Positive Volume Index – Read Crowd Behavior on Loud Days

A stock gaps up on earnings, volume triples, and every scanner on Twitter lights up. That is a Positive Volume Index day. The PVI only updates when today’s volume exceeds yesterday’s. It captures what price does when participation surges, and the assumption behind it is blunt: high-volume sessions reflect crowd-driven activity, not the quiet positioning of informed players.

The Positive Volume Index is the mirror of the Negative Volume Index, which tracks price action on low-volume days. Where the NVI filters for quiet accumulation, the PVI records the noise. That noise is not useless. It tells you whether the crowd is pushing in the same direction as the underlying trend or fighting against it. The divergence between the two is where the real signal lives.

I track the PVI on SPY and a handful of large-cap names because high-volume days on liquid instruments carry actual information. On April 14, 2026, SPY volume jumped to 63.5 million shares from 54.2 million the prior session. Price climbed from 686.10 to 694.46. The PVI updated. The crowd was buying into a rising market. That kind of alignment is easy to read. The harder question is what it means when the PVI starts disagreeing with the NVI.

How the Positive Volume Index Calculation Works

The PVI starts at a base of 1,000 (some platforms use 100). It changes only on days when volume rises compared to the prior session. On days when volume drops or stays flat, the PVI carries forward unchanged.

On an up-volume day:

On a down-volume day (volume equal to or less than the prior session):

The percentage change in price gets applied to the running PVI total, but only when participation increases. No smoothing. No lookback parameter. The indicator is cumulative from day one.

A mistake I see constantly: traders assume the PVI tracks volume itself. It does not. It tracks price change, filtered by volume direction. A day where volume doubles but price barely moves will produce a near-zero PVI update. A day where volume ticks up by 1% but price rallies 3% will produce a large PVI update. The volume is the gate. Price is the signal.

Why the PVI Gets Less Respect Than It Deserves

Norman Fosback, who developed both the PVI and NVI in the 1970s, found the NVI more statistically reliable for identifying bull and bear markets. His research showed the NVI above its 255-day EMA predicted bull markets 96% of the time. The PVI’s equivalent test was weaker. That asymmetry led most traders to favor the NVI and treat the PVI as an afterthought.

The dismissal is premature. The PVI’s weakness as a standalone trend filter does not make it useless. It makes it a different tool. The NVI tells you what informed money does on quiet days. The PVI tells you what the crowd does on loud days. You need both to see whether market participants are aligned or conflicted.

Where the conventional wisdom fails: treating the PVI as “the dumb money indicator” and ignoring it. High-volume days include institutional block trades, algorithmic rebalancing, and options expiration flows. Calling all high-volume activity uninformed is sloppy. The PVI captures a mix of participants. Its value is not in labeling them smart or dumb but in tracking whether high-participation days produce directional agreement with the trend.

Reading PVI Direction on Its Own

A rising PVI means price tends to advance on days when volume increases. That is a bullish crowd: more participants show up, and they buy. A falling PVI means price tends to decline when volume rises. The crowd is selling into volume spikes.

Take AAPL between April 6 and April 16, 2026. On April 7, volume surged from 29.3 million to 62.1 million. Price dropped from 258.86 to 253.50. The PVI updated with a negative move. On April 14, volume rose from 36.2 million to 48.4 million. Price slipped from 259.20 to 258.83. Another negative PVI update. But on April 15, volume rose again to 49.9 million, and price jumped from 258.83 to 266.43. That session produced a strong positive PVI update. Over this stretch, the PVI’s trajectory was choppy: two negative updates, one large positive. No clean trend signal. That is typical for the PVI in range-bound or transitional conditions.

What traders get wrong here: expecting the PVI to produce smooth trends like the NVI. It usually does not. High-volume days are more volatile. News events, earnings, macro data releases all cluster on high-volume sessions. The PVI line is inherently noisier, and trying to trade its short-term slope will generate whipsaws.

PVI and NVI Together: Confirmation and Divergence

The real utility of the Positive Volume Index shows up when you compare it to the NVI. Four states are possible, and each tells a different story.

Both rising: the trend is healthy. Price advances on quiet days (NVI up) and on loud days (PVI up). Buyers are present regardless of participation level. This is the easiest environment for swing trades in the trend direction.

NVI rising, PVI falling: this is the divergence that gets interesting. Quiet days show accumulation while loud days show selling pressure. I interpret this as informed players building positions while the crowd panics on headlines. It often appears during pullbacks within a larger uptrend. If I see this pattern and price is near a support level identified through support and resistance analysis, I am more willing to take the long entry.

PVI rising, NVI falling: the opposite scenario. The crowd is buying on volume spikes, but quiet-day action is weak or declining. This pattern can signal distribution. Informed players may be selling into the retail enthusiasm. I treat this as a warning to tighten stops or avoid new long entries, especially if price is extended above a anchored VWAP level.

Both falling: broad weakness. Neither high-volume nor low-volume days are supporting price. This is not the environment for aggressive positioning in either direction.

A critical nuance most guides miss: these four states are not always clean. The NVI and PVI update on different days, so their slopes can be hard to compare on a day-by-day basis. I look at the 20-day direction of each line, not individual updates. That smooths the comparison enough to be actionable without burying the signal.

Applying PVI Divergence to Swing Entries

Here is how I use PVI/NVI divergence in practice. This is not a mechanical system. It is a filter that shapes trade selection.

