mints, reaching 164M.

Mints outside the rounds arrive in lumps, one vote or one fund payment at a time. Under current escrow terms, a vote releases CC from the unclaimed pool only after a partner clears a milestone, which keeps their timing irregular.

The CC Token

CC is a burn-and-mint token: traffic burns it at a dollar price, while rounds mint it to the apps and operators that carry the traffic. Like ETH and TRX, it burns its fees; unlike them, it fixes its fee in dollars and pays no staking yield.

Disclosed lock wallets hold nearly half of CC's supply; the rest floats. Because circulating supply equals total supply, locked CC counts in market cap. Canton Strategic Holdings, Tradeweb, Broadridge, and Circle report CC on their balance sheets.

Locked balances in the top 100 came to 44.1% of supply on Nov 5

On Nov 5 the top 100 of Canton's 199,931 holders held 34.1B CC, 85% of supply. Disclosed lock wallets among them held 17.6B CC, 44.1% of supply. The rest of the top 100 held 16.4B. Holders below the top 100 held 5.9B. All of the locked CC still counts as circulating.

Treemap of 40.0B CC of supply on Nov 5, 2026. Disclosed lock wallets in the top 100 17.6B: Digital Asset 3.9B, Cumberland 3.28B, Canton Foundation 1.4B, SV Nodeops 1.35B, Broadridge 1.32B, Tradeweb 1.30B, Liberty City 1.30B, other lock wallets 3.77B. Other top-100 wallets 16.4B. Below the top 100 5.9B.

One owner can hold many wallets: Digital Asset's disclosed lock wallets alone sum to 3.9B CC, about a tenth of supply.

Tradeweb earned its 1.6B CC as a Super Validator. At Sep 30 those coins carried $18.3M of cost against $207.5M of filed fair value. Funds hold CC too: the 21Shares Canton Network ETF has traded on Nasdaq since May 7. Grayscale has filed for a Canton ETF, while the performance benchmark of a T. Rowe Price active crypto ETF weights CC at 1.60%.

The Development Fund refills from every round

Each round sets aside 5% of its issuance for the Development Fund under CIP-0082, since Canton has no large premine to draw on. That CC mints only when the Fund pays a grantee. Round mints leave it out. Issuance pays builders and keeps the core software funded without a treasury sale. Since Apr 28, 156 mints have paid out 123.21M CC, while the Development Fund held 305.3M CC available on Nov 11, before new accrual.

One bullet row. The Development Fund held 305.3M CC available on Nov 11, against marks at one to four quarters of Q3's 67.3M CC of payments.

By Aug 24, the Development Fund had committed 290.0M CC to 42 projects. Digital Asset's parties had received 57.2M CC by then, 85% of everything the Fund minted, for maintenance of Canton and its Splice node software, BFT ordering, and traffic-based rewards. BitSafe's 4.0M on Jul 9 was the largest payment to anyone else. Canton's creator takes its pay for that work in CC.

Canton Strategic Holdings (CNTN) is a listed route to the same token. It held 3.83B CC worth $482.1M at Sep 30, in its 10-Q for that quarter; at the Nov 10 close those coins were worth about $343M, 29% less. Its active seat weight rose from 0.5 units at Jun 30 to 8.4 on Nov 5. Among direct holders, Circle kept 379.9M CC on its own balance sheet.

The Canton Ecosystem

Canton's ecosystem has two sides: trading venues and wallets buy most of the traffic on the Global Synchronizer, while assets, bridges, DTCC's tokenization service, and Treasury repo carry the institutional work. Treasury repo is the largest by notional. The institutional programs' Global Synchronizer traffic burns CC under no separate label, while their dedicated synchronizers burn none.

Trading venues

DEX volume on Canton reached $1.0B in Q3, 3.4x Q2's $291.8M, led by Temple's $795.1M. Temple's pace stepped up late in the quarter: it traded about $6.8M a day from Jul 1 to Sep 21, then about $25.7M a day over the last nine days of September. On Sep 16 it launched Virtual Orderbooks, which lets institutions run their own venues.

Rocky is a spot and perpetuals venue on Canton whose trades run on a clock: exactly 96 a day from August 2025 through June 2026, then exactly 384 a day, one every 3.75 minutes, many of them the same size. A scheduled program made those trades, and their volume doesn't measure demand.

Stacked quarterly bars of DEX volume on Canton's tracked venues. Q3 2026: Temple $795.1M, Cantex $210.1M, other venues $0.7M, of $1.0B. Q2 2026: Temple $207.4M, Cantex $80.4M, other venues $4.0M, of $291.8M.

Temple kept most of its late-September pace after the quarter, at about $23.0M a day from Oct 1 to Nov 10, against $8.6M a day in Q3.

Rewards, not trading fees, fund most of Temple's traffic. From Aug 8 to Sep 30 it bought 355.7M CC of traffic and earned back 222.7M CC of app rewards, 0.63 CC for each CC burned, while its trading fees from Aug 2 to Sep 30 came to $153K. Net of those rewards and the 71.1M CC of validator rewards its own validator earned, its traffic cost it 61.9M CC over those weeks, 17% of what it burned. From Nov 5 the cap limits what each unit of reward weight earns, for Temple as for every featured app.

Wallets

Wallets paid 19.9% of the CC burned for traffic in Q3, down from 27.9% in Q2. They burned 77.0M CC less, which took 5.6 points off their share; the network's larger total took the rest. In late June, the Tokenomics Committee used CIP-0116 to pause two apps run by the operator of Send's Canton Wallet, though not the wallet itself. The wallet's traffic then halved in July. It burned 58.0M CC less in Q3. C8 Wallet fell to 9.7% of burn from 13.8%, while PixelPlex's Console Wallet rose to 6.6% from 4.4%.

Dumbbell of wallet shares of CC burned for traffic, Q2 to Q3 2026: all wallets 27.9% to 19.9%, C8 Wallet 13.8% to 9.7%, Canton Wallet 7.7% to 2.8%, Console Wallet 4.4% to 6.6%.

A wallet buys traffic for the parties it hosts. Since Sep 24, Local Traffic Management lets it see and charge that traffic per user.

Parties reached 866,882 at Sep 30, up 21.8% from 711,645 at Jun 30. One person or firm can hold many parties.

Line rises from January with a surge in March and reaches 866,882 at Sep 30. It steps down on Oct 27 when one validator offboards 117,804 parties.

On Oct 27, one validator behind PixelPlex's Console Wallet removed 117,804 parties. That one day erased most of Q3's party growth.

