On-Balance Volume
A running total of volume, signed by the direction of each price change. Aims to show whether volume is accumulating into rallies or selling into declines.
On-Balance Volume on GME, daily candles. Data via Financial Modeling Prep, cached server-side.
Quick reference
On-Balance Volume is the simplest volume-based indicator in technical analysis. It takes volume - which is just a number per bar with no inherent direction - and signs it with the direction of the price change. If price closed up, today's volume gets added to the running total. If price closed down, today's volume gets subtracted. If price closed unchanged, the running total stays.
That is the entire calculation. The simplicity is the point: OBV is not trying to measure how strong the buying was or how much money flowed. It is asking a binary question - did volume happen on up-days or down-days? - and accumulating the answer.
What OBV actually measures
OBV measures the net directional flow of volume over time. It is built on Joseph Granville's premise that volume precedes price. The idea: smart money accumulates ahead of price moves, leaving fingerprints in the volume data before the move shows up on the chart.
When volume is overwhelmingly happening on up-days, OBV rises persistently. When volume is overwhelmingly happening on down-days, OBV falls persistently. The slope of OBV reveals whether buying or selling has been dominant, integrated over the period you are looking at.
The cumulative nature is important. OBV is not "today's volume" or "this week's average volume." It is the running sum since OBV calculation began on the data series. Comparing OBV's absolute value is meaningless (it depends on when the data started); what matters is its slope and its divergence from price.
The formula
If today's close > yesterday's close: OBV = OBV(prior) + today's volume
If today's close < yesterday's close: OBV = OBV(prior) - today's volume
If today's close = yesterday's close: OBV = OBV(prior)
That is the entire formula. There is no smoothing, no period setting, no parameter to tune.
The result is a single line that walks up when price closes up on volume and walks down when price closes down on volume. The line's absolute value is arbitrary (it depends on the starting point of the data); the shape is what reveals the pattern.
A worked example
Suppose OBV currently reads at an arbitrary 1,000,000. The following five days print:
Day 1: Close higher than prior. Volume = 50,000. OBV = 1,050,000
Day 2: Close higher than prior. Volume = 60,000. OBV = 1,110,000
Day 3: Close lower than prior. Volume = 30,000. OBV = 1,080,000
Day 4: Close higher than prior. Volume = 40,000. OBV = 1,120,000
Day 5: Close higher than prior. Volume = 70,000. OBV = 1,190,000
OBV rose from 1,000,000 to 1,190,000 over five days. Volume on up-days totaled 220,000; volume on the single down-day totaled 30,000. The 7-to-1 ratio of up-volume to down-volume is visible in the steepening of the line.
Now imagine price barely moved during these five days - it ended the week roughly where it started. OBV would have risen 190,000 anyway, because more volume traded on up-closes than down-closes. That is OBV catching a pattern (heavier buying than selling on a flat price chart) that the price line itself does not reveal. This is what "stealth accumulation" looks like on OBV.
How traders actually use OBV
Three uses generate consistent edge. The fourth (treating OBV as a standalone trend indicator) does not.
1. Divergence
OBV's highest-value signal. Two flavors:
- Bullish divergence: price makes a lower low, OBV makes a higher low. The new low in price happened on less down-volume than the prior one - sellers are exhausted. Confluent with structure (a break of the prior swing high), this is a high-conviction long setup.
- Bearish divergence: price makes a higher high, OBV makes a lower high. The new high happened on less up-volume - buyers losing conviction. Confluent with structure breaking down, a high-conviction short setup.
Daily OBV divergence is more reliable than intraday. The longer the timeframe, the more meaningful the volume pattern.
2. Trend confirmation
The simple version. Price is in an uptrend; OBV is also trending up. Volume is supporting the move - the trend is healthy. Price is in an uptrend; OBV is flat or trending down. The trend is being driven by something other than broad participation - usually short-covering or low-volume drift, neither of which is durable.
OBV trending in agreement with price is a green light. OBV trending against price is a yellow light to tighten stops or reduce size.
3. Stealth accumulation / distribution
The most interesting OBV pattern. Price is sideways or even slightly down, but OBV is rising. Volume is happening on up-days even though price is not making progress. This suggests informed buyers are absorbing supply at current prices ahead of an eventual breakout.
