NEWS
Polymarket Protocol V2 Leaves the Old Book Running
Polymarket Protocol V2 moves new markets to pUSD on November 2, while old CTF positions stay put as Kalshi’s index perpetual is already approved.
Polymarket will put new prediction markets on Protocol V2 on November 2, with pUSD as the only collateral on that book. Rajath Alex, the company’s head of protocol, laid out the rebuild on October 5. Canary markets are already live on the new contracts.
Markets already running on the 2019 Gnosis Conditional Tokens Framework stay there. The new collateral layer opens beside that ledger, not on top of it.
One Token, One Router, Four Market Types
Alex said Polymarket has been running on Gnosis Conditional Tokens Framework code since 2019, a general-purpose kit that forced a new contract in front of every market type the venue later shipped. That meant an adapter for pUSD, a wrapped token for negative-risk markets, a second exchange for the second market type, and a separate UMA adapter for each hookup, each of them frozen once it went live.
Protocol V2 throws that pile out for a single ERC-1155 positions contract, one collateral token, one exchange for every market type, and one router. Each position ID states the market type, the market, and the outcome, so the contracts read that metadata off the identifier instead of walking a chain of adapters.
The first modules cover four structures. Each one also adds extra position operations that Alex said are meant to squeeze more use out of the same collateral.
WHAT V2 OPENS WITH
- Binary: A yes-or-no market whose shares live as ERC-1155 balances in PositionManager.
- Atomic neg-risk: A negative-risk book that settles as a set, rather than as a lone pair.
- Incremental neg-risk: A negative-risk book that can add outcomes without a fresh adapter contract.
- Combinatorial: A market whose result depends on more than one condition at once.
Those four types share ExchangeV3 and the same router for splits, merges, and redemptions. App and website users do not have to migrate wallets; they complete any new approval prompts the app shows the first time they touch a V2 market.
Old CTF Markets Stay on a Separate Ledger
The unification stops at new flow. Polymarket’s migration overview states that existing CTF holdings are not converted, and contract docs repeat that V2 balances sit on a separate ledger. Integrators keep CTF token IDs for V1 markets and use uint256 position IDs for V2.
pUSD is the sole V2 collateral. It wraps USDC and USDC.e at a 1:1 ratio through onramp and offramp contracts, and the same pUSD system still feeds the old CTF adapters so value can move between the two books. That is a bridge, not a merge. A holder can move some CTF positions into PositionManager only after Polymarket registers that condition or event, which keeps the old book alive until the company chooses to open a path.
CTF VERSUS PROTOCOL V2
| Piece | CTF book | Protocol V2 |
|---|---|---|
| Position token | Gnosis Conditional Tokens, live since 2019 | Single ERC-1155 PositionManager |
| Collateral | Adapters and wrapped tokens per market type | pUSD only, wrapping USDC and USDC.e 1:1 |
| Exchange | CTF Exchange plus a Neg Risk CTF Exchange | One ExchangeV3 for every market type |
| Split, merge, redeem | Layered adapters, one more approval each time | One Router, new approvals required |
| Old balances | Stay in CTF unless a condition is registered | Separate ledger; CTF approvals do not carry over |
On Polygon mainnet (chain ID 137), PositionManager owns the ERC-1155 balances and routes each operation to BinaryModule, NegRiskModule, or CombinatorialModule. The docs warn that protocol mints and unsafe transfers skip ERC-1155 receiver callbacks, so any vault that listens for those hooks has to count the balance change itself.
How Resolution Works After the OracleAggregator
Settlement is the other half of the rewrite. V2 sends results through an OracleAggregator that accepts pluggable reporter modules, with UMA’s optimistic oracle, a Chainlink reporter, and an externally owned-account reporter shipping first. The aggregator takes those votes and pushes a final result to the module that owns the condition, so a price market and a messy political market do not have to share one dispute path.
UMA still fits event questions that need a challenge window. Chainlink fits markets that already have a price feed and want faster finality. The design also leaves a socket for later oracles, which is why rival resolution shops immediately asked to be next in line. The aggregator is an invitation, not a finished map of who will settle which book.
But every new feature meant another layer, another approval for you, another contract for integrators, and another thing that can’t be changed once it’s deployed.
Rajath Alex, Head of Protocol, Polymarket, on X
He posted the full note, including the canary window and the Data API cut, from his own account.
Introducing Polymarket Protocol V2 https://t.co/FDzaZyiPnl
— Rajath Alex (@0xrajath) October 5, 2026
Bridging of positions, collateral, and resolution results is built into the contracts and stays dark until Polymarket turns on a multi-chain setup. Until then, the live change is simpler: desks that want the new market types talk to one aggregator instead of a fresh UMA adapter per listing.
Canary Markets Run Through October 30
Alex is not flipping the whole venue in one night. A small set of canary markets has been trading in production since October 5, and that window runs through October 30 so market makers and bot writers can point at live contracts. Net-new markets are penciled in for November 2. He called that date tentative, and the company is running weekly office hours plus a Discord channel for the four-week stretch.
A second clock is already louder for anyone who reads the Data API. Data API v2 is a Rust service on Polymarket’s own on-chain indexer, with a shared response envelope, snake_case fields, and cursor paging that replaces v1’s 10,000-row offset cap. Data API v1 is retired on October 24, a hard stop that lands before the protocol cutover and does not wait on November 2.