Assets and bridges

Featured assets earn on traffic other parties buy. USDCx held $7.09M of value on Nov 5 and CBTC $3.59M, at listed token prices. On Aug 25, World Liberty Financial began deploying its USD1 stablecoin natively on Canton, with BitGo as issuer, though none had been minted there by Nov 14.

Two bullet rows. The USDC bridge earned $6.73M of app rewards from Aug 8 to Sep 30, against $7.09M of USDCx on Canton on Nov 5. CBTC earned $6.43M, against $3.59M of CBTC.

From Aug 8 to Sep 30, the USDC bridge earned about $6.7M of app rewards, 95% of the USDCx on Canton five weeks later, on Nov 5. CBTC earned about $6.4M over the same days, 1.79x its Nov 5 balance. Guidance v2.0b cuts asset weight from Nov 5, while CIP-0104 ties rewards to traffic from February 2027.

Tokenized securities

DTCC's tokenization service turns eligible securities in DTC custody into tokens that can move as collateral within minutes, at a depository that holds more than $114T. Canton is the one public network in the launch, beside DTCC's private Besu network. On Jul 15 DTCC ran production trades across both with more than 30 firms, covering collateral pledges, securities lending, Treasury repo, and equities.

Timeline of programs from Tradeweb's Treasury trade, announced Jul 1, through DTCC's live date on Oct 28.

The service launched Oct 28 with 34 securities by Digital Asset's count and 20 DTC Participants signed, Bank of America, Citi, J.P. Morgan, and Morgan Stanley among them. Circle dates DTC assets on Arc for the second half of 2027, at least eight months after DTCC's Canton launch.

Repo and collateral

Treasury repo that settles atomically on Canton averages over $7.1T a month. Broadridge's Canton-built repo platform settled $7.5T in September, or $359B a trading day. MUFG began a proof of concept for Japanese government bond repo on Canton on Aug 13.

Monthly average daily repo value on Broadridge's repo platform: $280B in Aug 2025, $339B in Sep 2025, a peak of $385B in Oct 2025, and $359B in Sep 2026.

Repo notional measures value settled, while traffic measures megabytes bought. Daily repo volume held between $351B and $368B in each month of 2026, below the October 2025 peak of $385B. Broadridge expects contracted volumes to double, its DLX launch connects to DTCC's service, and its G7 extension widens the collateral.

Collateral trades ran on Canton through Q3 too. Vanguard and Wellington posted tokenized money market fund shares as collateral through Nasdaq Calypso in a pilot announced on Aug 24. In late August, Virtu, M1X Global, and Tradeweb completed an onchain repo against a Marshall Islands sovereign digital bond, settling both legs atomically in under 10 minutes.

Financials

Burn in dollars fell 17.6% in Q3, in step with traffic's 17.7% fall at a fixed dollar fee per MB. CC burned rose 9.2% because the burn-weighted CC price fell faster than traffic. Round mints held steady.

Round mints paid 72.7% of their CC to apps in Q3, against 42.1% a year earlier and 7.8% in Q3 2024, as the curve stepped down. The cap and Guidance v2.0b act on that share. Votes and the Development Fund minted another 163.5M CC outside the rounds, mostly to milestone partners and grantees.

Stacked quarterly shares of round mints by recipient. Apps rose from 7.8% in Q3 2024 and 42.1% in Q3 2025 to 72.7% in Q3 2026; Super Validators fell from 89.8% to 45.2% to 13.7%; validators held 13.7% in Q3 2026.

Validators and Super Validators each took 13.7% of round mints, which leaves the cap acting on an app tranche more than five times either of theirs. Validator rewards already had a cap, which left the app tranche for the cap that took effect on Nov 5.

Income statement

At each day's network price, round mints added $229.4M of CC to supply in Q3, while burn took back $157.5M. That difference is net round issuance at the network's own prices.

A lower CC price outweighed 221.6M CC of lost traffic burn

CC burned reached 1.37B in Q3 because a lower price made each dollar of traffic burn more coins. At Q2's price, less traffic would have taken 221.6M CC away.

Waterfall from 1254.2M CC in Q2 to 1369.0M in Q3. Less traffic takes 221.6M and a lower price adds 336.4M.

The burn-weighted CC price averaged $0.115 in Q3, 24.5% below Q2's $0.152. Every 10% off the CC price adds about 11% to CC burned for the same traffic. At Q2's price and with validator rewards moving with burn, Q3 burn would have covered 53.8% of round mints, against 64.4% in Q2.

Round mints per $1 burned fell to $1.46 from $9.24 a year earlier

Each dollar of burn still arrives with more than a dollar of new CC from the rounds, 73% of it in app rewards, which the cap targets. Round mints came to 1.99B CC in Q3, up 1.9% on Q2. In dollars they fell 11.2% over the year while burn in dollars rose 5.6-fold.

Statement of burn in dollars and round mints in dollars, with app, validator and Super Validator rewards, for Q3 2025, Q2 2026 and Q3 2026.

In CC, burn rose from 17.2% of round mints in Q4 2025 to 64.4% in Q2, before Q3's high. Over the year, CC burned rose 2.7x while round mints fell 57% as the curve stepped down. More burn explains about half of the rise in cover, while the shrinking mint schedule explains the rest.

Global Synchronizer traffic grew 5.7-fold in a year

Bought traffic reached 2.63M MB in Q3, 5.7x the year-ago quarter and 17.7% below Q2, averaging 28.5 GB a day against a February peak of 44.5 GB. Inside the quarter, the drop came in August: burn in dollars fell from $1.88M a day in July to $1.63M in August and held there in September. No single day drove the total: the ten heaviest burn days carried 13.2% of Q3's burn, against 10.9% for ten average days.

Quarterly bars: 0.46M MB in Q3 2025, 1.24M in Q4, a peak of 3.22M in Q1 2026, 3.19M in Q2 and 2.63M in Q3 2026, 5.7x the year-ago quarter.

Temple's traffic rose while the other payers' combined traffic fell.

Dumbbell of burn in dollars, Q2 to Q3 2026: Temple rose from $41.7M to $69.7M, up 67.0%; all other payers fell from $149.5M to $87.9M, down 41.2%; the network fell from $191.3M to $157.5M, down 17.6%.

The fall outside the six largest payers started before the quarter. Payers outside the six largest burned 8.8M CC a day in April, 4.5M in June, and 2.0M in September.

