POLYMARKET INTERNATIONAL · INDEPENDENT RESEARCH190,542 USDC
back to the funder.
In April, 33f’s treasury funded three wallet owners. In May, their trading wallets paid 190,542.330138 USDC to 33f. Each funding route and later payout has its own receipts.
pays three owner keys.
The route inputs and outputs are separate receipts. Ethereum and Polygon recipients are linked above.
Each key controls a trading wallet.
Historical owner states match the keys recovered from the May 28 execution signatures.
These three signing keys differ from . The records do not establish whether one person or firm controls them all.
The three wallets later pay .
Other funding and trading occurred between April and May. These payment totals do not measure profit.
33f-controlled B27 pays 15,000 USDC.e into three April bridge routes.
From the April funding to the May payouts.
Start at B27, follow the three named recipients, then follow their wallets’ later payouts to 33f.
B27 funds three owner keys.
Three 5,000-USDC.e inputs produce 14,987.919845 USDC of bridge outputs. B27’s sole recorded owner at each input was 33f.
Each key controls a trading wallet.
The three keys map to d2e8, 5eb2 and 73ed. Historical owner states and execution signatures establish the mappings.
Those wallets pay 33f.
Three redemptions deliver 190,542.330138 USDC to the original controller. Other flows and trading sit between these dates.
All three bridge inputs and all three outputs agree with the recorded recipient-specific routes.
The wallet owners are separate addresses. The funding relationship comes from actual payments, rather than shared use of a bridge service.
The April and May amounts measure different payments. The displayed difference is not a profit calculation.
Inspect each funding and payout route.
Three funding paths. Three distinct owner keys. Inspect the April and May records side by side.
75 maker rebates match token transfers.
Official activity records identify the payment type. Successful token-transfer receipts verify who received each amount.
Swipe to inspect every column →
| WALLET | PAYMENTS | USDC.e PAID | pUSD PAID |
|---|---|---|---|
d2e8 | 22 | 7,308.4327 | 6,475.8218 |
5eb2 | 17 | 3,488.1472 | 4,676.7180 |
73ed | 23 | 744.4937 | 873.5175 |
d950 | 13 | 0 | 56,599.5531 |
| Total | 75 | 11,541.0736 | 68,625.6104 |
April 23–May 28 payment query window. Every official MAKER_REBATE entry matches a unique token-transfer log. D950 is the separate fourth wallet that also paid 33f in May.
B27 and d950 matched in 22 transactions.
Exact bilateral event filters and successful receipts identify the direct matches. Shared transaction hashes alone do not enter the count.
Direct matched fills
B27 and financially connected d950 meet in 22 onchain transactions on September 20.
pUSD maker collateral
The 32 fills are complementary-outcome buys. This is the matched maker side of those fills.
pUSD rebate received
D950 receives a separate maker-rebate payment that day. The payment date is verified.
Shared transactions are not all direct matches.
The evidence distinguishes a direct match from several unrelated executions in the same settlement transaction.
The two account feeds share 2,504 transaction hashes. After checking actual counterparties, 22 transactions contain direct matching between them.
Most shared transactions bundle trades with other counterparties.16 contiguous event-filter queries cover the two v2 exchanges for September 20. Each of the 32 direct fills agrees with its successful transaction receipt.
Polygon blocks 94,104,891–94,162,484 inclusive.Sources and calculations.
Download the source records and verification code. Evidence rechecked September 21, 2026.
How the evidence is counted
The financing loop covers three specific earlier B27 routes and three later payouts. A separate fourth wallet, d950, supplies additional May funds and the September matched-fill example; it is not depicted as an April-funded recipient. The owner keys are distinct. Maker rebates are identified by the official activity category and exact transfer logs, with USDC.e and pUSD kept separate. The 32 direct fills are complementary-outcome buys, rather than same-token resales. These facts establish financial relationships, actual payments and a limited matching example; they do not establish common ownership of all keys or systematic artificial volume.