THE OCTOBER CUTOVER
- October 5, 2026: Alex publishes Protocol V2 and canary markets open in production.
- October 24, 2026: Data API v1 is retired; remaining reads must use the v2 routes.
- October 30, 2026: The canary window ends.
- November 2, 2026: Net-new markets are tentatively created on Protocol V2; the CTF book stays.
GET /v1/accounting/snapshot has no v2 twin and is not part of that retirement, so shops that still pull that one route keep the old path. Everything else in the mapping, including positions, user PnL, user stats, and resolutions, moves before October 24 or it goes dark.
Kalshi’s US500PERP Is Already Approved
The CFTC’s product file shows KalshiEX’s US500PERP listed as approved, dated October 3, as a future on an equity index. That is not a prediction-market rewrite. It is a perpetual-style index future on a U.S. designated contract market, sitting in the same week Polymarket asked desks to learn a new router.
On October 5 the CFTC’s Division of Market Oversight also told designated contract markets they may convert existing index futures into perpetuals if they warn holders and leave other terms alone. Those no-action positions expire on October 20. The staff letter is a green light for listed venues to strip expiries off broad index futures, which is the product shape Kalshi just cleared and Coinbase had already been pushing.
Polymarket US already trades as a CFTC designated contract market after the 2025 QCX deal, so the company is in the same regulatory club. V2 does not list an S&P-style perpetual. It tries to make event contracts cheaper to market-make by cutting approvals and putting every new share on pUSD. Whether that is enough depends on where the same desks park inventory, because Kalshi’s live crypto perpetual futures already give them a funded, onshore book that never expires.
Two ledgers on Polymarket and a newly approved index perpetual next door is a messy week to ask a market maker for yet another set of allowances. The people who feel that first are the ones who quote both venues, not the casual app user who taps through a fresh prompt.
Re-Approve the Router Before You Split
For anyone who signs orders or holds positions in a contract, V2 is a new permission surface. Existing CTF approvals do not carry over. Splits, merges, and redemptions now go through the Router in one transaction; the docs tell integrators not to pre-send tokens to a module and wait.
Amounts use six decimals, so 1,000,000 base units equal one pUSD or one share. Redemption uses outcome index 0 for YES and 1 for NO, replacing the old CTF index sets, and payouts are read from PositionManager.getPayout. Unresolved positions can revert; a losing payout of zero is valid.
NEW APPROVALS FOR V2
- Split: Approve the Router to spend pUSD from the holding account.
- Merge or redeem: Call setApprovalForAll on PositionManager for the Router.
- Buy orders: Approve ExchangeV3 to spend pUSD, including fees.
- Sell orders: Approve ExchangeV3 as a PositionManager operator, and sign against ExchangeV3 with EIP-712 domain version 3.
- Auto-redeem: setApprovalForAll on PositionManager for AutoRedeemer if the desk wants an operator to sweep resolved balances back to pUSD.
Keep CTF support in the same bot. The official guidance is to add V2 beside V1, pick position IDs or token IDs from the market’s version flag, and treat CONDITIONAL-V2 as the CLOB asset type for V2 allowance checks. Shops that ignore that fork will show empty balances on one book while the other still holds risk.
Six Firms Signed Off Before Canary Trading
Alex said the Protocol V2 contract code and audits were reviewed by Cantina, Certora, Quantstamp, SigmaPrime, Zellic, and Pashov, with Certora also running formal verification. Published reports on the repo run from May through August, covering the combinatorial module, the oracle subsystem, combo collateral return plus CCIP, and the V2 core. A bug bounty pays up to $5 million for critical findings.
The whole V2 set is upgradeable behind what Alex called a secure governance process, which is the point of leaving Gnosis CTF behind. New features can ship without bolting on another immutable adapter. Follow-on work already named in the same note includes scalar resolution and directional collateral return, neither of which is in the canary set.
Until November 2, the live test is narrower. Canary markets prove whether one router and one pUSD token cut the approval tax for the desks that quote this venue. Old CTF markets keep settling as they do now. If the tentative date slips, the split simply lasts longer, and the index perpetual Kalshi already cleared will have had more time to pull the same professional flow.
Disclaimer: This article is news reporting on a smart-contract upgrade and related CFTC product filings. It is informational only and does not constitute investment, trading, or legal advice. Readers should consult a qualified financial adviser or attorney before trading event contracts, perpetual futures, or any related crypto asset. Figures, contract statuses, and migration dates reflect the cited documents and may change if Polymarket revises the cutover or if regulators update a product file.
-
NEWS4 months agoThe Orchard Bug Forced Zcash to Open an Empty Pool
-
GAMING4 months agoThe $5.99 Game That Won Steam’s 2026 Summer Sale
-
ENTERTAINMENT6 months agoDon Lee Joins Hemsworth as Extraction 3 Starts Shooting
-
NEWS4 months agoYouTube Shorts Retires Dislikes and Swaps Likes for Hearts
-
NEWS4 months agoNEURA Robotics Is Spending Its $1.4B on a Machine Stack
-
AUTO6 months agoKawasaki Bets the New Z1100 Against Its Own Supercharger
-
BUSINESS5 months agoNorway’s Export Bank Backs Nscale’s $790 Million Narvik Loan
-
GAMING4 months agoGame of Thrones Dragonfire Follows Boston’s Conquest Playbook