Temple paid close to half of Q3's burn

Temple's share of the CC burned doubled from Q2's 21.6%. Burn now rests on fewer payers: Q3's burn was as concentrated as if 4.5 equal payers had paid all of it, against 11.5 in Q2. A 10% cut in Temple's traffic would now take 4.4% off total burn, twice the hit in Q2. Temple is also on DTCC's partner list and picked tZERO as its ATS in October.

Temple burned 336.2M CC more than in Q2, nearly three times the network's 114.8M CC rise in burn. Payers outside the six largest burned 322.5M CC less.

Waterfall from 1,254.2M CC burned for traffic in Q2 to 1,369.0M in Q3, ranked by contribution: Temple +336.2M, payers outside the six largest -322.5M, Yiksi +73.4M, Kora +69.0M, C8 Wallet -40.3M, Cantex -36.0M, Console Wallet +35.0M.

The six largest buyers paid 78.1% of the CC burned in Q3, against 50.4% for the same six in Q2. Lost featured status made up most of the smaller payers' fall. Q2 votes removed or paused featured status for some operators, and those operators burned 220.9M CC less, 68.5% of the 322.5M. Those votes include the late-June pauses under CIP-0116.

Payers paused or revoked in Q3 cut 24.7M CC more. Other featured apps, Super Validators, and large validators cut 104.6M while app rewards per $1 burned fell to $1.02 in August from $1.25 in June. Smaller validators added 27.8M.

Daily users averaged 109,155 in Q3

The Q3 average came to about an eighth of the parties at Sep 30. Daily active users averaged 144,603 in July, 82,877 in August, and 99,679 in September. They ran under 60,000 a day after Send's Canton Wallet went offline on Oct 22.

Daily line of active users from Jul 1 to Oct 31, 2026: 188,176 on Jul 1, a Q3 average of 109,155, 120,833 on Sep 30, and 36,110 on Oct 31 after a fall from Oct 22, the day Send's Canton Wallet went offline.

Of 1,562 validator licenses, 1,195 have ever earned rewards and 339 earned in Q3, about one license in five.

Market Data

From Jun 30 to Sep 30, CC's close fell 10.4% while BTC, ETH, and SOL rose 43% to 71%. Canton's traffic fell over the same quarter. The 775.6M CC net rise in live supply came to about $89M at the burn-weighted price, 7.3% of the quarter's spot trading volume. CC hit its quarter low on Aug 17, the day CIP-0104's first activation vote failed. That day CC fell 5.8% while BTC, ETH, and SOL rose 2.2% on average. That left CC 7.0% below what its link to them over the prior 60 days implies.

After the quarter, CC fell 29.2% from Sep 30 to Nov 10, in line with SOL's 29.5% and further than ETH's 23.6% and BTC's 11.8%. Since July, a 1% daily move in those three coins has come with about a 1% move in CC on average, against about a quarter of that in Q2.

CC closed Q3 at $0.126 and market cap at $5.02B

CC closed Sep 30 at $0.126, while market cap fell 8.6% in Q3 to $5.02B, a smaller fall than the price's 10.4%, because supply grew. With no maximum supply and every coin circulating, CC's fully diluted value is its market cap.

Four rebased lines across Q3 2026. CC ends at 89.6, down 10.4%, while BTC ends at 142.7, ETH at 171.0 and SOL at 160.5. Callouts mark CC's low on Aug 17, the day CIP-0104's activation vote failed, and the SEC's Innovation Exemption on Sep 17.

On its Q3 link to BTC, ETH, and SOL, CC would have risen about 60%; it fell instead. July and August carried more than 90% of that shortfall, September under a tenth. Realized volatility rose to 82% a year in Q3 from 53% in Q2.

CC traded $13.3M a day in Q3, with $2.97M of bids near the price on Nov 5

CC's spot trading volume averaged $13.3M a day in Q3, up from $11.7M in Q2. Perpetual futures traded about as much again, $1.22B over the quarter. Their open interest rose 39.5% to $25.6M at the Sep 30 close. Binance held about half of both, and Hyperliquid held 11% of open interest, the most of any onchain venue.

Dumbbell of CC perpetual open interest by venue from Jun 30 to Sep 30, 2026: Binance $8.49M to $12.65M; Bybit $4.26M to $6.88M; OKX $2.86M to $2.94M; Hyperliquid $2.27M to $2.71M. Binance traded 50% of Q3's $1.22B of perpetual volume.

Nineteen USD pairs held $2.97M of bids within 2% on Nov 5, led by Bybit, OKX, and Kraken. By Nov 11, those bids stood at $2.74M, while perpetual open interest stood at $23.7M, about 8.6 times as much.

Bullet rows of bids $2.97M and asks $3.06M across 19 pairs, Bybit and OKX bids, and Temple's CC/USDCx order book at $1.10M of bids, against about $0.35M a day that CIP-0105's first lock step can release.

On the supply side, CIP-0105's first scheduled step comes a year after its Mar 2 approval and lowers the top lock tier to 65% of lifetime rewards from 70%. On the 25.7B CC earned to Oct 4, that lets Super Validators unlock about 1.29B CC. Unlocks vest at 1/365.25 a day, about 3.5M CC or $0.35M a day at the Nov 5 price. Bids within 2% held 8.4 times that on Nov 5.

Outlook

Burn in dollars ran 8% below Q3's daily pace after the quarter. From Oct 1 to Nov 10, it ran at $1.57M a day, against $1.71M in Q3. CC burned covered 69.0% of round mints, just above Q3's level. Development Fund mints came to 28.4M CC in those weeks. Votes minted 4.0M more for Societe Generale's CIP-0115 milestone, which reached supply on Nov 6. Next quarter turns on traffic and the CC price, with reward weight setting how far the cap reaches.

The cap shows in its first five full days

From Nov 6 to Nov 10, the cap left 6.9M CC of app rewards unminted. App rewards fell 15.6% to $0.91 per $1 burned, from $1.08 over Oct 1 to Nov 4.

Waterfall from $1.08 of app rewards per $1 burned over Oct 1 to Nov 4 to $0.91 over Nov 6 to Nov 10: a lower CC price took $0.17, less traffic gave back $0.09, and the cap took $0.09.

A lower CC price took $0.17 and the cap $0.09, while less traffic gave $0.09 back. The cap held back 8.8% of the app rewards the uncapped formula would have paid. The lower price cut the dollar value of an app reward pool fixed in CC, while less traffic spread that pool over fewer dollars burned. The same window also carried a MainNet software upgrade on Nov 7 and new smart-contract models from Nov 10.

