Block Trades Explained: How They Work and What They Signal
Ivan Royle
Block trades are single trades of at least 10,000 shares, the size FINRA uses as its working definition. They are how pension funds, hedge funds and private equity owners move large positions without walking the price against themselves, often in dark pools or through a bank that buys the whole block at once.
Block trades are also getting rarer. In the first nine months of 2026, trades of 10,000 or more shares and $200,000 or more made up just 3.73% of dark pool volume, down from 9.69% in 2017, according to FINRA data.
This guide is for self-directed investors who see "block trade" alerts or news of a big holder selling and want to know what it means for the stock. You will see how blocks are priced, a real $253.5 million block worked through line by line, and why a block print is a clue, not a signal.
Key takeaways
FINRA Rule 5270 treats a trade of 10,000 shares or more as "generally" a block, and smaller trades can count too. For small stocks, measure a block against average daily volume.
Large blocks sold by big holders are usually priced below the last trade. Torm's Oct 6, 2026 block was sold at $40.05, 2.44% below the prior close of $41.05.
Research on 5,625 upstairs blocks found 83% were seller-initiated. A block print does not tell you the buyer was the one in a hurry.
Blocks fell from 9.69% of dark pool volume in 2017 to 3.73% in 2026. Most institutions now slice big orders into small trades of about 79 shares.
What is a block trade?
A block trade is one large, privately negotiated trade in a stock, usually 10,000 shares or more. Instead of sending a huge order to an exchange, where everyone would see it, the buyer and seller agree on a size and price through a broker, a bank or a block-focused dark pool.
"A transaction involving 10,000 shares or more of an equity security ... is generally deemed to be a block transaction, although a transaction of fewer than 10,000 shares could be considered a block transaction."
— FINRA Rule 5270, Supplementary Material .03
Rule 5270 exists to stop front-running: a firm that knows a client's block order is coming may not trade ahead of it. That rule tells you why blocks matter. Information about a big order is valuable, so the whole system is built to keep it quiet until the trade is done.
How big is a block trade in practice?
Block size depends on the stock. FINRA's own statistics count trades of 10,000 shares or more, trades of $200,000 or more, or both. For a stock with average daily volume of 300,000 shares, a 10,000-share trade is 3.3% of a day's trading. For Nvidia, which traded about 224 million shares off exchange in the single week of Oct 19, 2026, it is routine. The useful measure is block size as a percentage of average daily volume (ADV).
Average daily volume (ADV) is the number of shares a stock trades on a typical day, usually averaged over 20 to 90 days. A block worth 100% of ADV is as big as a full day of normal trading.
How do block trades work?
Block trades work by matching a large buyer and seller away from the public order book, either through a venue that hides the order or a bank that takes on the risk. There are four common routes, and each leaves a different trail.
| Route | How it works | Who takes the risk | What the public sees |
|---|---|---|---|
| Algorithmic slicing | A broker's algorithm splits the order into thousands of small trades over hours or days | The fund, through price drift | Small prints that look like normal flow |
| Block-crossing dark pool | Venues such as Liquidnet or BIDS match large buy and sell orders anonymously | Nobody until a match appears | One large print on the tape, no names |
| Negotiated dealer block | A bank's desk finds the other side by phone or chat, or buys the block itself | The bank, if it commits capital | A large print, sometimes a news headline |
| Underwritten block sale | A bank buys registered shares from a big holder overnight and resells them to investors at a set price | The bank, for a fee | A press release and an SEC prospectus filing |
For retail investors, the underwritten route is the most useful one to study because it is the only one where the price, size, discount, seller and fee are all published. That is the case worked through below.
What does a real block trade look like? The Torm example
A real block trade shows exactly how the discount and fees work. On Oct 6, 2026, OCM Njord Holdings, an entity of funds managed by Oaktree, sold 6,329,874 shares of the shipping company Torm (TRMD) through J.P. Morgan as sole underwriter. The SEC prospectus (Form 424B7) lists every number.
- Size and price. 6,329,874 shares × $40.05 = $253,511,453.70.
- Discount. The last sale on Oct 5 was $41.05. ($41.05 − $40.05) ÷ $41.05 = 2.44% below market.
- Bank's fee. J.P. Morgan earned $0.205 a share, or 0.51% of the price: 6,329,874 × $0.205 = $1,297,624.17.
- Seller's proceeds. $40.05 − $0.205 = $39.845 a share, or $252,213,829.53 in total.
- Market reaction. Torm closed at $39.01 on Oct 6, according to Intrinio data: 4.97% below the prior close and 2.60% below the block price. Buyers of the block were briefly under water.
