OBV Indicator: How Volume Confirms the Trend

Price printed a new high. Clean, pretty, punching straight through a round number. Except there was no money behind that high, and twenty minutes later the coin was back where it started, taking out the stops of everyone who bought the breakout.

On a bare chart, moves like that look identical. Candle up, the volume histogram shows some bar, now figure it out yourself. The OBV indicator gathers those bars into a single line and answers one question: is real flow building underneath the move, or is price being dragged along on leftover momentum.

Let's go through the formula, the limitations, two setups that actually work, and the link to order flow — without which the OBV indicator turns into a pretty line under the chart.

What OBV is

On Balance Volume is a cumulative signed volume counter. A candle closes above the previous one, its entire volume gets added to the running total. Closes below, the entire volume gets subtracted. The resulting line doesn't show how much was traded — it shows which side the balance leaned toward across the whole calculation history.

Joseph Granville came up with the indicator in the early sixties. His claim: volume leads price, and a large participant builds a position before the move becomes obvious. The claim is debatable and holds up poorly in its pure form. But the idea of splitting volume by direction of close outlived its author and became the basis for a whole family of tools, from the accumulation/distribution line to cumulative delta.

The main thing to get straight right away: OBV doesn't show who was buying. It shows where the candle closed and assigns its entire volume to that side. That difference matters.

The On Balance Volume formula

The calculation fits in three lines.

If Close > Close(prev):   OBV = OBV(prev) + Volume

If Close < Close(prev):   OBV = OBV(prev) - Volume

If Close = Close(prev):   OBV = OBV(prev)

No smoothing, no periods, no settings. Hence the first quirk: the indicator has no parameters you can tune to an instrument. It's identical on BTC and on some coin ranked three hundredth.

Let's run it on real numbers. BTC/USDT, five-minute chart, starting value set to zero.

Look at the 10:15 bar. Price rose by two dollars at a price of 67 thousand — 0.003%. Technically that's an up bar, so all 95 BTC went into the plus column. The indicator doesn't distinguish a two-dollar move from a four-hundred-dollar one.

That binary treatment is the core weakness of the design. Granville was working with daily bars on stocks, back when intraday data simply wasn't available, and a crude plus/minus split gave acceptable accuracy. Today we have per-trade data with an aggressor flag, and next to that OBV looks like an abacus sitting beside a calculator.

Second property of the formula: it accumulates without ever resetting. The value depends on which bar your terminal started counting from, so the absolute number on the scale carries no meaning at all. Only the slope of the line and its structure are readable.

What it shows

OBV is plotted in a separate pane below the chart as a single line. There are three things to read in it.

Slope. A rising line says volume is accumulating on up closes. A falling one, on down closes. A flat line while price is moving means buys and sells are cancelling each other out, and the rally is running on empty.

Extremes of the line. OBV draws its own highs and lows, and they don't have to line up with price highs and lows. Comparing the two sets of extremes is where the analysis comes from.

Sharpness of change. A vertical jump in the line on a single bar means burst volume: a liquidation cascade, or a large participant stepping in with one block.

What OBV doesn't show is levels. No overbought zone, no 70 or 100 marker, no neutral midpoint. Let's compare it with its neighbours to place the tool properly.

The logic goes like this. OBV gives a crude but long view of balance: is anything piling up over the last few hours. Delta in footprint clusters gives a precise but short one: who is pressing right now at a specific level. The general mechanics of turnover are covered in the piece on crypto trading volume, and volume distribution across price zones and POC in the Volume Profile article.

If the basics of volume are still shaky, watch the free lesson from our full trading course on YouTube: it walks through, on a live screen, how the order book differs from the tape and how volume is distributed inside a candle.

How to read OBV

In practice the work comes down to comparing two pictures: what price is doing and what the line is doing. They agree — the trend is backed by flow. They diverge — you've got a question.

OBV rising with price = a healthy trend

The baseline state, where nothing needs doing. Price makes new highs, OBV makes its own. Every leg up comes with an increase in cumulative volume, pullbacks run on less volume than the impulses themselves.

The mechanics are simple. Buyers lift the offer with market orders, candles close up, their volume goes into the plus column. Sellers work the pullbacks half-heartedly, the subtraction is weak, and the line steps its way higher. A trend like this gets traded long from pullbacks with no extra questions about a reversal. In a downtrend the picture is mirrored.

And here are three states where the trouble starts.

Price rising, OBV flat. A rally without inflow doesn't last. Usually the move is being carried on thin liquidity: the order book is empty, there are no large orders on either side, and price drifts up simply because nobody is selling it. The first serious market sell puts it all back.

Price ranging, OBV rising steadily. Textbook accumulation in the Granville sense. Someone is building a position by taking the offer inside the range without pushing price up. A rare signal, and on the daily it deserves attention.

Price ranging, OBV falling steadily. Distribution: a large participant is unloading into the range, selling to everyone who thinks they're watching accumulation.