Since composed transactions earn added weight, app and validator rewards still ran at $1.10 per $1 burned, above the general-app ceiling. December is the first full month under the cap.

Daily line from Sep 1 to Nov 10, 2026. App and validator rewards ran between $1.13 and $1.43 per $1 of burn from Sep 1 to Nov 4, then between $1.05 and $1.16 from Nov 6 to Nov 10, after the cap took effect on Nov 5, marked on the line. Reference lines mark the new ceilings of $1.15 at 1.2 weight and $0.99 at 1.0 weight.

Guidance v2.0b's lower asset weight took effect with the cap on Nov 5. It moved app rewards from asset and bridge apps to the payers that buy traffic. Against Oct 1 to Nov 4, daily rewards to the USDC bridge, Ember, cETH, and CBTC fell by 63.8% to 80.4% over Nov 6 to Nov 10, while Kora's, C8 Wallet's, and Temple's rose.

Dumbbell of app rewards a day before and after the cap: Temple 3.58M to 4.36M CC, Kora 1.96M to 3.46M, C8 Wallet 0.90M to 1.40M, USDC bridge 2.00M to 0.50M, Ember 1.66M to 0.42M, cETH 0.99M to 0.36M, CBTC 1.37M to 0.27M.

The cap itself sets one rate per unit of reward weight for every featured app. The move of rewards between apps came from the asset-weight cut, while the cap cut what each unit of reward weight claims and left the rest in the unclaimed pool.

Traffic near February's peak covers round mints

Burn's share of round mints turns on traffic and the CC price, with traffic the far more welcome driver. February's traffic at Q3's price brings burn to about 100% of round mints with no cap. A lower CC price gets part of the way too. Without the cap, Q3's traffic would cover 85.4% of round mints at the Nov 10 close. At $0.074, it would cover all of them.

Grid of burn to round mints across five traffic cases and five CC prices. Q3 pace at the Q3 average price reads 68.9%. February's peak at the Q3 price reads 100%.

Burn meets round mints at about 44.6 GB a day at Q3's price and reward pool. At that traffic, apps claim the whole app reward pool with or without the cap. The cap then no longer changes what rounds mint.

Reward weight sets how far the cap raises burn's share

On Q3's traffic and price, the cap raises burn's share of round mints from Q3's level to between 74.7% and 84.6% by cutting app-reward mints. The reward weight apps report sets where in that range it lands. At 1.0 weight it cuts net round issuance to 0.99B CC a year, from 2.45B at Q3's pace.

A composed transaction earns at most 1.2 reward weight. At that weight, burn reaches the low end of that range, as low as it goes while apps report under Guidance v2.0b. The first capped days paid $0.91 of app rewards per $1 burned, close to the $0.95 that 1.2 reward weight pays. At 1.0 weight, burn still reaches 80.0% of round mints if traffic falls by the same 25.4% the cap takes from app rewards.

Bullet rows: Q3 pace 68.9%, cap at 1.0 weight 84.6%, cap at 1.2 weight 74.7%, and cap at 1.0 weight with traffic down 25.4% at 80.0%, against the line where burn equals round mints.

Capped rewards fall with burn, which keeps net round issuance near 1B CC a year at 1.0 weight even if traffic falls. Extra traffic also mints more: a general app's rewards rise at the capped rate with each dollar its traffic burns, until the app pool runs out. Rewards the cap removes stay in the unclaimed pool, where votes have minted before. The cap also leaves Super Validator rewards out, which is why burn stops short of 100%.

Universal burn adds dedicated traffic to burn once its operators report it. Illustrative universal burn terms stack commitment, collateral, and loyalty discounts that take $60 per MB to about $6 after four years, before the volume curve. Q3's traffic fell about 16 GB a day short of breakeven. At about $6 per MB, dedicated traffic would need about 160 GB a day to make up the shortfall. That is about 9.6M median-size transactions a day, 5.6 times Q3's average Global Synchronizer traffic.

Rewards on that traffic are weighted by the CC actually burned, tying any discount to the rewards it earns. Until the app pool runs out, each new dollar of burn on a general app still mints close to a dollar of app and validator rewards. Early dedicated traffic lifts gross burn faster than it cuts net supply.

Valuation

Market cap stood at 8.0x annualized burn in dollars on Sep 30, near the top of its range since April and far below its fee peers. Burn retired 1.37B CC in Q3, about 3.4% of supply, against 2.15B CC minted by every route.

Since April, market cap has set the multiple's swings

From December to February the multiple fell from as high as 19.8x trailing 92-day burn as burn grew into the price. From April it ran between 5.0x on Aug 7 and 8.7x on Sep 27. Market cap drove most of the multiple's daily swings. In Q3, dollar burn also fell faster than market cap, which put the Sep 30 reading above 92% of the days since April.

Line of market cap over annualized trailing burn from December 2025 to October 2026. It fell from 19.8x on Jan 1 to March, then ran between 5.0x on Aug 7 and 8.7x on Sep 27, read 8.0x at the Sep 30 close and 6.6x on Oct 31.

Q3's burn in dollars annualizes to $625M, yet net supply still grew about 3.1B CC a year, or 7.8% of supply, after rounds, burn, votes, and the Development Fund.

What the price implies

Across its range since April, the Sep 30 market cap needs $574M to $995M of burn a year. Q3's pace lies near the low end of that burn range, the level the highest multiple since April implies. Burn from Oct 1 to Nov 10 ran at that floor. After a fall that matched SOL's, CC's Nov 10 close put market cap at $3.58B, about 6.2x that later burn pace.

At Tron's 104x, the Sep 30 market cap needs $48.3M of burn a year, or 7.7% of Q3's traffic at an annual pace. Solana's 278x brings that down to $18.1M. Tron also prices fees by vote, yet the market pays 13x Canton's multiple for Tron's fees. Tron's supply grew about 0.5% a year in Q3, a fraction of Canton's dilution.

Bullet rows of the burn a year the Sep 30 market cap implies: $994.9M at CC's 5.0x low since April and $573.6M at its 8.7x high, against a Q3 pace of $625.0M; $48.3M at Tron's multiple, $18.1M at Solana's and $2.1M at Ethereum's.

At 1.0 weight on Q3 traffic and price, the cap cuts net round issuance to about 2.5% of supply a year, from about 6.2% in Q3. With votes and Development Fund mints at Q3's pace, supply grows about 4.1% a year at 1.0 reward weight and about 6.3% at 1.2, both below Q3's total supply growth.