Step one: check whether the NVI is above its 255-day EMA. If yes, the broad trend filter is bullish. If no, I reduce exposure regardless of what the PVI says.

Step two: compare the 20-day slopes of the PVI and NVI. If both slope upward, I take standard swing entries on pullbacks. If the NVI is rising and the PVI is falling or flat, I still take longs, but I look for deeper pullbacks and want to see the PVI stabilize before entry. That falling PVI tells me the crowd is not yet committed, so I want a better price as compensation.

Step three: if the PVI is rising while the NVI is flat or falling, I am cautious. The crowd is enthusiastic, but quiet-day behavior is not confirming. I either skip new entries or take smaller positions. This state preceded the late March 2026 selloff in several names. SPY’s high-volume days were still producing positive moves, but quiet sessions were not following through.

I combine this with volume oscillator readings to gauge whether the volume trend itself is expanding or contracting. The PVI tells me what happens on high-volume days. The volume oscillator tells me whether those high-volume days are becoming more or less frequent. Both together paint a clearer participation picture than either alone.

Where the Positive Volume Index Breaks Down

The PVI has specific failure modes, and knowing them prevents bad reads.

Options expiration weeks inflate volume without adding directional information. Quad witching Fridays can see volume double on mechanical rebalancing. The PVI will update on that day, but the price move may reflect gamma hedging flows rather than any directional conviction. I discount PVI updates during the last two trading days of monthly options expiration.

Earnings season produces a similar distortion. A stock reporting after the close will see the next session’s volume spike on the reaction. The PVI captures that move, but earnings gaps are driven by fundamental re-pricing, not the volume-participation dynamic the PVI was designed to measure. The PVI reading on an earnings gap day is real data but carries different meaning than a normal high-volume session.

Low-liquidity instruments produce unreliable PVI readings for the same reason they produce unreliable NVI readings. If average daily volume is 50,000 shares, a day at 60,000 triggers a PVI update. That 20% volume increase might be a single institutional order, not a meaningful shift in participation. I only track the PVI on names averaging at least 1 million shares daily.

A subtler problem: the PVI does not weight the magnitude of volume change. A day where volume goes from 50 million to 51 million gets the same treatment as a day where volume goes from 50 million to 150 million. The PVI gate is binary: volume up or not. This means a trivial volume increase can trigger an update that carries the same weight as a genuine volume surge. Be aware of this when a single PVI update seems to change the indicator’s trajectory.

PVI on SPY: A Recent Walk-Through

Walk through SPY from April 6 to April 16, 2026, and track which days the PVI would update.

April 6: volume 39.1 million. If the prior session was higher (and it was, at 68.4 million on April 2), volume declined. PVI stays flat. NVI updates. Close: 658.93.

April 7: volume 70.0 million, up from 39.1 million. PVI updates. Price moved from 658.93 to 659.22. Small positive PVI update.

April 8: volume 93.6 million, up from 70.0 million. PVI updates. Price from 659.22 to 676.01. Large positive PVI update.

April 9: volume 57.1 million, down from 93.6 million. PVI stays flat. NVI updates. Close: 679.91.

April 10: volume 42.3 million, down from 57.1 million. PVI stays flat. NVI updates. Close: 679.46.

April 13: volume 54.2 million, up from 42.3 million. PVI updates. Price from 679.46 to 686.10. Positive update.

April 14: volume 63.5 million, up from 54.2 million. PVI updates. Price from 686.10 to 694.46. Positive update.

April 15: volume 58.2 million, down from 63.5 million. PVI stays flat. NVI updates.

April 16: volume 49.9 million, down from 58.2 million. PVI stays flat. NVI updates.

The PVI updated on four out of nine sessions. Every update was positive. That is a clean signal: when the crowd showed up, they were buying. Combined with a rising NVI over the same period (the NVI article walked through those same dates), both lines were pointing up. Full confirmation of the uptrend from early April.

PVI Settings and Platform Considerations

Default settings across most platforms: starting value of 1,000, with an optional moving average overlay. Fosback’s original research paired the PVI with a 255-day EMA, same as the NVI. Some traders use a shorter EMA for the PVI since its signal is noisier. I have tested 100-period and 255-period EMAs on the PVI and found the 255 more useful for filtering trend direction. The shorter period just picks up more noise.

TradingView, MetaTrader, and most institutional platforms include the PVI as a built-in indicator. If yours does not, the calculation is simple enough to script. You need daily close and daily volume. Compare volume to prior day. If higher, apply the price percentage change to the running total. If not, carry forward.

One thing to watch: some platforms calculate both PVI and NVI on the same chart with the same starting value. Others use separate panels. There is no analytical difference. I prefer separate panels because it makes slope comparison easier visually, but that is personal preference.

Crowd Behavior Is Data, Not Noise

The Positive Volume Index tracks what happens when participation increases. That information is worth having even though it is noisier than the NVI’s quiet-day signal. High-volume days are where trends get tested. If price advances every time volume surges, the trend has broad support. If price retreats on every volume spike, the crowd is selling into rallies, and the trend is fragile.

Use the PVI alongside the NVI to read the full participation picture. Track their 20-day slopes for confirmation or divergence. Let the NVI handle the trend filter (above or below its 255-day EMA). Let the PVI tell you whether the crowd agrees. When they align, trade with confidence. When they diverge, trade with caution or step aside.

The crowd is not always wrong. Sometimes they are early. The PVI helps you figure out which one it is.

Educational content only. Not investment advice. Trading involves risk. You are responsible for your decisions.