Level 2 and tape reading are the two real-time order-flow tools scalpers and short-term traders use to see intent and execution beneath the last price. Level 2 (market depth / the order book) shows the resting limit orders stacked on the bid and ask at multiple price levels — what people are willing to do. The tape (Time & Sales) shows every executed print — size, price, time — i.e. what people actually did. The core tension is that the book is a statement of intent that can be faked or pulled in milliseconds, while the tape is the ground truth of who won. A scalper reads them together: Level 2 frames the supply/demand landscape, the tape confirms whether that landscape is real.

How it's read

Level 2 anatomy. Two stacked columns: bids (buyers) sorted high-to-low, asks/offers (sellers) sorted low-to-high. The top of each is the inside bid/ask (the NBBO — National Best Bid and Offer). Each row shows price, aggregate size, and, on some montages, the venue or market-maker ID (a four-letter MPID such as NSDQ, ARCA, or a participant code). Size is usually shown in round lots (1 lot = 100 shares); confirm your platform's units. Per Center Point and Lightspeed, the MPID lets a trader spot which venue/participant is repeatedly providing or pulling liquidity in a given name.

The tape (Time & Sales). A scrolling log of executions, color-coded by most platforms: prints at/above the ask are aggressive buys (green), prints at/below the bid are aggressive sells (red), prints between are mid/neutral. The three fields are price, size, time. Reading the tape is reading the speed, size, and location of prints, not any single line.

How it's used in practice

• Order-flow imbalance. The single most-watched read: compare resting size on the bid vs the ask, and aggressive prints hitting each side. Markedly more size and aggressive buying on the bid implies near-term upward pressure; the reverse for the ask. This is the practitioner version of the academically documented Order Flow Imbalance (OFI) effect (Cont, Kukanov & Stoikov, 2014).
• Support / resistance from size. An unusually large resting order ("a wall") marks a price where a big participant is defending. Scalpers lean long just above a real bid wall (tight stop below it) and fade into ask walls — if the wall holds when tested.
• Absorption. Price sits at a level, the tape prints heavy aggressive selling, yet price does not drop — a hidden/iceberg buyer is absorbing the supply. This is one of the higher-conviction tape signals: aggressive sellers exhausting themselves into a buyer often precedes a bounce. The mirror (heavy buying that can't lift price) signals distribution.
• Speed/tempo of the tape. Acceleration of prints (size and frequency) signals momentum and conviction; a fast green tape into a breakout confirms it. A tape that goes quiet after a move warns the move is stalling.
• Confirmation rule. The disciplined workflow is book frames, tape confirms. A bid wall is only a buy thesis once the tape shows it actually absorbing sells; a breakout level is only valid once the tape shows aggressive prints lifting the offers, not a single 100-share fill.

Standing & evidence

The honest landscape splits into two very different claims.

The microstructure effect is real and documented. Cont, Kukanov & Stoikov ("The Price Impact of Order Book Events," Journal of Financial Econometrics, 2014; 50 NYSE stocks) found that short-horizon price changes are driven mainly by order-flow imbalance at the best bid/ask, with a near-linear relation whose slope is inversely proportional to market depth, robust across stocks and time scales. The predictive content, however, is very short-horizon — seconds to tens of seconds — and decays fast. This is why order flow is genuinely useful to scalpers and HFT and largely irrelevant to multi-day swing positions.

Discretionary "tape reading as edge" is contested and harder to defend. No peer-reviewed evidence shows that manual tape reading produces durable retail profits, and several structural forces have eroded it: (1) decimalization (2001) thinned displayed size and made the book noisier; (2) HFT posts and cancels orders in microseconds, so a human's view of the book is stale on arrival; (3) hidden and iceberg orders mean "not all orders show up in Level 2" (Center Point); and (4) spoofing/layering — placing large non-bona-fide orders to fake imbalance, then canceling — is explicitly illegal precisely because it manipulates these displays. In futures it was criminalized by Dodd-Frank §747 (CEA §4c(a)(5), which names "creating an appearance of false market depth"), the basis for the first criminal spoofing conviction, U.S. v. Coscia, 2015; in equities the prohibition runs through the Securities Exchange Act (e.g. §9(a)(2), §10(b)) rather than §747. Practitioners detect it by watching a wall vanish the instant the tape approaches it. Treat the book as probabilistic and adversarial, never as fact.

Strengths & limitations

Strengths. For scalping and intraday execution, order flow is the highest-resolution signal available — it leads price on the second-to-minute horizon (Cont et al.), and absorption/exhaustion reads have no equivalent on a chart. Even for swing traders, the tape adds execution value: timing a fill, sizing into a level, or bailing when promised support evaporates.

Limitations / failure modes. (1) Spoofing — the #1 misuse is trusting a wall at face value; size you can see is the size most easily faked. (2) Stale view — in fast or low-liquidity names the displayed book lags real fills. (3) Hidden liquidity — icebergs make the book understate true depth. (4) Horizon mismatch — the edge is seconds-long; using Level 2 to forecast tomorrow is a category error. (5) Over-trading — the constant flicker invites impulsive entries; it is an execution and scalping tool, not a thesis generator. Works best in liquid, high-volume, tight-spread names; near-useless in thin small-caps where one participant can paint the entire book.

System relevance

Augustus is a swing / short-term setup agent operating mostly on the daily-to-multi-day horizon, where Level 2's seconds-long predictive edge has decayed to noise — so this node should not feed multi-day directional theses. Its legitimate role is execution context at the moment of entry/exit: confirming that a chart-derived support level shows real bid absorption before sizing in, or warning that displayed depth is suspiciously thin/spoofy. This complements sibling scalping nodes and the broader order-flow / market-microstructure branch; for the directional setup itself, defer to the chart-pattern and swing-application nodes. Hard caveat for the agent: never treat resting book size as committed liquidity — it is intent, not execution, and is a documented vector for manipulation (spoofing/layering).

Sources

• Cont, R., Kukanov, A., Stoikov, S. — The Price Impact of Order Book Events, Journal of Financial Econometrics 12(1), 2014 (SSRN 1712822; arXiv:1011.6402). Order-flow-imbalance / price-impact evidence and its short horizon.
• Center Point Securities — How to Interpret Level 2 Data (MPIDs, hidden orders, spoofing caveats, "no perfect way to interpret").
• Lightspeed — How to Read Level 2 Market Data (book anatomy, market-maker IDs).
• Warrior Trading — Time and Sales / Tape Reading and Level 2 Definition (color-coding, practical reads).
• CFTC / Dodd-Frank §747 amending CEA §4c(a)(5) — statutory spoofing prohibition (futures), incl. "false market depth" language; U.S. v. Coscia (N.D. Ill. 2015, aff'd 7th Cir.) — first criminal spoofing conviction, a futures case. Equities spoofing prohibited under Securities Exchange Act §9/§10(b). (Sources: CFTC fact sheet; Morgan Lewis / Davis Polk client alerts; Wikipedia "Spoofing (finance)".)
• Bookmap — Cracking the Spoofing Code and Beyond Level II (HFT noise, spoofing detection via tape).

Dispute flag: the microstructure OFI effect is well-evidenced but very short-horizon; "manual tape reading as a durable retail edge" is folklore-heavy and not academically established — the two should not borrow each other's credibility.