OBV divergence with price

Divergence is a mismatch between the direction of price and the direction of the indicator, where a new price extreme isn't confirmed by cumulative volume. It's the most valuable thing the tool gives you. And the most common way to blow up an account, if you trade it head-on.

Bearish divergence. Price makes a higher high, OBV makes a lower one. Buyers got further, but bought less.

Bullish divergence. Price makes a lower low, OBV makes a higher one. Sellers pushed price lower, but it took less volume to do it.

A numeric example. ETH/USDT, hourly. First high: price 3,482, OBV 1.24M ETH cumulative. The second one eleven hours later: price 3,516, OBV 1.11M. Price went 0.98% higher, cumulative balance dropped 10.5%.

Then comes the part that separates a trader from someone reading tutorials. Divergence is not a signal. It can drag on for days, refreshing four times over, while price walks calmly higher and takes out the stops of everyone who shorted the first mismatch.

What turns a divergence into an argument for a trade:

• the second high forms with fading aggression on the tape — the large buys are gone and flow has turned choppy
• the footprint clusters at the upper level show absorption: big volume went through and price didn't budge
• there's a density level sitting above the second high that holds for more than a few minutes
• the divergence is visible on the hourly chart, not the five-minute

On timeframes, bluntly. OBV divergence on 5m in crypto almost never works. The minimum horizon that makes sense is 15 minutes; a proper one is hourly and above. Every type of divergence, hidden ones included, is covered in a separate article "Divergence: Types of Divergence and How to Trade Them".

Strategies

Two setups where the OBV indicator is genuinely useful rather than an illusion of understanding, plus the mandatory link to order flow.

Breakout confirmation

The main working function. Price approaches a level, and the question is the usual one: real breakout or a fakeout.

The rule in one line. A price breakout of the level has to come with OBV breaking its own high, and the line has to go above the value it held at the previous test of that same level.

Here's the logic. If the level is being tested for a third time and cumulative volume is the same each time, the limit order at the level hasn't gone anywhere — it simply hasn't been eaten. If by the third test OBV is noticeably higher than at the first, volume was being accumulated in that zone and there's nothing left to defend the level with.

Checklist before entering:

• Mark the level and the OBV values at each touch
• Wait for price to move beyond the level with a bar close
• Check whether the line made a new local high
• Look at the order book: did the density level get filled, or pulled
• Check the tape for large market orders in the direction of the breakout

Three of the five points have nothing to do with the indicator. That's normal. OBV here is one vote among several, not the judge.

Case study. SOL/USDT, 15 minutes, level 178.40. Price approaches it for the third time in a day.

First touch at 04:15, OBV 812 thousand, rejection down 1.9%. Second at 11:30, OBV 940 thousand, rejection of 0.8% — noticeably weaker. Third at 16:45, OBV 1.19M, the line 46% above the first test.

I check the order book. At 178.42 there was an order for 34 thousand SOL; on the first two approaches it wouldn't let price through. By the third it had thinned to 9 thousand, and the order lifetime timer showed eight minutes instead of over an hour. The tape took off: market buys in chunks of 800-1,500 SOL, the rest of the density level chewed through in forty seconds.

Entry at 178.55 after the bar closed above the level, stop at 177.90 (0.65 dollars away). Account 10,000, risk 0.5% — 50 dollars — so position size = 50 / 0.65 = 76 SOL. Take at 181.20, where the nearest large density level sits above, risk-to-reward 1 to 4. It ran to 180.85 and I closed it manually. Plus 175 dollars.

The key point: the reason for the entry was the eaten density level and the acceleration on the tape. OBV only confirmed that the level had been prepared in advance. Without the order book I wouldn't have taken that trade.

OBV + levels

The second setup is less obvious. The levels are drawn not on price but on the indicator line itself: cumulative volume, like any other series, forms horizontal zones and sloping channels. The line runs into its own resistance and bounces, which means flow dried up on the approach to the same mark as last time.

How to use it without fooling yourself:

• Draw a trendline along OBV lows in an uptrend. While it holds, price pullbacks are bought. A break of that trendline is the first sign the structure is cracking, before the price low even breaks
• Mark a horizontal level where the line stalled twice. The third approach, with acceleration, usually comes with price leaving the range
• Overlay a 20-period moving average. Above it, look for longs; below it, shorts
• Compare accumulation between correlated coins. Bitcoin is flat, its OBV is flat, but the alt's OBV is going vertical — money is coming into that coin independently of the market

About the lead time Granville promised: in crypto on intraday timeframes it amounts to one or two bars. Useless for scalping, fine for reading context.

OBV, the order book and the tape

Here's the honest conversation about where OBV trading ends and order flow begins.

The indicator looks backwards at closed bars and assigns volume a sign by a crude rule. The order book shows limit orders sitting there right now, not yet filled. The tape shows real aggression with a side flag on every trade. A footprint cluster is the distribution of volume across prices inside a single candle: at which level the exchange happened and which side won.