Comps

Canton ranks first of eight settlement chains on Q3 fees. With its voted price per MB unchanged since January, its fees moved with traffic alone. Peers' dollar fees move with activity and token prices, Tron's with voted rates too. Each peer matches one part of Canton's model: Ethereum is the reference settlement chain, Solana the high-throughput one, Tron prices its fees by vote, Stellar serves regulated payments, and Avalanche sells institutions their own subnets. Hedera shares Canton's dollar fee and council of firms. XRP Ledger burns every fee and sells cross-border payments to institutions.

Canton's fee lead rests on traffic bought at a set price

Canton's Q3 fees of $157.5M were about twice Tron's $77.8M. Canton took 47.2% of the eight chains' Q3 fees, against 6.4% a year earlier, when it ranked fourth of eight. Canton's fees aren't net of rewards: in Q3, apps and validators earned $198.1M of new CC against that burn, $40.6M more than the network burned. From Nov 5 the cap holds a general app under a dollar of app and validator rewards per $1 burned.

Bump chart of eight chains over five quarters, by fees. Canton rises from fourth of eight in Q3 2025 to first in Q3 2026, ahead of Tron, Solana and Ethereum.

The contest for institutional share runs elsewhere: Circle's chain Arc went live Sep 16 with DTCC and Visa among its validators. Solana released a DvP program on Oct 6. DTCC and Visa work on Canton too, leaving none of this exclusive; Canton's institutional edge rests on the DTCC date and its privacy model.

Canton records almost nine in ten dollars of represented real-world assets

A represented asset stays off chain, while a network records who holds it or how it settles. On Sep 17, Canton recorded 88.4% of represented real-world asset value across the networks, against 95.2% in December. All of Canton's value is Broadridge's repo platform.

Stacked shares of represented real-world asset value by network, one read a quarter from Dec 22, 2025 to Sep 17, 2026. Canton fell from 95.2% to 88.4% while Provenance rose from 3.6% to 6.1%. On Sep 17 Avalanche held 3.1% and XRP Ledger 1.1%. All other networks made up the rest.

Avalanche rose to 3.1% from 0.2%, and Provenance to 6.1% from 3.6%. Canton's lead rests on a single repo platform.

Canton's fees grew the most of the eight over the year

Canton's fees fell with traffic in Q3, while the fee per MB held. Tron's fees also fell when counted in TRX. Ethereum's fell when counted in ETH, though ETH itself rose in Q3.

Slope of fees for Canton, Tron, Solana and Ethereum from Q3 2025 to Q3 2026. Canton rose to $157.5M from $28.0M, while Tron fell to $77.8M, Solana to $62.9M and Ethereum to $34.6M. The four smaller chains in the peer table each took under $1M in Q3.

In Q3, Solana was the one large peer whose fees rose in its own coin, up 36.2% in SOL, while Avalanche's and Stellar's smaller fees rose in their coins too. Over the year, Canton rose from fourth to first because its fees grew 5.6-fold, while Tron, Solana, and Ethereum lost about half or more of theirs.

Burn over all new coins counts burn against burn plus the net rise in supply. On that basis, burn equaled 63.8% of all the CC that every route issued in Q3, against 65.1% of Tron's new TRX and 1.6% of Ethereum's new ETH. Solana burns half of each base fee; in Q3 that burn came to about 1.3% of the SOL it minted. XRP Ledger burns every fee too, but its fees are tiny beside the escrow releases that add to its circulating supply. Stellar and Hedera burn none: Stellar's fees go into a fee pool that no rule pays out, and Hedera pays its fees to nodes and its treasury.

Canton's multiple runs far below its fee peers'

Canton's multiple falls below Tron's 104x and Solana's 278x, while Ethereum and Avalanche run near 2,400x, Stellar and Hedera near 16,000x, and XRP Ledger above 500,000x. Each multiple divides Sep 30 market cap by annualized Q3 fees.

Peer table of eight chains with Q3 fees, change on Q2, change on Q3 2025, market cap and the multiple of annualized fees.

At about a dollar a transaction, Canton's price keeps out the small payments and trades that cheaper chains carry. Most peers' fee rates rise and fall with demand for block space. CIP-0129's volume discounts apply only to dedicated synchronizers, while traffic on the Global Synchronizer keeps paying the full price per megabyte at every volume, which leaves that barrier in place for public traffic.

Significant Updates

The quarter set up the changes, most of which landed after it. Q3 brought DTCC's production trades and six new Super Validator approvals, while October and November brought the universal burn vote, the cap, and DTCC's launch.

The SEC's Q3 exemption for tokenized stocks reached Canton on Nov 4

On Nov 4, Cumberland bought tokenized CNTN shares on TSV LLC, a venue under the SEC's Innovation Exemption of Sep 17. The trade settled atomically on Canton, seven weeks after the exemption opened that route for tokenized stocks.

Timeline of Q3 2026 events: Tradeweb Treasury trade Jul 1, DTCC production trades Jul 15, Digital Asset adds Shinhan and SC Ventures Jul 21, CIP-0104 activation vote fails Aug 17, Arc goes live Sep 16, SEC Innovation Exemption Sep 17, LSEG approved as a Super Validator Sep 22.

On Sep 30, Lloyds and Visa made public a round-the-clock cross-border settlement pilot; Lloyds ran its own Canton node.

The consensus move and the SEC letter came after the close

MainNet moved to the CantonBFT consensus protocol on Oct 10 after a delay for DevNet results, with almost no downtime across more than 800 validator nodes. The 13 Super Validator nodes also canceled the Splice 0.9.0 node software upgrade on Oct 19 over a query issue, rolling out the fix two days later.

On Oct 13, Derivnex launched the first OTC swap on Canton, with DRW and Marex as counterparties, adding a derivatives workflow to the collateral and repo set. Ripple Custody added Canton Coin and tokens on Canton's token standard in its Oct 5 release, which lets its clients hold and transfer CC through its API beside their other digital assets. On Oct 30, Uber and HSBC made public a pilot in which Uber bought tokenized money market funds with HSBC deposits and could sell part of them over a weekend while they kept earning.

Timeline of events after Q3 2026: MainNet moves to CantonBFT Oct 10, universal burn approved Oct 23, DTCC service live Oct 28, SEC staff letter on CC Nov 3, app cap effective Nov 5, Nomura approved as a Super Validator Nov 10, Paxos weight passes Nov 11.

