Dark pool prints on a small-cap tape: off-exchange trades and what they show
Much small-cap volume executes off-exchange and is reported to a FINRA facility after the fact. Those prints are real trades, but on their own they carry little information about direction.
Start in plain English
On a small-cap tape, many prints carry a venue code that means "reported off-exchange". Traders call them dark pool prints, though most of them come from other sources. A large one, or a cluster, often gets read as "institutions buying".
The core idea: off-exchange prints are real trades, reported after they happen, and on their own they say little about who is buying or what comes next. A large share of US stock volume executes away from the exchanges, mostly through wholesalers that fill retail orders and through alternative trading systems (ATSs, the dark pools). The tape shows that a trade happened, its price, size and time. It does not show which side started it, whether it was one order or a piece of a larger one, or whether it was a hedge.
US Day Trading Fundamentals covers the Time & Sales window in Time & Sales; Tape reading covers reading prints against the quote.
Two paths from the same start
All prices, sizes and costs in this example are Illustrative. Ostrander Mining is fictional.
The setup at 13:41:20 ET. Ostrander has drifted between $3.50 and $3.62 since noon. In 70 seconds the tape shows three large off-exchange prints, 50,000, 75,000 and 40,000 shares, all three at sub-penny prices between the bid and the offer, two of them at $3.5875, just under the $3.59 offer. Chat posts "dark pool buying". The quote is $3.58 × $3.59.
Ostrander Mining (fictional), Time & Sales to 13:41:20, ET (Illustrative)
| Time | Price | Size | Side | Venue |
|---|---|---|---|---|
| 13:40:10 | 3.58 | 300 | bid | exchange |
| 13:40:12 | 3.5875 | 50,000 | between | TRF |
| 13:40:31 | 3.59 | 200 | ask | exchange |
| 13:40:44 | 3.58 | 500 | bid | exchange |
| 13:40:55 | 3.5875 | 75,000 | between | TRF |
| 13:41:02 | 3.59 | 400 | ask | exchange |
| 13:41:15 | 3.585 | 40,000 | between | TRF |
| 13:41:18 | 3.58 | 100 | bid | exchange |
side = print price against the quote (bid, ask or between); venue: exchange = exchange print, TRF = reported off-exchange to a FINRA facility (market center D). Range since noon $3.50–$3.62. Prices at or above $1 with more than two decimals ($3.5875, $3.585) are sub-penny trades.
The plan, written before 13:41:20. Buy with a marketable limit order (ceiling $3.63), filled at $3.62 after the offer lifts. Stop $3.53. Target $3.80 (2R). Time stop 14:00. Planned risk $0.09 per share. Size: $49.50 ÷ $0.09 = 550 shares. Costs: $0.005 per share round trip (Illustrative).
Path AContinuation
Exchange prints start hitting the offer. Price clears $3.62, trades $3.70 by 13:46 and $3.81 at 13:52. The $3.80 limit sell fills.
Path BNothing behind it, then a fade
No further large prints appear. Price slips back into the range and below it. The $3.53 stop triggers and fills at $3.52.
Path CNothing at all
Price falls back into the noon range ($3.50–$3.62) but holds between $3.55 and $3.62, above the stop. At 14:00 the time stop exits at $3.61.
The review note is the same for all three: "Entered after a cluster of large off-exchange prints; plan followed; 1R planned." In Path A the prints get called "smart money"; in Path B, "a hedge" or "a cross". The label arrived with the outcome. The rest of this lesson covers where these prints come from, how they are reported, and what public data says about them, with a delay.
Where off-exchange volume comes from
In June 2026 the SEC wrote that off-exchange trading has, since the end of 2024, been "regularly exceeding 50% of overall volume", citing staff data showing 51.9% for Nasdaq-listed stocks and 47% for NYSE-listed stocks as of January 2026 (SEC Release No. 34-105655). The main sources:
- Wholesalers. Retail brokers route most of their customers' marketable orders to a small group of off-exchange market makers, who fill them from their own inventory, often at prices slightly better than the quote. In 2022 the SEC put the share above 90% of individual investors' marketable orders.
- Alternative trading systems (ATSs). Venues registered with the SEC that match orders without displaying them publicly. These are the "dark pools" in the strict sense.
- Other broker-dealers. Firms filling client orders against their own inventory, or crossing two clients.
For a small cap with heavy retail interest, a large part of the tape can be wholesaler fills of retail orders, which by definition are many unrelated people.
How the prints reach your tape
Trades executed off-exchange in NMS stocks are reported to a FINRA facility: one of the Trade Reporting Facilities (the FINRA/Nasdaq TRF Carteret, the FINRA/Nasdaq TRF Chicago and the FINRA/NYSE TRF) or the Alternative Display Facility. FINRA rules require reporting as soon as practicable, and no later than 10 seconds after execution. On the consolidated feed for Nasdaq-listed stocks, these trades carry the market center code D for FINRA (FINRA and UTP sources, checked 2026-10-05).
What that means on your screen:
- Venue, not counterparty. "D" tells you the trade was reported off-exchange. It does not say which wholesaler or ATS, or who the customer was.
- Timing. A print can appear up to 10 seconds after execution, and some carry condition codes for late or out-of-sequence reports. A "large print at 13:40:55" may have executed slightly earlier.
