Order Flow Trading Strategy: Know What the Feed Measures

Distinguish the measurements

A transaction feed can report executions; aggressor-side labels may come from the venue or be inferred by a vendor. Buy-classified minus sell-classified volume is delta. Cumulative delta sums that quantity from a declared reset point. Neither quantity is net shares entering the market: every matched trade has two sides. Depth of market shows displayed orders at quoted prices. CME Market by Order adds anonymous order-level information and queue position. Orders can be changed, executed or canceled. A disappearing bid does not by itself identify the owner or guarantee a price decline.

• A footprint groups classified volume into price rows within bars; row size changes the display.
• A volume profile aggregates activity across a selected range. Its highest-volume row describes activity, not a guaranteed support level.
• OBV signs an entire bar’s volume by its close-to-close direction. That is a different calculation from transaction-level aggressor delta.
• Record venue, symbol, delay, trade-size units and session boundaries with any result.

What TradingView currently makes available

TradingView’s footprint calculations classify lower-interval volume by intrabar price movement. They are not a universal record of trades executed at the bid or ask, and do not identify institutional accounts. Its chart documentation also describes how changes in available intrabar resolution can change historical footprints. Pine now supports request.footprint(), which can provide bar and row volume information. Current official documentation requires TradingView Premium or Ultimate, allows only one footprint request per script and returns no footprint for bars without available data. An unavailable bar must not silently become zero delta. TradingView access requirements are separate from a Pineify subscription.

Read a delta example without inventing a prediction

Suppose a feed classifies a bar’s volume as 700 buy and 500 sell. Its delta is +200 and total classified volume is 1,200. If cumulative delta was +300 at the preceding close, it becomes +500, assuming no reset. These are arithmetic examples, not evidence of future returns. A later price high above an earlier high, accompanied by lower cumulative delta, is a bearish divergence under the chosen comparisons. Price may still rise. Compare delta at the same defined price pivots, use the same reset interval and avoid selecting different swing points merely because the picture looks persuasive.

Specify a divergence experiment before measuring it

For an illustrative short-only ES study, use a named futures contract, standard five-minute bars and a documented delta source. Reset cumulative delta at 09:30 America/New_York on full-length US equity sessions; exclude shortened sessions and do not compare pivots across resets. If any required delta bar is missing, exclude the session and report that exclusion. A price pivot high must exceed the highs of the two bars before it and the two bars after it; ties do not qualify. It becomes known only when the second following bar closes. Store cumulative delta at the pivot bar. Compare successive confirmed pivot highs within the session: trigger only if the newer price high is higher and its stored delta is lower.

• Enter only while flat, with one short entry per session, on a confirmation bar closing after 09:30 and before 14:00. Submit a market order for the next available tick; do not backdate entry to the pivot.
• For this untested experiment use one contract with sufficient capital and a stop 3 points above the newer pivot high. Set the target 6 points below the actual entry fill; this is not necessarily a 2R trade.
• Inspect when protection becomes active. If an entry gap is already beyond the stop, request exit and retain the trade. Keep the protective stop until flat.
• If still positioned at the close of the bar ending 15:50, request a market exit and record its fill. Include per-contract fees and tick-based slippage.

Make the comparison falsifiable

Test the divergence rule against a declared baseline, such as the same price-pivot condition without the delta filter. Use the same contract dates, session exclusions, position sizing and execution assumptions. Separate the dates used to select filters from the evaluation period, and report rejected variants. A total-volume filter changes the sample; it does not automatically improve accuracy. If tested, define the volume window using data available at confirmation. Record trade count, average gains and losses, drawdown, costs and missing-data coverage. There is no verified 62% or 71% result supplied for this experiment.

Match the platform to the data task

Spot FX feeds, crypto venues, stocks and futures differ in coverage. A vendor’s FX volume need not represent the entire decentralized market; one crypto exchange does not represent every venue. A futures print is not inherently institutional, and an equity print is not inherently noise. Choose the feed needed for the hypothesis rather than ranking markets by presumed participant identity. Pineify can draft supported Pine calculations. Compile in TradingView and compare inputs and selected trades with the intended definitions. Strategy Optimizer requires Advanced or above and searches supported exposed inputs; it does not supply missing order-book history. A live execution workflow also needs a configured connection, order controls and monitoring.