The difference, in a concrete situation. A five-minute bar closes 12 points above the previous one on 380 BTC of volume, and OBV adds all 380 to the plus column. Yet the footprint for that bar might show 160 BTC of market buys against 220 of sells, delta minus 60, with price rising only because the seller pulled their limit orders from the upper levels. The indicator books the bar to the buyer, even though the buyer was the weaker side.

Which gives the rule I arrived at after several expensive lessons. OBV is good for background context and for divergences on higher timeframes. The entry decision is made on order flow.

Secret Terminal's tools cover all three layers of flow:

• Footprint clusters with delta. Volume distribution across prices inside the candle, with total volume and delta in the footer
• Order lifetime timer in the order book. A density level that sits there for a long time is real interest; one that appears and vanishes instantly is spoofing — a decoy order pulled before it can be filled
• Density map across all exchanges. Ready-made levels to anchor your entry and stop to
• Anomalous volume highlighting. Tells you immediately what caused that vertical jump in the line

Take the order book and the tape out of the chain and you're left guessing at a line.

Mistakes people make with OBV

Five common ones, each of which cost somebody money.

Trading divergence head-on. A mismatch looks convincing and creates a feeling of inside knowledge. In my experience, divergence without confirmation from the tape and the footprint clusters breaks down about a third of the time, and more often than that in a strong trend.

Watching OBV on the minute chart. On M1 and M5, half the bars close a single tick apart, so the sign of the volume is effectively decided by a coin flip. I don't open the indicator below 15 minutes.

Comparing absolute values. OBV depends on where the calculation started: two charts with different history depth will show different numbers on the same instrument.

Calculating it on somebody else's exchange. The same hour on BTC gives different volume on Binance, Bybit and OKX, and spot and perpetuals live their own lives. Calculate it where you trade.

Using it on listings and in tight ranges. This is where the indicator flat-out doesn't work. A cumulative line needs history, and a first-day coin has none: the move happens in the opening minutes, while OBV has managed to accumulate three or four bars. In a compressed range, the binary sign of the close turns the line into noise. What works there is spread width, tape speed and order book depth.

A breakdown of order flow on a live market is in the free lesson of the full crypto trading course on YouTube: how professionals read the market through the order book and footprint clusters, with entry examples.

FAQ

• What does the OBV indicator show, in plain terms?

  Cumulative volume signed by the direction of the candle's close. A candle closes above the previous one, its entire volume is added; below, subtracted. A rising line means volume is piling up on up closes. The absolute value means nothing — only the slope and structure of the line are readable.

• What OBV settings are best for crypto?

  There are no settings; the formula contains no periods. All you can change is the timeframe, plus adding a moving average over the line. A working setup for crypto: hourly chart for context, 15 minutes for divergences, a 20-period moving average as a regime filter.

• Which timeframe does OBV work best on?

  Hourly and above. On the five-minute and one-minute charts a significant share of bars close one or two ticks apart, so the sign of the volume is close to random. The line gets noisy and divergences appear and vanish every half hour.

• How is OBV different from a regular volume indicator?

  A volume histogram shows turnover for each bar separately and says nothing about direction. OBV sums those turnovers on a running basis, assigning a sign by the bar's close. The first tool answers "was there a lot of trading"; the second answers "which side did the balance lean toward".

• OBV or cumulative delta — which is more accurate?

  Delta is more accurate. It counts real market buys and sells with an aggressor flag on every trade, while OBV assigns a bar's entire volume to one side by the sign of the close. A bar can close up on negative delta, and the cumulative line will wrongly book it to the buyer.

• Can you trade on OBV alone?

  No. The indicator gives no entry points, shows no levels for a stop and doesn't distinguish the strength of a move. In combination it works as a context filter and a source of divergences, while the entry decision is made on the order book, the tape and the footprint clusters.

• Why does OBV show different things on different exchanges?

  Volume is calculated from a specific venue's data, and spot and perpetuals trade with different activity. The same hour on BTC produces mismatched turnover across exchanges, so the cumulative lines can diverge in direction. Look at the data from the venue where you're opening the position.

What to take away

The OBV indicator does one thing: it turns scattered volume bars into a continuous line of balance, from which you can see whether flow is building under a move. Everything else needs verification.

The way to use it fits in four steps. Hourly OBV gives you the background. The chart gives you the level. The order book gives you the order your stop hides behind. The tape gives you the moment.

And the main rule, the one that saves the most money. A divergence between OBV and price is a reason to take a closer look, not a reason to open a position against the move. The difference between those two statements is measured in blown accounts.

Watch volume where the flow is visible.

Secret Terminal puts the chart with volume indicators, the order book with a density map and the tape in one window. Footprint clusters with delta show who won inside the candle; the order lifetime timer separates a real density level from spoofing. Your divergence gets confirmed before the entry instead of after the stop.