- Sub-penny prices. Rule 612 limits the increments of quotes and orders, but, in the SEC's words, "it does not restrict trading in sub-penny increments." Wholesaler price improvement and midpoint executions produce prints like $3.5875.
What public data shows, with a delay
FINRA publishes off-exchange volume by venue through its OTC Transparency data. Under FINRA Rule 6110, weekly ATS volume for Tier 1 NMS stocks is published no earlier than two weeks after the week ends, and four weeks for other NMS stocks; non-ATS volume is published monthly (checked 2026-10-05). Most small caps fall outside Tier 1, so their ATS data arrives about a month later.
That data answers "how much traded in dark pools last month", not "who is buying now". Daily short-sale volume files are a different dataset again; FennEdge's article Short interest vs. short volume explains why they do not show open short positions.
Reading a large print without a story
When a large off-exchange print appears, four questions keep the reading honest:
- How large is it against the stock's normal minute volume? 50,000 shares in a stock trading 3 million shares a minute is routine; in one trading 30,000 a minute it is notable.
- Where did it print against the quote? At the offer, at the bid, or between them.
- Is it alone or repeated? One print can be a single client cross. Repeated prints over an hour are a different observation.
- What did the lit market do next? Exchange prints and quote changes after the cluster are observable; the cluster's intent is not.
These questions describe the print. None of them turns it into a direction.
Observe vs. infer
Go back to 13:41:20 in the Ostrander example (fictional; all figures Illustrative).
| What the tape shows (observed) | What a trader might infer | Why it is not established |
|---|---|---|
| Three off-exchange prints totalling 165,000 shares in 70 seconds | "Institutions are buying" | Fits a wholesaler filling many retail orders, a crossed block between two clients, or a hedge |
| Two prints at $3.5875, just under the $3.59 offer | "Buyers paid up" | Sub-penny prices fit price improvement for either side |
| The large prints show venue TRF (market center code D) | "It came from a dark pool" | TRF and code D mean reported to a FINRA facility; most such volume is wholesaler fills, not ATS trades |
| The exchange quote did not move | "Hidden demand is waiting" | An off-exchange trade does not touch the displayed book |
Only the left column was usable at 13:41:20.
Odds box
Close to zero on purpose. If off-exchange size carried the directional information chat often assumes, a count would show it. Until someone counts, the cluster is a starting condition, not a signal.
What shifts the odds
These are hypotheses to record, not findings:
- Size relative to the stock's average trade size and minute volume.
- Price relative to the quote. At the offer, at the bid, at the midpoint.
- Repetition. One cluster vs. repeated clusters over an hour.
- Exchange prints that follow. Whether lit-market buying lifts the offer afterward.
- News timing. Large prints around a headline vs. in a quiet afternoon.
Cost of being wrong
Using the Ostrander plan at 550 shares (Illustrative):
| Outcome | Exit | R after costs | Dollars after costs |
|---|---|---|---|
| Path A: target | $3.80 | +1.94R | +$96.25 |
| Path B: stop, 1¢ slippage | $3.52 | −1.17R | −$57.75 |
| Path C: time stop | $3.61 | −0.17R | −$8.25 |
| Path B, thin afternoon book, 5¢ slippage | $3.48 | −1.61R | −$79.75 |
Afternoon books in small caps can be thinner than morning ones, so the slippage row deserves weight.
How many trades before this means anything
To test "large off-exchange prints matter" you need two counts under the same plan: entries after clusters and entries without them. With 20 in each group, 30% and 40% Path A shares have overlapping 95% intervals (Wilson, about 15%–52% and 22%–61%, computed on hypothetical counts). See How many trades before your numbers mean anything.
What this lesson does not claim
- It does not claim off-exchange prints show institutional buying or selling.
- It does not claim dark pools are harmful or helpful. It describes how their trades appear on a retail tape.
- Market-structure figures are from SEC and FINRA sources as dated; they change over time.
- Ostrander Mining and its prints are fictional and Illustrative.
Go deeper
- FennEdge first. FennEdge does not offer a tick-by-tick tape or venue breakdown as of 2026-10-05. The Scanner (paid: Plus, Pro, Max) triggers on rapid rises and drops, regular-session highs and Nasdaq-listed LULD halts. The article Short interest vs. short volume covers the related public datasets.
- Wholesaler internalization and routing. Why your own retail order may not reach an exchange: Order routing.
- Sub-penny rules. Tick size and sub-penny rules.
- Related: Market maker Level 2, Iceberg orders.
Exercises
Five exercises follow this lesson: three concept checks, one replay of a static tape (classify observation vs. inference at 13:41:20, then Path B is revealed), and one scenario on reading a cluster of large prints. Exercises, answers and progress are available after signing in.
Primary sources
- SEC. Release No. 34-105655 (June 11, 2026), proposed rescission of Rules 611 and 610(e), including off-exchange volume data.
- SEC. Release No. 34-101070 (September 18, 2024), on minimum pricing increments and sub-penny trading.
- SEC. Release No. 34-96495 (December 14, 2022), proposed Order Competition Rule, later withdrawn; cited for its 2022 routing figures.
- FINRA. Rule 6380A, Trade Reporting Facility and Rule 6110.
- UTP Plan. Binary output specification (market center codes).
Sources checked 2026-10-05. Ostrander Mining and all prints, outcome splits and costs in this lesson are fictional and Illustrative. This lesson is education, not a recommendation to buy or sell any security. Past patterns do not indicate future results.