TLDR
Script 18's ownership graph module constructs a directed network of 33 static edges (corporate registrations, officer appointments) plus dynamic wire flow edges, then identifies entities where inflows approximately equal outflows with minimal retention. These "pass-through" nodes — entities that function as financial conduits rather than economic actors, simply receiving money and forwarding it along — include Harlequin Dane, Birch Tree Br, and DKI PLLC. They add complexity to the money trail without adding economic substance (PAPER TRAIL Project, 2026).
What a Pass-Through Looks Like
In legitimate corporate finance, money flows through entities that perform economic functions. A holding company receives dividends from subsidiaries. A law firm receives retainer payments and disburses to vendors. An operating company receives revenue and pays expenses. In each case, the entity retains some portion of what flows through it — profit, fees, reserves — because it is providing a service or generating value.
A pass-through entity is different. Money enters and exits in roughly equal amounts. The entity retains little or nothing. Its function is not economic — it is structural. It exists to add a node to the transaction graph, to change the account number and institution name on the wire, and to make the path from source to destination harder to follow.
Detecting pass-through entities requires seeing the full flow picture: all inflows, all outflows, and the net retention. Transaction-level monitoring catches individual wires. Graph-level analysis catches the pattern.
The Ownership Graph
Script 18's ownership graph module constructs this view by combining two types of edges. The 33 static edges come from corporate registrations and officer appointments: Indyke as sole signer on Harlequin Dane, Kahn as sole authorized signer on HBRK Associates, the shared address at 6100 Red Hook Quarter linking seven USVI entities. These edges do not change over time — they represent the structural skeleton of the corporate network (PAPER TRAIL Project, 2026).
The dynamic edges come from the 224 parsed wire transfers. Each wire creates a directed edge from originator to beneficiary, weighted by the transfer amount. When static and dynamic edges are overlaid on the same graph, the result is a network that shows both who controls what and where money actually moves (PAPER TRAIL Project, 2026).
Three Pass-Through Patterns
Harlequin Dane LLC is the clearest example. It received $7.75 million across seven wires from Southern Financial LLC. It disbursed funds to Seaford Avenue Capital, Signature Title Group LLC, JR Watersports Inc., and Jetsmarter. It also sent $221,050 back to Southern Financial as "loan payments" — a circular flow that returned money to the originator. Harlequin Dane was incorporated in Florida on August 14, 2018, with Indyke as sole signer. A short-lived entity with a single funding source, multiple disbursement targets, and a circular return: this is the textbook pass-through profile (TD Bank, 2019).
Birch Tree Br LLC follows a similar pattern at smaller scale. It received $300,000 in two wires from Southern Financial and Southern Trust, then disbursed to Roadruck Investigations (a Miami private investigator) and general business expenses. The entity's primary function appears to have been routing investigative services payments through an intermediary rather than paying the PI directly from a high-profile account (TD Bank, 2019).
DKI PLLC — Darren Indyke's law practice — received $414,741 across six wires from multiple Epstein entities: Southern Trust, HBRK Associates, and JSC Interiors. Outflows went to Black Bag Media ($100,000 to an Arlington, Virginia entity with no public information), Lesley Groff ($1,188.72), and other payees. A law firm receiving client funds and disbursing to vendors is not inherently suspicious. But when the law firm's sole client is also the sole controller of every entity funding the account, the economic justification for the intermediary step diminishes (TD Bank, 2019).
Geographic Clustering
The ownership graph also reveals geographic concentration that amplifies the pass-through pattern. At least seven entities share the registered address 6100 Red Hook Quarter B3, St. Thomas, USVI 00802: Southern Financial LLC, NES LLC, JEGE LLC, F T Real Estate Inc., Bella Klein TTEE, Southern Country International, and Financial Trust Company. When multiple pass-through entities share a single physical address, the structural separation they provide is purely notional — the same office, the same staff, the same filing cabinets (PAPER TRAIL Project, 2026).
Script 18 flags this geographic clustering as a pass-through concentration indicator. Entities at a shared address that exhibit high-flow, low-retention wire patterns are more likely to be structural conduits than independent economic actors.
What Detection Reveals
Pass-through detection does not prove that the entities involved were used for money laundering. It identifies a structural pattern — minimal retention, high throughput, shared control — that is consistent with layering (disguising the source of funds by moving them through multiple accounts or entities). The distinction matters because the same pattern can arise from legitimate business structures: a holding company that distributes dividends, a trust that disburses to beneficiaries, or a law firm that acts as escrow.
The difference lies in economic substance. Harlequin Dane had no employees, no revenue, no clients, and no business operations beyond receiving and disbursing funds controlled by the same person. Birch Tree Br had no apparent business purpose beyond routing payments to a single investigator. When pass-through detection surfaces entities with zero economic substance, the legitimate explanations narrow.
The ownership graph makes these patterns visible by combining corporate structure with money flow. Neither dataset alone is sufficient. Corporate filings show who controls what but not where money moves. Wire transfers show money movement but not who controls the entities involved. The graph layer — static edges plus dynamic edges — reveals both simultaneously, and the pass-through nodes emerge from the intersection.
References
PAPER TRAIL Project. (2026). Entity ownership research [Data set].
PAPER TRAIL Project. (2026). Institutional analysis: Ownership graph module [Script 18].
PAPER TRAIL Project. (2026). Ownership edges and nodes exports [Data set].
TD Bank. (2019). Suspicious Activity Report (BSA-31000155070501). Filed October 1, 2019.