Reference data (financial markets)

In order to facilitate rapid, error-free completion of financial transactions across multiple geographical markets and in support of increasingly complex products, the financial service industry and regulatory agencies have pursued a policy of standardizing the reference data that define and describe such transactions.

At its most basic level, reference data for a simple sale of a stock in exchange for cash on a highly liquid stock exchange that involves a standard label for the underlying security (e.g., its CUSIP), the identity of the seller, the buyer, the broker-dealer(s), the price, etc. At its most complex, reference data should ideally cover all relevant particulars for highly complex transactions with multiple dependencies, entities, and contingencies.

The background for this policy is the risk that transactions fail and are reversed because contractual terms were misunderstood or ambiguous. In addition, the lag between the trade and ultimate settlement of the transaction may include various events that affect various elements of the transaction.

Efforts to standardize reference data are complicated by a number of factors, including:
* Semantic differences in common terminology
* The sheer number of data elements that make up transactions
* Rapidly changing markets, products, and underlying events

As a result, work to standardize reference data is broadly considered to be an ongoing effort rather than a series of discrete programs.