- A week later. On Oct 13, Torm closed at $41.125, back above both the block price and the pre-deal close. By Oct 16 it was at $38.21 again.
This was the second Oaktree block in a month. On Sep 14, 2026, the same seller sold 9,000,000 shares at $32.25, or $290,250,000, an 8.74% discount to the $35.34 last sale, with a fee of $0.50 a share (1.55%). Torm closed at $34.93 the next day, 8.3% above that block price. IFR reported that the October sale completed Oaktree's exit from the stock.
Why do sellers accept a discount?
Sellers accept a discount because selling 6.3 million shares on the open market would take days and push the price down anyway, while news of the selling leaked. A single block at a known price, delivered the next day (Oct 7 in Torm's case), removes that risk. The discount is the price of speed and certainty.
Why do block trades move prices?
Block trades move prices for two reasons: the market has to absorb a large amount of stock at once, and the trade may carry information. Research separates these into a temporary effect, which reverses, and a permanent effect, which does not.
Keim and Madhavan studied 5,625 upstairs block trades from 1985 to 1992. Of these, 4,688 were initiated by sellers and 937 by buyers. They found that prices often started moving before the block printed, a sign of information leaking as brokers shopped the trade, and that the temporary price impact grew less than proportionally as blocks got bigger.
"While prices go up on buys and down on sells, they revert after sells but remain high after buys, creating the permanent price impact asymmetry."
— Gideon Saar, "Price Impact Asymmetry of Block Trades: An Institutional Trading Explanation" (2000)
Holthausen, Leftwich and Mayers reported a similar asymmetry in 1987. A seller often has a reason that has nothing to do with the company, such as a fund closing or a private equity owner exiting. A buyer of a large block more often has a view. That fits Torm's October block: the selling was a known, one-time exit, and the price was back above the block level within a week.
Are block trades getting rarer?
Yes. Block trades now make up a much smaller share of dark pool volume than they did in 2017. FINRA's monthly statistics show trades of 10,000+ shares and $200,000+ fell from 9.69% of all dark pool shares in 2017 to 3.73% in January to September 2026.
Block trades' share of dark pool volume, 2017 to 2026
The trend reflects how big funds trade today. Algorithms slice large orders into small pieces, and the average dark pool trade fell from 210 shares in 2017 to 79 shares in 2026. In 2026, only about one dark pool trade in 9,500 was a block, and the average block was about 28,300 shares. FINRA counted 654,111 blocks in dark pools from January to September 2026.
This matters for anyone hunting blocks. A big print is now an exception, which makes it stand out, but it also means most institutional buying and selling never shows up as a block at all. Watching only blocks means missing most of the flow.
Where do block trades happen?
Block trades are concentrated in a few dark pools built for size. In September 2026, 33 dark pools reported 50.38 billion shares, but blocks made up 99.9% of Dealerweb's volume and 90.0% of Liquidnet Negotiation's, against 1.4% at Intelligent Cross.
Share of each dark pool's volume traded in blocks, September 2026
BIDS ATS handled the most block volume in absolute terms: 20.3% of all dark pool block shares in September 2026, across 15,695 blocks. Level ATS handled 13.3% and Dealerweb 10.4%. Dealerweb's average trade was 117,006 shares, while the average across all dark pools that month was 74.
The split shows two kinds of dark pools. Block venues match a few very large orders, often negotiated, while high-volume pools run by banks and electronic firms match millions of tiny ones. Our guide on what a dark pool is explains how the two types work.
Does a block trade mean smart money is buying?
No. A block trade has a buyer and a seller of exactly the same size, and the print alone does not tell you which side started it. In the research above, 83% of negotiated blocks were started by sellers, so the default guess for a big negotiated block should be "someone needed to sell".
This is the contrarian point of this guide. Many alert services call every large print "smart money buying". The Torm case shows the opposite: the biggest prints in the stock that autumn were a private equity owner leaving. Even TruthSayer's own block-based trades show that size alone does not decide the outcome.
Block size vs trade result: six Dark Pools AI alerts, 2026
WinLoss
The chart shows the six most recent closed Dark Pools AI trades that we could match to a published block size. The largest print, Sun Life Financial at 792% of ADV, lost 3.48%. A print at 130% of ADV in Relay Therapeutics gained 30.87%. Three won and three lost. What mattered was price confirmation and risk rules, not the size of the print.
A block print's price is not proof of buying. Negotiated blocks are priced by agreement between the two sides, so the print rarely shows which side was more eager.
How can you see block trades as an investor?
