Level 2 vs. Time & Sales: displayed orders are not completed trades

Two windows answer different questions

Start with Bid, Ask and Spread: quotes describe offers to trade; the last price describes a reported execution. Different updates within the same second need not concern the same event.

First establish the coverage of your depth subscription

Trading platforms commonly use “Level 2” for a depth window, but feeds and subscriptions differ. Nasdaq’s TotalView description concerns displayed depth on the Nasdaq book. A NYSE-listed stock can trade on Nasdaq; that does not turn one venue’s book into a complete sum of every venue.

Before comparing terminals, check instrument, venue coverage, real-time or delayed status, aggregation, quantity units and filters. One interface may show shares while another shows lots or hundreds. Similar-looking numbers can represent different settings.

Displayed depth does not include every hidden order or future intention. A thin visible level alone cannot establish how quickly a larger order would complete. Routing, queue position, changing quotes and broker conditions also matter.

A 10,000-share bid shrinks: what can you actually conclude?

Suppose a covered venue displays 10,000 shares at $5.00 at 10:02:15 ET. It later shows 2,000, while the tape reports 3,000 shares at $5.00. Record two observations: displayed size decreased by 8,000, and the feed reported a 3,000-share trade.

Do not conclude that the other 5,000 were necessarily canceled. A row may aggregate orders, a print may involve another venue, and updates may arrive asynchronously. Strict attribution needs sufficiently detailed order events and timestamps; an ordinary screenshot usually lacks that evidence.

Tape colors do not identify buyers, sellers or short positions

Every completed trade has both a buyer and a seller. A terminal may color prints relative to contemporaneous Bid/Ask or the preceding price; read its actual convention. Green does not mean only buying occurred, and red does not establish a newly opened short position. Quote and trade arrival timing can affect classification.

Conditions matter too. DAS’s Time & Sales flag documentation describes records that can count toward volume without qualifying for high, low or last-price statistics. For a print far from the market, inspect conditions, reporting time and corrections before treating it as an executable price.

Use the DAS guide or DAS scripting course to keep market observation separate from order actions. A visible price, submitted order, accepted order and fill are distinct states.

Use a repeatable observation sequence

Record instrument and ET time, then Bid/Ask, feed coverage and delay status. Observe size changes, check contemporaneous prints and their conditions, and leave unresolvable attribution explicitly unknown. Avoid replacing missing evidence with a story about trader intent.

After FennEdge Scanner surfaces activity, add source context in FennEdge Research and event notes in FennEdge Desk. Check depth and tape in your actual market-data terminal; this research workflow does not imply FennEdge supplies every venue’s depth subscription.

For review in FennEdge Journal, preserve actual order responses: accepted, partially filled or unfilled. Compare them with order-type mechanics to explain execution rather than relying on “the book looked strong.”

Three common beginner questions

Does a large bid guarantee support? No; displayed size can change or cancel. Does a large print identify an institution? Price and size alone generally do not establish identity. Does disagreement prove a software bug? First reconcile coverage, units, filters, conditions and update timing, then investigate genuine anomalies.