A stock prints flat for three weeks. Closes hug the 20-day moving average. Daily volume bars look ordinary. Nothing on the chart screams. Then one Tuesday it gaps 6% on a news catalyst and a trader watching the price-only chart is left wondering where that came from.
The On-Balance Volume line on that same chart had been sloping up for the entire three weeks. Same data, different bookkeeping. That gap was not a surprise to anyone reading OBV.
This guide walks through what the OBV indicator measures, the calculation behind it, the divergence reading that is the only reason most traders pay attention to it, and the misreads I see most often. It does not pretend OBV is a complete system. It is one tool. I want it to be a tool you use accurately.
What the OBV indicator actually measures
OBV stands for On-Balance Volume. Joe Granville published it in his 1963 book Granville’s New Key to Stock Market Profits, and the math has not changed since.
The rule is plain. If today’s close is higher than yesterday’s close, add today’s volume to the running total. If today’s close is lower, subtract today’s volume. If the close is unchanged, the total holds where it was. That is the whole indicator.
An example with round numbers. Day 1 closes at 50.00. Day 2 closes at 50.50 on 2,100,000 shares, so OBV is now +2,100,000. Day 3 closes at 50.20 on 1,800,000 shares, so OBV drops to +300,000. Day 4 closes at 50.20 on 900,000 shares, the close did not change, so OBV stays at +300,000. Day 5 closes at 51.10 on 3,200,000 shares, so OBV jumps to +3,500,000.
The absolute number means nothing. A stock that has traded for ten years can have an OBV value in the billions; a stock that listed last quarter will have a value in the tens of thousands. You never compare the level of OBV between two tickers. You read its slope and the relationship between its peaks and troughs and the price chart’s peaks and troughs.
The idea behind the bookkeeping
Granville’s working theory was that volume precedes price. If accumulation happens quietly over weeks, the up-days will print fatter volume than the down-days even when the price chart looks balanced. OBV catches that asymmetry by adding the up-volume and subtracting the down-volume into a single running line.
This sits in the lineage of Richard Wyckoff’s tape reading, and the same instinct shows up in modern volume tools: the assumption is that informed buyers leave footprints in the volume long before the price chart confirms them.
OBV is not magic and does not see anything you could not see by counting up-volume and down-volume by hand. It is a bookkeeping trick that surfaces an imbalance most traders would miss because their eye gets caught by the height of individual volume bars rather than the cumulative bias of the up-day versus down-day stack.
Reading OBV divergences
Divergences are the reason most traders watch OBV at all. There are two shapes worth knowing.
A bullish divergence prints when price makes a lower low and OBV makes a higher low. Price says the sellers won the round. OBV says the down-day volume was thinner than the previous down move and the up-days in between carried more weight. The read is that the supply that pushed price down was light.
A bearish divergence is the mirror. Price makes a higher high. OBV makes a lower high. The breakout looks valid on price but the volume signature of the second push was weaker than the first.
I look for these on the daily chart, anchored on swing pivots a trader could mark with a ruler. A worked case: a stock prints a swing low at 182.40, rallies to 198, pulls back, then prints a marginal new low at 181.95. OBV at the first low was 12,400,000; at the second low it is 18,700,000. That is a bullish divergence by definition.
What this is not. A divergence is not a trade signal on its own. It tells you the volume profile under the second move is healthier than the price chart suggests, and that is all. A trader using divergence as an alert still needs a structural trigger, usually a higher low confirming on price or a break of the most recent lower-high. Skipping the trigger and acting on the divergence alone is the most common way I see this indicator fail people.
OBV trend lines and the early break
The cumulative line is a price chart in its own right and you can draw trend lines on it. A rising OBV that holds an ascending support line for six weeks and then breaks below it often precedes a similar break in price by several sessions.
This is not a guarantee. The relationship is statistical, not mechanical. I have seen plenty of OBV trend breaks that resolved sideways while the price kept making fresh highs. What the OBV trend break does is shift the burden of proof. Before the break, the assumption is continuation. After, the assumption is at-risk until proven otherwise.
The cleanest pairing is with concrete price structure. Volume confirmation on breakout candles is the short-horizon read; OBV is the longer-horizon read. They answer different questions, and a trader who keeps them separate avoids the trap of forcing one to do the other’s job.
Where the OBV indicator gets contaminated
The calculation has three holes that cost traders money.
The first is the binary treatment of price. A close one cent above prior treats all of that day’s volume as positive. A close one cent below treats all of it as negative. A stock that prints an inside day with a strong intraday rally but closes a hair red contributes a negative volume number that is the same magnitude as if it had collapsed from the open.
The second is earnings. A 12% earnings gap on triple the average volume can swing OBV by the equivalent of a month of normal trading, and that distortion sits inside the indicator for the full length of the chart window. If you read the line without marking the catalyst dates, you will mistake the earnings spike for organic accumulation.
The third is thinly traded stocks. A small-cap with average daily volume under 200,000 shares produces an OBV line that jumps on individual block trades. The signal-to-noise on those tickers is poor and divergence reading is unreliable. I will not run OBV on anything below 500,000 average daily volume.
OBV is not the same as Chaikin Money Flow
This trips up newer traders, so worth being specific. OBV uses the close-versus-prior-close test and treats the day’s full volume as either positive or negative. Chaikin Money Flow uses the close’s position within the day’s high-low range, weights the volume by that position, and sums over a fixed lookback (usually 20 days). They will often disagree because they measure different things. CMF sees a strong intraday close as bullish even on a red day; OBV does not.
OBV is also not the same as the Negative Volume Index or its positive sister. NVI only updates on days when volume contracts. OBV updates every day. NVI is built around the theory that informed money trades on quiet days. OBV makes no such assumption.
If you want a signal-line variant of cumulative volume, the Klinger Oscillator takes the cumulative idea and converts it into a fast-and-slow EMA difference. That makes it easier to spot turns mechanically, at the cost of the directness OBV has.
How I use OBV in a swing-trading workflow
I run OBV second, not first. The setup starts on price. I want a base, a clear pivot, a defined invalidation level. OBV joins the read after that to answer one question: is the volume under this base supportive of an upside resolution or not.
The mechanical version of my filter. The 50-day OBV slope has to be positive, measured by least-squares regression on the last 50 cumulative values. The most recent OBV swing low has to be above the prior OBV swing low. If both are true, the setup stays on the list. If either fails, I drop it even if the price chart looks textbook.
On a 4-hour chart, OBV divergences resolve in roughly 6 to 12 bars from the second low. On a daily chart the same setup takes 2 to 4 sessions. I time-stop divergence-based watches at twice that horizon; if price has not confirmed by then, the divergence reading has decayed and the chart deserves a fresh look.
This is the same disciplined approach William O’Neil built into CAN SLIM around volume confirming breakouts. The exact metric is different. The instinct, that the volume chart has to vote with the price chart before you commit capital, is the same.
What OBV is actually telling you
OBV is a measurement, not a forecast. It tells you whether the up-day volume or the down-day volume has been heavier across the chart window you are looking at. That answer is useful when you want to know whether a flat price chart is hiding accumulation or whether an apparently strong breakout has thin volume underneath it. It is not useful when you treat it as a buy-or-sell trigger in isolation.
Use the line, draw on it, watch its slope, mark its divergences against price, and respect what it cannot do. Read it together with the price chart and at least one other volume tool. That is the whole job.
Learn the pattern. Ride the trend. Keep the gains.
Educational content only. Not investment advice. Trading involves risk. You are responsible for your decisions.