A trade is a fact shared by two parties, and its identifiers exist to make that shared fact provable. Where the order record named an instruction, the trade record has to name a completed, bilateral event — one that two counterparties, and often two regulators, must each be able to point to as the same thing.
Three identifiers are common to almost any executed trade. The Legal Entity Identifier (LEI) names each counterparty. The International Securities Identification Number (ISIN) — or, for many derivatives, the Unique Product Identifier (UPI) — names the instrument. The execution timestamp fixes the moment. Together they answer who, what, and when; but none of them, on its own, names the transaction, and it is naming the transaction that makes reconstruction possible.
Here the path splits by what was traded and where it is reported. For derivatives reported to a trade repository, the transaction is named by a Unique Transaction Identifier (UTI): a single code that both counterparties must report against their own side of the trade — so a regulator can pair two independently submitted records and confirm they describe one trade. Because only one side can create that shared code, a fixed order of precedence (known as the generation waterfall) decides who does: for a centrally cleared trade the clearing house generates it, otherwise the trading venue, otherwise the two counterparties agree between themselves, and if they cannot, a deterministic rule based on sorting their identifiers settles it. The rule exists to guarantee exactly one code, generated once, rather than two sides colliding or each waiting on the other. For equities executed on a venue under the Markets in Financial Instruments Regulation (MiFIR), the equivalent is the Trading Venue Transaction Identification Code (TVTIC), generated by the venue for the market side of the execution. In the United States, no single shared code is used at all; instead the execution is linked into the order-event chain of the Consolidated Audit Trail (CAT) through linkage keys, and the record is stitched together rather than paired.
The distinction matters more than it first appears. A UTI is a shared code, so two reports can be matched to each other; a TVTIC is a venue code, so a report can be matched back to the venue's own record; a CAT linkage key ties an execution back up its own chain of order events. Each is a different mechanism for the same purpose — the hook a later reconstruction hangs on. Every auditability question that follows in this Primer depends on these hooks already being understood.