Accumulation Distribution vs OBV
Michael Whitman
·
Trading since 2017, full-time since 2020
·
Published 2026-09-08
·
Updated
·
Reviewed
On-balance volume adds a bar's entire volume to one side based only on whether the close was higher or lower than the previous close. The accumulation distribution line weights that volume by where the close sat within the bar's own range, which is the single difference between them.
Two running totals built from the same input. Both accumulate volume and both are read by direction rather than by level. The one thing that differs is how each bar’s volume gets allocated, and on thirty-two years of SPY that one difference was enough to send the two lines to very different places.
What each one is
On-balance volume adds the whole bar’s volume when the close is higher and subtracts it when lower. Joe Granville first explained it in his 1963 book Granville’s New Key to Stock Market Profits. The size of the move is irrelevant; only its sign counts, and an unchanged close adds nothing. On balance volume covers it.
The accumulation distribution line weights the volume by where the close sat inside the bar. Marc Chaikin’s money flow multiplier is (close − low) minus (high − close), divided by the high-low range, so it runs from +1 at the high to −1 at the low and is zero at the midpoint. Accumulation distribution covers that calculation.
Neither has a meaningful level. Both start from an arbitrary point, so only the slope carries information, in exactly the same way for both.
Where they differ
How a bar is allocated. A bar closing a fraction higher gives all its volume to the up side on one and, if that close sat near the middle of the range, almost none on the other. On SPY the typical multiplier was a bit under 0.6, so the accumulation line counted 56.92% of all volume and on-balance volume counted all of it.
How gaps are handled. On-balance volume compares to the previous close, so a gap is fully reflected. The accumulation line looks only inside the bar, so a bar that gapped up and closed near its low counts as selling.
Where they disagree. On days that closed up but in the lower half of the bar, and days that closed down but in the upper half. That is a genuine difference in what is being measured, not noise.
Which is blunter. On-balance volume, by a wide margin, and that bluntness is also why it is easier to reason about.
Why the two lines split on a gap day
23 March 2026 is the plain case. SPY had closed at $648.57. The next session never traded that low: its range ran from $653.94 to $662.62, and it closed at $655.38, above the previous close and near the bottom of its own bar.
On-balance volume added all 134,802,700 shares, because the close was higher. Chaikin’s multiplier was −0.67, so the accumulation line subtracted about 90.08 million. One line recorded the day as buying and the other as selling, and both did exactly what their formulas say.
Neither is wrong. The question is whether a higher close or a weak finish inside the bar is the thing you wanted counted. In 2026 alone, 21 sessions through 2 October closed up but in the lower half of their range.
Which one to use: Accumulation Distribution or OBV?
Use on-balance volume when you want the simpler tally. Its behavior is easy to predict and easy to explain, and being able to say exactly why the line moved is worth more than a small gain in precision.
Use the accumulation line when closing position is information you want counted. On instruments with long wicks, where a bar’s close within its range genuinely says something, the extra weighting is real rather than decorative.
Use on-balance volume where gaps are common, because the other one cannot see the jump from the previous close.
And run one, not both. They agree most days, so having both looks like confirmation while being one measurement with two weightings, and on the days they disagree neither says which is right.
Thirty-two years of both lines on SPY
They ended up far apart. Starting both at zero at the end of 1993, the accumulation line finished near its high and on-balance volume finished below a peak set almost five years earlier. Part of the reason is plain arithmetic: SPY closed higher on more days than lower, but closed in the upper half of its bar more often still, so the weighting that looks inside the bar kept adding.
On-balance volume’s 2021 peak became a divergence that lasted years. SPY fell through 2022 and then climbed far above its late-2021 price, while on-balance volume never got back to its peak. Read as a warning, that line said the same thing for nearly five years, through both the fall and the rise.
Calendar years tell the same story. Over the 32 full years, the accumulation line rose in more years than price did, while on-balance volume matched the direction of price more often.
The original data
The sample. Yahoo Finance daily SPY bars, highs, lows, closes and volume as reported, from 29 January 1993 to 2 October 2026, measured from 3 January 1994: 8,243 sessions. Opens are not used. Both lines start at zero on the last session of 1993. A bar with no range adds nothing to the accumulation line.
Day by day. The two lines moved the same way on 6,598 sessions (80.04%) and opposite ways on 1,560 (18.93%); on 85 one of them did not move. SPY closed up but in the lower half of its range on 670 sessions and down but in the upper half on 890. Overall it closed up on 4,431 sessions and down on 3,762, and closed in the upper half of the bar on 4,664 and the lower half on 3,544.
Over 20 sessions. Comparing each line’s change over the previous 20 sessions, they had the same sign on 5,876 of 8,223 sessions (71.46%). Cut into 412 separate 20-session blocks so no stretch is counted twice, they agreed on 290 (70.39%). In those blocks, on-balance volume moved the same way as price in 335 (81.31%) and the accumulation line in 315 (76.46%).
Where they ended. On 2 October 2026 the accumulation line stood at +45.51 billion shares, close to its high of 45.70 billion on 4 August 2026, and on-balance volume at +10.37 billion, below its 11.10 billion peak of 24 November 2021. SPY closed at $469.44 that day and at $769.64 on 2 October 2026, 63.95% higher. Total volume over the period was 702.1 billion shares.
By calendar year, 1994 to 2025. SPY rose in 23 of the 32 years, on-balance volume in 22 and the accumulation line in 29. On-balance volume moved the same way as price in 29 years, the accumulation line in 24. Every session is in the SPY on-balance volume and accumulation file, and every year in the yearly file.
When it fails
The failure is running both and treating agreement as confirmation, and the agreement is arithmetic rather than evidence. Two cumulative totals built from the same price and the same volume moved the same way on four SPY days in five. Seeing both rise reads as two independent measures concurring, when it is one measurement computed twice with different weights, and the confidence that produces is not supported by anything.
The second failure is quoting the level. It depends on the start date.
A third is feeding either one an incomplete volume record. The total can only be as complete as its input.
A fourth is finding divergence by looking. On-balance volume stayed below its 2021 peak for nearly five years while SPY ended 63.95% above that day’s close.
A fifth is expecting the accumulation line to see gaps. It looks inside the bar only.
And a sixth is treating on-balance volume as buying pressure. It reads the sign of the close.
Related
Accumulation distribution covers the weighted version. On balance volume covers the signed tally. And volume analysis sets the ceiling on what either can tell you.
Run one of these, not both, and decide in advance what you want it to count. If a bar that gapped up and closed weak should count as selling, use the accumulation line; if any higher close should count as buying, use on-balance volume. Then write down what a divergence has to look like before you look for one.
— Michael Whitman
This page is educational, not financial advice. Test every idea on your own charts before risking money.