The Canton Foundation warned validator operators in late October. Send's Canton Wallet went offline Oct 22 after a security incident. Send says no customer wallets or funds were reached and that it moved funds under the affected keys to safety early. The Canton Foundation's Oct 26 advisory warned that attackers target operators' networks to reach node keys. It told operators to run Splice 0.9.4 and Wallet Gateway 1.13.0 and to move large balances into external keys. Send's Canton Wallet came back online on Nov 10 on a new validator, and Super Validator votes re-featured Send's apps under new party IDs the next day.

SEC staff set out their position on CC on Nov 3. Division of Corporation Finance staff view CC as a digital commodity and won't recommend enforcement action on the facts that Digital Asset's counsel described. The letter rests on counsel's representations and "expresses the Division's position on enforcement action only and does not express any legal conclusion."

Around DTCC's launch, the SEC letter, and Send's outage, CC's two-day moves stayed inside its usual range against BTC, ETH, and SOL.

Roadmap

Universal burn carries the longest dates on the roadmap: Phase 1 in Q1 2027 and Phase 2 in Q3 2027, both its authors' estimates. CIP-0104's switch is set for February 2027, inside universal burn's Q1 2027 estimate.

Table of planned steps: CIP-0104 dry run in December; CIP-0104 switch in February 2027; universal burn Phase 1 in Q1 2027 and Phase 2 in Q3 2027, both estimates; CIP-0127 lock enforcement with no date.

Each participant joins universal burn on its own timeline, as it moves to Daml 3. Those moves set when institutional traffic starts paying in CC.

Catalysts Ahead

CIP-0104's dry run of the amended formula runs in December, the first full month under the cap.

Traffic at September's 27.1 GB a day or better would show volume held under the cap. App rewards at or under $0.98 per $1 burned for Q4 would show the cap and Guidance v2.0b biting at their five-day pace. A CC price below its Nov 10 close also pushes app rewards per $1 burned under that mark. App rewards under that mark point to the cap only if CC stays near or above its Nov 10 close.

Scorecard of five readings, each with its Q3 value, Oct 1 to Nov 10, the capped days Nov 6 to 10, the Q4 mark and a status set from the capped days. Traffic ran 24.5 GB a day in the capped days against a mark of 27.1 or more, behind. App rewards ran $0.91 per $1 burned against $0.98 or less, and burn 77.9% of round mints against 73.8% or more, both on track. Mints by vote came to 4.0M CC against a mark under 174.0M CC. Universal burn stays pending until a second CIP is filed.

In the first five capped days, burn covered 77.9% of round mints, against 72.9% had the cap held nothing back. At a CC price near or above its Nov 10 close, Q4 burn at or above 73.8% of round mints would show the lift holding.

At its five-day pace, the cap leaves about 78M CC of app rewards unminted in Q4, against 223M over Q4's 56 capped days at 1.0 weight. Capped rewards stay in the unclaimed pool, where votes can mint them later. Vote mints would offset that saving only above about 174M CC in Q4, 78M more than Q3's 96.2M.

Conclusion

In Q3, trading venues paid most of Canton's burn, not the institutions Canton was built for. App rewards returned most of what the largest buyers spent, while round mints paid apps as a group more than the network burned. From Nov 5, a general app's traffic earns less in app and validator rewards than it burns.

In Q4, burn at 73.8% of round mints or more, at a CC price near or above the Nov 10 close, would show the cap's lift holding. Universal burn then brings the institutions' dedicated traffic into CC from an estimated Q1 2027, at discounts a second CIP sets.

Appendix

Quarter figures run Jul 1 to Sep 30, 2026 UTC; later readings carry their own dates. Three definitions govern the quarter figures: round mints, which leave out votes and the Development Fund; burn in dollars, which prices each day's burn at that day's network price; and the Global Synchronizer, which leaves out dedicated traffic.

Definitions

Q4 marks. The traffic mark is September's daily average. The app-reward mark blends the pre-cap rate for Q4's first 36 days with the rate over the first five capped days for the last 56, weighted by daily burn. The burn-share mark blends the 67.8% of Q4's first 36 days with the rate over the first five capped days. The vote-mint mark is Q3's 96.2M; the 174M line adds the cap's five-day pace, about 78M CC for Q4.

Annualized. Q3 flows times 365/92. The multiple path uses trailing 92-day burn on the same basis.

App rewards per traffic dollar. App rewards in dollars over burn in dollars for the same window. Validator rewards per traffic dollar work the same way.

App windows. Per-app figures for Q3 run Aug 8 to Sep 30, 2026; after the quarter they compare Oct 1 to Nov 4 with Nov 6 to Nov 10. Temple's fee record starts on Aug 2, 2026. Burn by payer runs to Nov 4, the last full day before the cap.

Burn in dollars. All CC burned each day times that day's mean network price, summed. Every Q3 burn falls under the traffic label, and burn in dollars equals traffic spend in Q3.

Burn to round mint. CC burned over round mints, in CC, for the same window. Burn to all mints adds mints by vote and Development Fund mints to the denominator. At a constant price, Q3's traffic in MB is priced at Q2's dollars per MB and burned at Q2's burn-weighted CC price of $0.152.

Capped rewards per traffic dollar. App and validator rewards over burn. The vote that closed Nov 4 caps app rewards at $0.79 per unit of reward weight. The pair stops at $0.99 per $1 burned at 1.0 weight, about $1.07 at 1.10, and about $1.15 at 1.20. Super Validator rewards fall outside that limit.

CC authorized by vote and CC minted by vote. Authorized counts accepted vote records from the unclaimed pool. Minted leaves out records that never reached supply: Zenith's Aug 21 record and Franklin Templeton's Sep 14 record, which named the wrong party ID.

CC close. The close of a UTC day reads from the next day's 00:00 row, which puts the Sep 30 close in the row dated Oct 1.

Dedicated synchronizer. Capacity on Canton's main network outside the Global Synchronizer, run by its own operator.

Represented asset value. Real-world assets that stay off chain while a network records their holders or settlement, read per network at the time of each capture: Dec 22, 2025, Mar 20, Jun 27 and Sep 17, 2026. A child network's value counts under its parent.

Disclosed lock wallets. Wallets that Super Validators list for locking. The balance of such a wallet counts as locked; locks stand by commitment until CIP-0127 ships.

Global Synchronizer traffic. Bytes of traffic validators buy each UTC day on the Global Synchronizer, converted at 1 MB = 10^6 bytes.

Live CC supply. Market cap over price in the same-day row. Its quarter change is the live rise.

