Qm
Core

The EMO Trade Classification Rule

A method for labeling each trade in a tape as buyer-initiated or seller-initiated by combining the tick test with a check on whether the trade occurred at the bid or ask, improving on either rule used alone.

Prerequisites: Bid-Ask Spread Decomposition

Trade data typically records a price and a size but not whether the buyer or the seller was the one crossing the spread to make the trade happen. Reconstructing that "who initiated" label matters enormously for order-flow research, since buy-initiated and sell-initiated volume carry very different information about near-term price pressure. The EMO rule, named for authors Ellis, Michaely, and O'Hara, is one standard way to make that call.

EMO classifies a trade as buyer-initiated if it executes exactly at the posted ask price, and seller-initiated if it executes exactly at the posted bid price. For the remaining trades, those that print somewhere between the bid and ask, which happens often for actively negotiated or dealer-facilitated trades, EMO falls back to the tick test: a trade is classified as buyer-initiated if its price is higher than the previous trade's price, and seller-initiated if lower. This combination was found, in direct comparison against the alternative Lee-Ready algorithm, to classify Nasdaq trades more accurately, because at-the-quote trades are unambiguous and don't need the tick-test fallback, while the tick test remains a reasonable default for the ambiguous middle-of-spread cases.

No classification rule is perfect; all of them make systematic errors on trades that occur inside the spread during fast-moving markets, and researchers typically validate a chosen rule's accuracy against a labeled subsample before relying on it for a full order-flow study.

The EMO rule classifies a trade as buyer-initiated if it prints at the ask and seller-initiated if it prints at the bid, falling back to the tick test (comparing against the previous trade's price) only when a trade prints inside the spread, a hybrid found to classify trades more accurately than either rule alone.

Discussion

Sign in to join the discussion · reading is open to everyone

💡 Discussion rules

  1. Ask and answer about this concept. Off-topic gets removed.
  2. No homework dumps. Show what you tried first.
  3. Corrections are welcome. Cite a source when you claim an error.

Loading discussion…

Related concepts

Practice in interviews

Further reading

  • Ellis, Michaely, O'Hara, The Accuracy of Trade Classification Rules (2000)
ShareTwitterLinkedIn