Distribution: Selling Into Strength

Michael Whitman · Trading since 2017, full-time since 2020 · Published 2026-09-06 · Updated 2026-10-04 · Reviewed 2026-10-04

Covered on this page: TradingView.

A large holder cannot sell a big position in one order without pushing the price down through its own remaining shares. Distribution is the name for doing it slowly, near a high, into whatever buying arrives. The word also has a second financial meaning: a fund’s payout of income or capital gains. This page is about the first one.

How it works

Distribution is a large seller working out of a position. Not one order and not one day, but a supply of stock released gradually into demand.

It is the mirror of accumulation. Same mechanics, opposite direction, and the opposite end of a move.

The visible result is a sideways range near a high. Buying keeps arriving and keeps being met, so price stalls instead of falling, which is why the pattern looks like an ordinary pause while it is happening.

The Wyckoff framework adds two marks. A push above the top of the range that fails and closes back inside, the upthrust, and heavier participation on falling days than on rising ones. Both are claims about evidence, so both can be counted.

Distribution or a pause: the same shape

The range itself carries no label. A month of sideways trading under a high precedes both a further rise and a top, and nothing in the price pattern says which is underway until one boundary goes.

The ownership story cannot be seen. Large orders are worked through algorithms built not to show, and every share sold in the range was also bought, so the transfer of stock has to be inferred.

That leaves the countable parts. How wide the range is, how close it sits to the high, where the latest close sits inside it, the share of trading on down days and whether pushes above the top failed. Each of those has a number, and the section below measures them.

A worked example

SPY in early 2026 looked like the textbook. For the 20 sessions from 12 December 2025 to 12 January 2026 it traded between a low of $671.20 (17 December) and a high of $696.09, set on 12 January, which was also its highest price of the previous 52 weeks. That range was 3.71% wide: ($696.09 − $671.20) ÷ $671.20.

Then it pushed above the top seven times without closing above it. On 27, 28 and 29 January, 2 and 3 February and 10 and 11 February the high cleared $696.09 (the highest was $697.84 on 28 January) and the close finished back under it. In Wyckoff terms each one is an upthrust.

The volume was even. Inside the 20 sessions that formed the range there were 10 down closes and 10 up closes, and the down days carried 51.31% of the shares traded on the two.

On 12 March it closed at $666.06, below the $671.20 floor. That was 41 sessions after the range was first in place. Only from that close onward does the word distribution describe these weeks. A month earlier the same chart was also a range near a record that might have broken upward, as most did.

The original data

The sample. SPY daily bars from Yahoo Finance, highs, lows, closes and volume as traded, January 1993 to 2 October 2026; ranges can start from January 1994, once a full year of history exists. Opens are not used.

The rule, fixed before counting. A range is the 20 sessions ending on a given day, from the highest high to the lowest low, no more than 4% wide. It counts as near a high when its top is within 2% of the highest high of the 252 sessions ending that day. Once it qualifies the boundaries are frozen; it resolves on the first later close above the top or below the bottom. The search resumes the session after a resolution, so one range can share sessions with the one before. Down-day volume share is volume on down closes divided by volume on up and down closes in the 20 sessions. A failed push is any session after the range forms, and before the close that resolves it, whose high clears the top while the close does not.

How the ranges resolved. The rule found 237 tight ranges near a 52-week high; one, formed on 17 September 2026, was still open on 2 October. Of the other 236, 172 (72.88%) broke upward first and 64 (27.12%) downward. The median range was 3.50% wide and resolved 4 sessions after it qualified. Tight ranges away from a high were rare: 31, of which 26 broke upward, too few to compare reliably (Fisher p = 0.27).

After a downward break. The lowest close in the next 60 sessions sat a median 6.43% below the old range top, and 20.28% below it in the worst case.

Position inside the range came first. Where the close sat on the day the range qualified, as a share of the way from bottom to top, split the outcomes sharply:

Heavier selling volume looked like a sign, until position was allowed for. Ranges with more volume on down days than up days broke downward 27 times in 49 (55.10%), against 37 in 187 (19.79%) for the rest. But those ranges had usually drifted lower inside themselves: their median close sat 52% of the way up, against 89% for the rest. Within each third the gap mostly closed. In the top third, 4 of 17 down-heavy ranges broke downward against 23 of 161 others, 23.53% against 14.29% (p = 0.30); in the middle third, 7 of 14 against 7 of 18 (p = 0.72).

Failed pushes above the high pointed the other way. Of 73 ranges with at least one, 10 (13.70%) broke downward, against 54 of 163 (33.13%) without. In the top third, where 63 of those ranges sat, the rates were 9 of 63 and 18 of 115, 14.29% and 15.65% (p = 1.00). The other ten broke downward once.

This is one index, one rule and a few hundred ranges. A broad fund made of hundreds of stocks is a hard place for one seller’s footprint to show, and a single stock might behave differently. Every range, with its dates, levels, failed pushes and volume share, is in the SPY ranges file.

The other meaning: a fund’s distribution

Funds also make distributions. A mutual fund or ETF passes the dividends, interest and realized gains it collects on to its holders, usually on a schedule, and the fund’s value per share falls by about the amount paid. That meaning has nothing to do with a range near a high. Fund payouts are covered with dividends; none of the figures above measure them.

When it fails

The base rate is the whole problem. On SPY nearly three tight ranges in four near a high broke upward, so calling each one distribution produces a run of early short positions, each with a story attached.

The second failure is the retrospective read. After a fall, every earlier range looks like distribution, and the ones that broke upward are forgotten.

A third is treating a failed push as confirmation. In this sample ranges with one were no more likely to break downward, and in the lower two thirds less likely.

A fourth is reading volume without position. Down-heavy volume mostly told you price had already slipped toward the floor, which the chart showed anyway.

And a fifth is shorting inside the range. Until the floor gives way the range is intact, and the early-2026 example held its floor for 41 sessions, through seven failed pushes at its top, before it broke.

Related

Accumulation is the same process at the other end of a move, with its own count of how post-fall ranges resolved. The Wyckoff method is the framework both phases come from. And the Wyckoff upthrust covers the failed push above the high in detail.

This page is educational, not financial advice. Test every idea on your own charts before risking money. Some links on this page earn a commission if you buy through them. It costs you nothing and it does not decide what appears here or in what order — how these pages are made is set out in our methodology.