Mints outside rounds. Mints by vote plus Development Fund mints.

Net round issuance. Round mints less CC burned.

No vote on record. A node with no cast vote on a proposal: every such entry lacks a cast time.

Multiple range since April. Market cap over trailing 92-day annualized burn, Apr 1 to Oct 31, 2026: a low of 5.044x on Aug 7 and a high of 8.749x on Sep 27. The burn a market cap needs at a multiple is that market cap over the multiple.

Network CC price. The CC price the 13 nodes post each round, which sets CC burned per traffic dollar.

Party. An identity on the network. One person or firm can hold many. A validator can offboard them in bulk.

Seat types. An operator seat pays the party that runs its node. A hosted seat pays a party other than the one that runs the node; that party can belong to the same organization. An escrow seat holds seat weight for a partner whose rewards go to the unclaimed pool until milestones clear.

Unclaimed pool. Rewards no party claimed, including escrow seats' share, held for later minting by vote.

Vote counts. One vote per node, latest cast.

Restatements

Development Fund mints. The network's mint record counts 67.5M CC minted to grantees by Aug 24, 2026, 57.2M of it to Digital Asset's parties. The Development Fund's Aug 24 report lists 62.2M CC across 23 projects. Its list leaves out three projects whose mints the network records: BitSafe's 4.0M CC on Jul 9, Obsidian Systems' 1.05M CC for Daml training, and Bit Dynamics' 0.55M CC for settlement patterns. It also lists 0.3M CC more for Digital Asset's projects than had minted by that date.

Vote mints. Nodes accepted vote records for 122.5M CC in Q3, and 96.2M CC of it minted. Two records named a party ID that could not take the mint. Zenith's Aug 21 record for 25.42M CC under CIP-0091 carried an incorrect party ID and minted nothing; a Sep 4 vote sent the same 25.42M CC to Zenith's verified validator party, and that vote minted. Franklin Templeton's Sep 14 record for 926,060 CC under CIP-0118 also used the wrong party ID; a corrected vote passed on Nov 11.

Method notes

Supply walk. The walk follows the change in live supply. In Q3 live supply rose 158.9M CC more than round mints less burn, within 4.6M of the 163.5M CC minted by vote and to the Development Fund.

Traffic and reward dollars. Monthly Global Synchronizer traffic on monthly app reward dollars, Feb to Oct 2026: each extra $100K of app rewards a day goes with about 1.5 GB a day more traffic, an R² of 0.83, while a time trend alone gives 0.94. Monthly changes, eight in all, give a p of 0.09.

Payer groups. Temple counts its four parties. Other trading venues are Kora and Cantex. Wallets are Cantor8's C8 Wallet, PixelPlex's Console Wallet, Loop, and Send's Canton Wallet. Each group's burn is the daily change in its parties' cumulative traffic purchases.

Payer concentration. The Herfindahl-Hirschman index sums each payer's squared share of CC burned: 866 in Q2 and 2,231 in Q3 on a 10,000 scale, the same as 11.5 and 4.5 payers of equal size. Burn by the smaller validators outside the named groups counts as many small payers; counted as one payer, the index reads 911 and 2,298.

CC against the market. Daily CC returns regressed on an equal-weight basket of BTC, ETH, and SOL: slope 1.04 in Q3 and 1.04 from Oct 1 to Nov 10, against 0.25 in Q2. Event moves subtract the slope times the basket's move, with the slope from the 60 days before each event; Aug 17's one-day move, 7.0% below the model, is the largest of the four events, with a p of 0.016. Over Q3, the slope times the basket's daily moves compounds to +60.7%, against CC's -10.4%; by month, July carried 26.0 and August 27.2 of the 58.4 log points of shortfall, September 5.2.

Multiple stability. Market cap over annualized trailing 92-day burn, daily from Dec 1, 2025 to Oct 31, 2026. A unit-root check on its log reads t of -1.36 against a 5% bound of -2.87 and doesn't reject a unit root; split into its legs, market cap carries 86% of its variance since March and all of it since April, when burn moved slightly against it.

Burn over all new coins. Burn over burn plus the net change in supply, by quarter. Canton reads its live supply. Tron reads TRX minted and TRX burned each day on Tronscan. Ethereum reads the sum of all ETH balances (ultrasound.money) against base and blob fees burned (Etherscan): 4,185 ETH burned against 263,634 ETH of net new supply in Q3. Solana burns half of each base fee; its Q3 burn came to $6.7M against $502M of staking rewards minted, about 5.6M SOL, at the network's inflation rate of about 3.6% a year. Avalanche burns every C-Chain fee, and staking rewards mint new AVAX at the end of each staking period. XRP Ledger burns every fee, against escrow releases that add about 5.4% a year to its circulating supply.

Daily active users. CantonScan's rolling 24-hour count ran 4.5% to 25% below The Tie's UTC-day count in the three reads compared.

Peer fee definitions. Ethereum counts base, priority, and blob fees. Solana counts base, priority, and vote fees, without Jito tips. Tron counts TRX burned, at voted rates. Avalanche counts C-Chain burn. Stellar counts fees charged. Hedera counts transaction fees paid to nodes and the treasury. XRP Ledger counts XRP burned as fees. Canton's fee is a voted dollar price.

Rewards against token value. The USDC bridge and CBTC reward values use the mean network price over Aug 8 to Sep 30, about $0.1085, against supply read on Nov 5.

Reward-traffic fit. Eight monthly changes in traffic against app reward dollars, February to October 2026, give a p of 0.09, above the usual 0.05 bar.

Peer fee basis. Each peer's fees come from its own chain record where one exists. Tron's are TRX burned for fees on Tronscan. Avalanche's are C-Chain fees from the Avalanche Metrics API. XRP Ledger's are the daily fall in total XRP, since every XRP fee is burned. Stellar's are the daily rise in the network's fee pool, read from Stellar's history archive. Hedera's are DefiLlama's daily fees, with three short days (Jul 2, 2025, Jun 19 and Jun 20, 2026) replaced by the mirror node's full-day sums. Ethereum's and Solana's are DefiLlama's: on Aug 15, 2026, Ethereum's 7,167 block receipts sum to 81.74 ETH of fees, within 0.5% of DefiLlama's day, and Solana's quarters match Token Terminal within 0.1%. Each chain's coin amount is valued at that day's close.

Peer fees in native coins. Peer fees in TRX, ETH, and SOL divide each day's fees by that day's close, summed over Q2 and Q3.

