Market Data & Analysis

Trading Account Net Position Changes

Explanatory Notes

The Trading Account Net Position Changes data, published by the Commodity Futures Trading Commission (CFTC or Commission) on June 30, 2011, builds on the CFTC’s efforts to improve market transparency. The data identifies, for a given week, the daily-average aggregate net position change at the trading account level for 28 futures markets. The data depicts trading that changes or creates an end-of-day position, as contrasted with trading that has no effect on a trader’s end-of-day net position, such as spread or day trading.

The new market-specific data augments the volume and open interest data that exchanges routinely provide. The weekly data covers futures markets for 20 physical commodities and 8 financial products.1 The data is being published as a one-time report together with historical data going back to April 2010. The data release should provide the public, academia and traders with further insight into market liquidity.


The Trading Account Net Position Changes data is constructed from transaction data that is associated with each trading account and provided to the Commission by the exchanges. An account represents the combination of a clearing member firm identifier and each of the account numbers used by that clearing firm when executing trades. Staff examined the futures-only transactions associated with each account that had trades cleared on a given day.2 In instances where an executing or allocating firm executed a trade and then “gives up” the trade, post-execution, the “give-ups” were assigned to the claiming or destination account.

As a first step, staff calculated, for each account, the daily all-futures-combined net purchases (excluding option trades).3 For example, if an account purchased 300 futures contracts and sold 200 futures contracts over the course of a day, that account had 100 net purchases. Staff then calculated an aggregate of daily net purchases/sales by summing the net purchases across all trading accounts. Staff calculated the daily average net purchases/sales for a given week as the average of the daily aggregate net purchases for that week. Finally, the report also shows the daily-average futures (only) trade volume data for each market. The latter figure may differ from what the exchange originally published for the same dates due to various trade anomalies.

Each trading account that had a net purchase clearly had a net position change at the account level. Net purchases may result in reducing a short position, moving from a short position to a long position, creating a new long position or increasing an existing long position. Likewise, net sales may result in reducing a long position, moving from a long position to a short position, creating a new short position or increasing an existing short position. Therefore, the type of trading captured by this report is directional in nature, i.e., trading that results in the changing or creating of an end-of-day position at the account level.

Equal Purchases and Sales

Net purchases for an account are calculated using transaction data from all contract months for a given futures market. As a result, the execution of equal purchases and sales on a given day net to zero and will not show up as net purchases. For example, an account’s net purchases across all futures months will be zero when a trader rolls a long or short position from one futures month to another, establishes a calendar month spread position, offsets a calendar month spread position or rolls a calendar month spread. Likewise, other types of trading that result in equal purchases and sales show up as zero net purchases, including, typically, many forms of day trading, high frequency trading and algorithmic trading.

Data Issues and Limitations

The transaction-level data in this report provide a different view of directional trades than that derived from position changes extracted from the Commission’s Large-Trader Reporting System (LTRS). Owners and traders in the Commission’s LTRS can have multiple trading accounts at multiple clearing members, but, because that ownership information is not available in the transactional database, each separate trading account is treated as a separate entity for this report. As a result, some purchases and sales that would offset each other under a given owner/trader will not appear that way if done in separate trading accounts. For example, if an owner/trader bought 150 contracts in one trading account and sold 100 contracts in another account, the net, daily position change for that owner/trader would be 50 contracts. In contrast, in this Trading Account Net Position Changes data, the net, daily purchases and sales would reflect the purchasing account’s 150 contracts of purchases without the netting of the 100 contracts sold in the selling account.

In the transaction data associated with this data, transactions cleared through a single trading account at a clearing member could represent transactions by a single trader or many traders. For example, many traders could execute trades through a single account, or futures commission merchants may execute transactions for multiple traders through omnibus accounts. Because the account-level data is not all directly related to a particular large trader in the Commission’s large-trader reporting system, the Commission’s different types of large-trader classifications cannot be applied. Similarly, these data cannot determine whether trades were placed as part of market speculation, an arbitrage, an option strategy, or a hedging program.

1 Data from ICE Futures U.S., Minneapolis Grain Exchange and the Kansas City Board of Trade could not be incorporated into the analysis as the Commission lacked the necessary data for the entire study period and due to changes in the structure of the data that the Commission received.

2 Focusing on the cleared-on date, as opposed to the transaction date, allows staff to account for out-trades. A trade may remain “out” over night due to a discrepancy in the way in which buy and sell data are matched.

3 The same computation of net purchases/sales could have been carried out by summing the net sales. Because the transaction data cover all futures trades, and, because there is a net sale for every net purchase, the aggregate of net purchases must necessarily equal the aggregate of net sales.