You can see block trades through free public filings and FINRA statistics, or through paid real-time feeds. The free sources are slower but tell you more about who and why.
| Source | What it shows | Timing | Cost | Best for |
|---|---|---|---|---|
| SEC Form 424B7 and press releases | Seller, size, price, discount, fee | Same day or next morning | Free | Underwritten blocks by named holders |
| Form 4 and Schedule 13D/G filings | Insider and 5%+ holder changes | 2 business days (Form 4) | Free | Seeing who sold or bought |
| FINRA ATS block statistics | Block volume per dark pool | About 1 month | Free | Market structure trends |
| Real-time print feeds | Every large print with size and price | Seconds | Paid | Active traders who filter themselves |
| TruthSayer Dark Pools AI | Filtered block alerts with entry, stop and exit | Within 1 to 2 hours of the signal | Free signals | Investors who want a tested rule set |
What does it cost to follow block trades?
SEC filings and FINRA statistics are free. Real-time print feeds are paid because they rely on licensed consolidated tape data. TruthSayer's signals are free; the app's AI chat costs $10.99 a month or $71.88 a year after a 7-day trial.
Which source is best for you?
- Long-term investors: 424B7 filings and Form 4s. They tell you who sold and why, which matters more over months than a single print.
- Swing traders holding weeks: filtered real-time alerts with a clear stop, because the edge, if any, fades within days.
- Not a fit: day traders who need order book data, and anyone who wants to copy a print without a plan for the exit.
How do you read a block trade, and what are the risks?
Read a block trade by asking how big it is relative to normal volume, who was likely behind it and what price did next. This table sorts common readings from weak to strong, from a buyer's point of view.
| Clue | Weak clue for buyers | Worth a closer look | Stronger clue for buyers |
|---|---|---|---|
| Size vs ADV | Under 10% of ADV | 10% to 100% of ADV | Above 100% of ADV |
| Known seller | Fund exit or lockup expiry | Unknown seller | Insider or activist buying, filed on Form 4 or 13D |
| Price vs prior close | Discount above 5% | Discount of 1% to 5% | At or above the prior close |
| Price over 2 to 3 days | Falls back to the block price | Flat | Holds away from the block price on rising volume |
| Timing | Index rebalance or quarter end | Near earnings | No scheduled event |
These thresholds are our rules of thumb from the data in this article, not fixed laws. The main risks of trading around blocks:
- Wrong side. Most negotiated blocks are sales, so buying because "a block printed" can mean buying from a motivated seller.
- Follow-on supply. A holder who sold once may sell again. Oaktree sold Torm blocks on Sep 14 and Oct 6, 2026.
- Short-lived moves. Torm fell 4.97% on its block day and recovered within a week, then fell again. Short windows are noisy.
- Survivorship. Alert services that show only winning prints hide the losses. Ask for the full list, as our guide to judging a signal service explains.
How should you judge a block trade? The TruthSayer Block Trade Test
The TruthSayer Block Trade Test is a five-question check before you act on any block trade headline or alert. If you cannot answer at least three, treat the block as noise.
- How big is it relative to ADV? Under 10% is routine. Above 100% is unusual.
- Who was the likely seller? Check for a 424B7, a press release, a Form 4 or a known fund exit.
- What was the discount? A deep discount points to a motivated seller; a print at market points to balanced demand.
- Is more supply coming? Look for remaining stakes, lockup expiries and repeat sellers.
- What did price do next? Wait two to three days. Act only if price holds away from the block level, with a stop in place.
Run on Torm's October block, the test gives a clear answer: a known seller completing an exit, a 2.44% discount, no stake left to sell and a price that went back to the block level within days. That is supply, not a buy signal.
How does TruthSayer AI trade block trades?
TruthSayer AI's Dark Pools AI scans for unusually large institutional transactions, filters for statistically significant activity and confirms price direction before it trades. Every signal is executed in TruthSayer's own funded account before the alert goes out, with a 5% position size and a trailing stop.
A real example: on Mar 10, 2026, Dark Pools AI flagged one dark pool trade of 3,070,000 shares in Relay Therapeutics (RLAY), 130% of its average daily volume and roughly $33 million at the entry price. The model went long at $10.83 with a 20% trailing stop.
- Position. 150 shares × $10.83 = $1,624.50 invested.
- Exit. Price hit the 3× ATR trailing stop on Apr 28 at $14.24: 150 × $14.24 = $2,136.00.
- Result. $2,136.00 − $1,624.50 = $511.50 gross. After a $9.98 commission, $501.52 net, or +30.87% in 49 days.
The same rules also produced losses. A short in ON Semiconductor after a 3,140,000-share block lost 13.95%, and a long in Mirum Pharmaceuticals after a block of 275% of ADV lost 15.41%. All of these trades are in the public trade list.