Data Availability

Rocky. Dune Labs runs Rocky, a spot and perpetuals venue on Canton. Its spot volume stays out of the venue totals. Its daily candles show exactly 96 trades a day from August 2025 through June 2026 and 384 a day in July. On Aug 10, 2026 its hourly candles show 16 trades in every hour, many of them 0.206223 CBTC. Its CBTC-USDCB and CBTC-USDCX pairs print identical candles on 345 days.

Data Sources

CC burned. CantonScan.

CC burned for traffic. CantonScan.

CC burned as other fees. CantonScan.

Round mints. CantonScan.

App rewards. CantonScan.

Validator rewards. CantonScan.

Super Validator rewards. CantonScan.

Global Synchronizer traffic. CantonScan.

Burn in dollars. CantonScan.

App rewards in dollars. CantonScan.

Validator rewards in dollars. CantonScan.

Round mints in dollars. CantonScan.

Network CC price. CantonScan.

Explorer running supply. CantonScan.

CC authorized by vote. CantonScan.

CC minted by vote. CantonScan.

Development Fund mints. Loop Data.

Vote mints received by each recipient. CantonScan.

CC close. CoinGecko.

Market cap. CoinGecko.

Trading volume. CoinGecko.

DEX volume on Canton by venue. DefiLlama, protocol breakdown. Temple's own records agree within 0.11% for Q3.

Canton DEX volume. DefiLlama, the sum of the venue series.

Daily active users. The Tie.

Burn in dollars, Jul 1 to Sep 29. CantonScan.

Live CC supply. CoinGecko.

Seat weight. CantonScan.

Disclosed lock wallets. CantonScan.

Top 100 holders. CantonScan.

Featured apps listed. CantonScan.

Parties. CantonScan.

Featured app grants voted. CantonScan.

Featured app revocations voted. CantonScan.

Bid depth within 2%. CoinGecko.

Ask depth within 2%. CoinGecko. Bids and asks read at 12:43 UTC on Nov 5 and at 11:19 UTC on Nov 11.

Perpetual open interest and volume. Each venue's API (Binance, Bybit, OKX, Hyperliquid), with daily history for seven venues from Coinalyze. Open interest counts one side of each position at the UTC close; Bybit, Hyperliquid, Gate, and Lighter count both sides, and their figures are halved. Bybit's quarter-end reads match its one-side field exactly.

Bitcoin close. CoinGecko.

First lock step. CIP-0105's top lock tier falls to 65% of lifetime rewards from 70% a year after its Mar 2, 2026 approval. The step on the 25.7B CC earned to Oct 4 is 1.29B CC, which vests at 1/365.25 a day, priced at the Nov 5 CoinGecko close.

Token prices and balances on Canton. CantonScan.

Ethereum fees. DefiLlama.

Ethereum close. CoinGecko.

Ethereum market cap. CoinGecko.

Solana fees. DefiLlama.

Solana close. CoinGecko.

Solana market cap. CoinGecko.

Tron fees. TRX burned for fees, Tronscan.

Tron close. CoinGecko.

Tron market cap. CoinGecko.

Stellar fees. The network's fee pool, Stellar history archive.

Stellar close. CoinGecko.

Stellar market cap. CoinGecko.

Avalanche fees. C-Chain fees, Avalanche Metrics API.

Avalanche close. CoinGecko.

Avalanche market cap. CoinGecko.

Hedera fees. DefiLlama, with three days from the Hedera mirror node.

Hedera close and market cap. CoinGecko.

XRP Ledger fees. The daily fall in total XRP, read at each UTC midnight (XRPL).

XRP Ledger close and market cap. CoinGecko.

Burn to round mint. CC burned as a share of round mints.

Net round issuance. Round mints less burn.

App rewards per traffic dollar. App rewards over burn in dollars.

Validator rewards per traffic dollar. Validator rewards over burn in dollars.

Capped rewards per traffic dollar. App plus validator rewards over burn in dollars.

Traffic in MB. Global Synchronizer traffic at 1,000,000 bytes per MB.

Super Validator rewards in dollars. Super Validator rewards each day times that day's mean network price, summed.

App and validator rewards in dollars. App rewards in dollars plus validator rewards in dollars.

Minted outside rounds. CC minted by vote and to the Development Fund since Jun 30, 2026, Q3 only.

Live supply above the explorer's running supply. The rise in live CC supply less the rise in the explorer's running supply since Jun 30, 2026.

Breakeven CC price. The CC price at which Q3 burn in dollars burns as much CC as rounds would then mint, with validator rewards rising with burn: $0.0737. Q3 traffic reads 100% of round mints at that price.

Canton fees. Burn in dollars, set beside peer fees. Canton's fee is a voted dollar price.

Represented real-world asset value by network. RWA.xyz network table, each network's non-transferable value, from Internet Archive captures on Dec 22, 2025, Mar 20, Jun 27 and Sep 17, 2026. The network rows sum to the page's represented total within 1%. Broadridge's own September print ($7.5T, $359B a trading day) is the first-party check on Canton's row.

References

Protocol rules: CIP-0067, CIP-0082, CIP-0104, CIP-0105, CIP-0116, CIP-0124, CIP-0127, CIP-0129, Featured App Guidance v2.0b, tokenomics docs, dedicated synchronizers, Universal Burn deck.

Quarter events: DTCC, Jul 15, Tradeweb and USDCx, message #386, Broadridge DLX, Broadridge G7 repo, Broadridge September repo print, LSEG appointment, SEC Innovation Exemption, Development Fund report.

Events after the quarter: DTCC service launch, Uber pilot, Send's notice, Foundation notices, Foundation board framework, SEC staff letter, counsel's letter, SV operations schedule.

Filings: Tradeweb 10-Q, Broadridge 10-Q, Circle 10-Q, Canton Strategic Q3 10-Q.

Data: CantonScan, Loop Data, CoinGecko, DefiLlama.

Peers: Ethereum fees, Solana fees, Tron fees, Stellar fees, Avalanche fees, Hedera fees, XRP Ledger fees, Arc, Solana DvP.

Disclosures

This report on Canton was not paid for or commissioned.

Members of the Alea Research team, including those directly involved in the analysis above, may have positions in the tokens discussed.

This content is provided for educational purposes only and does not constitute financial or investment advice. You should do your own research and only invest what you can afford to lose. Alea Research is a research platform and not an investment or financial advisor.

Canton: Q3 2026 — IR deckCanton: Q3 2026 — IR deck
IR deckThe same research in 40 slidesOpen the IR deck