The reverse pattern: price is holding sideways or slightly up, but OBV is falling. Sellers are unloading into strength, often ahead of a breakdown.
These setups have lead times of weeks. They do not produce immediate signals. They are watchlist criteria - "this stock looks like it is being accumulated; watch for the breakout when it comes."
The trap most retail traders fall into
The most common OBV mistake is treating absolute OBV values as meaningful.
OBV at "5 million" tells you nothing on its own. The number depends entirely on when the data series began. A chart showing OBV at 100 million on Stock A and OBV at 50 thousand on Stock B does not mean Stock A has "more buying" than Stock B - it means Stock A has more data history.
What matters with OBV is its slope (rising? falling? flat?) and its relationship to price (confirming? diverging?). The absolute number is just a coordinate on the line.
The second trap: trading OBV crosses through zero or other arbitrary levels. There are no meaningful threshold levels in OBV. The indicator has no natural overbought or oversold. Trying to "trade OBV at the 0 line" or "buy when OBV crosses above its 20-day average" is misapplying it.
The third trap: using OBV on instruments with poor volume data. Volume on small-cap stocks, illiquid futures contracts, or thin altcoins can be wildly inaccurate or manipulated. OBV computed on bad volume data is garbage. Stick to liquid instruments with reliable exchange-reported volume.
OBV vs other volume indicators
vs Accumulation/Distribution Line. A/D weights each bar's volume by the close's position within the bar's range. A bar that closes near its high gets more positive weight; a bar that closes near its low gets more negative weight. OBV treats every up-bar equally, regardless of whether it closed at the high or barely above the open. A/D is finer-grained; OBV is simpler. A/D handles "indecision bars" (close in the middle) more accurately.
vs Money Flow Index (MFI). MFI is bounded 0-100 like RSI and weights volume by price changes. MFI handles "how big was the move" while OBV ignores it. MFI produces overbought/oversold signals; OBV does not.
vs Volume bars themselves. Volume bars show per-period volume. OBV integrates direction into volume to produce a cumulative line. Volume bars are useful for spotting individual high-volume events; OBV is useful for spotting cumulative patterns.
vs Chaikin Money Flow. CMF combines OBV-style cumulative volume with bar-position weighting (similar to A/D) over a fixed lookback. CMF is bounded and easier to read for systematic strategies. OBV is unbounded and useful for divergence reads on longer horizons.
Common questions
Why does OBV look different on two platforms? Two reasons: starting point of the calculation (different platforms can begin OBV at different dates depending on their data archive) and how they handle the "unchanged close" rule. Most platforms correctly leave OBV unchanged on flat closes; some incorrectly add a small amount.
Should I change any settings on OBV? There are no settings to change. OBV is parameter-free - no period to set, no smoothing constant to tune. You either plot it or you do not.
Does OBV work on intraday charts? It can, but with limitations. Intraday volume is more erratic than daily volume (gaps, opens, market-on-close orders). OBV computed bar-by-bar on a 1-minute chart picks up a lot of noise. The most reliable OBV reads are on daily and weekly charts.
Can OBV predict reversals? It can identify them earlier than price alone, via divergence. It cannot reliably predict them in isolation. Combine with price structure and at least one other indicator.
What is the difference between OBV and money flow? "Money flow" can mean either the simple typical-price-times-volume calculation (which underlies A/D and MFI) or the more general concept of cumulative volume-direction. OBV is one specific implementation of cumulative direction; "money flow" indicators are a family of related but distinct measures.
Is OBV useful in crypto? Yes, but only on instruments with reliable exchange-reported volume. Bitcoin and Ethereum on major exchanges have usable OBV. Small-cap altcoins with self-reported or wash-traded volume have OBV readings that are essentially garbage.
When to use OBV and when not to
Use OBV when:
- You are looking for volume divergence between OBV and price on a daily or weekly chart
- You are scanning for stealth accumulation / distribution patterns
- You want a simple, parameter-free volume confirmation tool
Skip OBV when:
- You are trading instruments with unreliable volume data (microcaps, low-volume futures contracts, sketchy crypto exchanges)
- You need a bounded indicator with overbought / oversold levels - use MFI instead
- You want bar-position-weighted volume - use A/D Line instead