As of Nov 11, 2026, the Dark Pools AI index shows a 33.62% return since inception against 29.44% for the S&P 500, a 67.3% win rate and a Sharpe ratio of 2.78. TruthSayer reports performance under the GIPS standards; our guide on GIPS compliance explains what that checks.
The six trades in the chart are the most recent closed Dark Pools AI trades whose alerts state both the block size and its percentage of average daily volume. We included all of them, wins and losses. The Torm example was chosen because its SEC filing publishes the seller, discount and fee in full.
What should you do next?
Next time you see a block trade headline, run the TruthSayer Block Trade Test before you act. Most blocks will fail it, and that saves you from buying someone else's exit.
- Learn how to read FINRA's weekly and daily files in dark pool data: where it comes from and how to read it.
- Read what a dark pool is for how block venues fit into the market.
- Use our checklist on how to judge a signal service before paying for block alerts.
- See live block alerts with entries, stops and exits, losses included, on the Dark Pools AI page.
The bottom line: Block trades are real evidence that a large holder moved a lot of stock, but not of which way the smart money leaned. Most negotiated blocks are sales, blocks are now just 3.73% of dark pool volume, and price after the print matters more than the print. Use the TruthSayer Block Trade Test: size against ADV, likely seller, discount, future supply and price follow-through.
Frequently asked questions
How many shares is a block trade?
FINRA Rule 5270 says a trade of 10,000 shares or more is generally a block, and smaller trades can qualify. FINRA's dark pool statistics also count trades of $200,000 or more. For small stocks, compare size with average daily volume; a block above 100% of ADV equals a full normal day.
Are block trades bullish or bearish?
Neither by default. Every block has a buyer and a seller. Research on 5,625 upstairs blocks found 83% were seller-initiated, and prices tended to revert after sales but stay higher after buys. Check who sold, the discount and price over the next two to three days.
Why are block trades sold at a discount?
The buyer, often a bank, takes on the risk of reselling millions of shares quickly. In Torm's Oct 6, 2026 block, 6,329,874 shares sold at $40.05, 2.44% below the $41.05 prior close, and the bank earned $0.205 a share. Sellers accept that cost for speed and certainty.
Should I sell a stock I own after a big block sale?
Not automatically. Find out who sold and whether more stock is coming. Torm fell 4.97% on its Oct 6, 2026 block day, then closed at $41.125 on Oct 13, above the pre-deal price. A one-time fund exit differs from an insider selling, which shows on Form 4 within two business days.
How quickly do block trade price moves fade?
Seller-driven moves often fade within days. Torm recovered its full 4.97% block-day drop within a week in October 2026. Academic studies found temporary impact reverses after sales, while buyer-initiated blocks keep more of their move. Judge a block after two to three trading days, not in the first hour.
Where can I see block trades for free?
SEC EDGAR shows underwritten block sales on Form 424B7, usually by the next morning, with seller, price and fee. FINRA publishes monthly block statistics for each dark pool, about one month late; September 2026 data appeared on Nov 2. TruthSayer's Dark Pools AI alerts are also free.
Sources
- FINRA Rule 5270, Front Running of Block Transactions (retrieved Nov 11, 2026)
- FINRA, OTC Transparency data: ATS block statistics (monthly, January 2017 to September 2026; September 2026 published Nov 2, 2026)
- Torm plc, Prospectus Supplement (Form 424B7), 6,329,874 shares (Oct 6, 2026)
- Torm plc, Prospectus Supplement (Form 424B7), 9,000,000 shares (Sep 14, 2026)
- IFR, "Oaktree liquidates Torm stake via US$253.5m block" (Oct 9, 2026)
- Keim, D. and Madhavan, A., "The Upstairs Market for Large-Block Transactions," Review of Financial Studies (1996)
- Saar, G., "Price Impact Asymmetry of Block Trades: An Institutional Trading Explanation," SSRN (2000)
- Holthausen, R., Leftwich, R. and Mayers, D., "The effect of large block transactions on security prices," Journal of Financial Economics 19(2) (1987)
- Investor.gov, "Insider Transactions and Forms 3, 4, and 5" (Jan 26, 2021)
- Intrinio daily and 15-minute price history for Torm (TRMD), September to October 2026, via TruthSayer AI's market data feed
- TruthSayer AI, Dark Pools AI performance, trade list and alerts (Nov 11, 2026)
This article is for educational purposes only and is not investment advice. Past performance does not guarantee future results.
Ivan Royle
Ivan Royle is Co-founder and Chief Revenue Officer of TruthSayer AI, bringing approximately 25 years of leadership experience across financial services, investment management, and wealth management. His background includes serving as Global CMO of BNY Mellon Investment Services, a $10 billion revenue business, and VP of Corporate Communications at GE Capital Europe.
