Negative Volume Index: Tracking What Happens on the Quiet Days

Most volume indicators shout about the big days. The negative volume index does the opposite. It only moves on days when turnover is lower than the day before, and it simply ignores the busy, headline-grabbing sessions. That sounds odd at first. But once you see why it was built this way, it becomes a useful second opinion on whether a rally in the Nifty or in a single stock has real support underneath it.

We will go through where the idea came from, how the number is calculated, a full worked example you can redo in a spreadsheet, and the places where it tends to mislead. Everything below is for learning; the example figures are hypothetical and chosen to make the arithmetic easy.

What the negative volume index actually measures

According to StockCharts’ ChartSchool, the Negative Volume Index (NVI) is a cumulative indicator that uses changes in volume to judge when the “smart money” is active. Paul Dysart first developed it in the 1930s. His assumption was simple: informed money does its work on quiet days when volume falls, while the less-informed crowd piles in on days when volume rises.

So the NVI is a running total that only updates on lower-volume days. On a day when volume goes up, the line stays flat, no matter how big the price move. Over weeks and months, what you end up with is a record of how prices behaved when nobody was making a fuss.

There is a sister indicator, the Positive Volume Index (PVI), which does the reverse and updates only on days when volume rises. Wikipedia’s history of the indicator notes that Dysart called the PVI the “majority” curve and that, in a 1967 Barron’s article, he described the NVI as having “proved to be the most valuable of all the breadth indexes”. He even gave it a name, AMOMET, short for “A Measure Of Major Economic Trend”.

Dysart’s version versus Fosback’s version

Here is a detail many explainers skip. Dysart did not use index prices at all. StockCharts explains that in the original version he built a cumulative line by adding Net Advances, meaning advancing issues minus declining issues, on days when volume fell. It was a market breadth tool.

Norman Fosback, author of Stock Market Logic (which, per Wikipedia, reported his findings in 1976), changed that. He swapped Net Advances for the percentage price change of a market index. That is the version most charting platforms use today, including StockCharts’ own SharpCharts. Wikipedia lists three Fosback changes:

• He added up the daily percentage change in the index instead of advances minus declines.
• He suggested starting the cumulative count at a base level such as 100.
• He generated buy or sell readings based on whether the NVI sat above or below its one-year moving average.

StockCharts starts its line at 1,000 instead of 100 and uses a 255-day exponential moving average as the one-year signal line. The base number does not matter; it only sets the scale. What matters is the direction of the line and where it sits relative to its long average.

How to calculate the negative volume index

The calculation fits in four lines. Following the StockCharts method:

1. Start the cumulative NVI at 1,000.
2. If today’s volume is lower than yesterday’s, adjust the NVI by today’s percentage price change.
3. If today’s volume is equal to or higher than yesterday’s, leave the NVI unchanged.
4. Plot a 255-day EMA of the NVI as the signal line.

Wikipedia writes Fosback’s formula in multiplicative form. If C_t is today’s close and C_y is yesterday’s close, then on a lower-volume day:

NVI today = NVI yesterday + NVI yesterday × (C_t − C_y) / C_y

That is the same as multiplying yesterday’s NVI by (1 + today’s percentage change). On any other day, you add zero. The PVI uses the identical formula, but only on days when volume rises.

In Indian terms, for the index version you would take the Nifty 50 close and the total traded volume or turnover for the session. For a single stock, use that stock’s daily close and its daily traded quantity from the exchange bhavcopy or your charting software.

A worked example, step by step

Let us take a made-up index that starts at 20,000 and trade it through five sessions. Volumes are in million shares and are purely illustrative.

Walk through it slowly.

Day 1. The index jumps 1% on heavier volume, 360 million against 300 million. Because volume rose, the NVI stays at 1,000. The PVI picks up the move and goes to 1,010.

Day 2. The index slips 100 points, which is about 0.495% of 20,200. Volume drops to 310 million, so this day counts for the NVI: 1,000 × (1 − 0.00495) = 995.05.

Day 3. A clean 1% gain to 20,301 on even lighter volume of 280 million. The NVI moves to 995.05 × 1.01 = 1,005.00.

Day 4. A small dip of roughly 0.25% on a busy 400 million day. The NVI ignores it. The PVI falls to about 1,007.46.

Day 5. The index climbs about 0.75% to 20,402 while volume eases to 350 million. The NVI goes to 1,005.00 × 1.0075 ≈ 1,012.54.

Over the five days the index itself gained about 2.01%, from 20,000 to 20,402. The NVI rose about 1.25%, and the PVI about 0.75%. The interesting part is not the totals but the split: most of the gain on lower-volume days held up, while the busy days added little net progress. Dysart would have read that as a constructive sign, because prices were holding and rising when the crowd was not chasing them.

The NVI is a record of what the market did when nobody was shouting. If prices keep climbing on the quiet days, that rally has a different quality from one that only moves when volumes spike.

Reading the signal: the 255-day line

Fosback’s rule is the one you will see quoted everywhere. StockCharts summarises it this way: the odds favour a bull market when the NVI is above its 255-day EMA, and favour a bear market when it is below. The odds are not symmetrical, though. Fosback found a 96% chance of a bull market when the NVI was above its long average, but only a 53% chance of a bear market when it was below.

Two caveats matter here, and both come straight from the sources. Wikipedia notes those probabilities were derived from a 1941 to 1975 test period on US market data, and that modern tests might give different numbers. So treat 96% as a historical finding about a different market in a different era, not a promise about the Nifty in 2026.

The second caveat is the asymmetry itself. A 53% reading is barely better than a coin toss. In plain words, an NVI above its long average has historically been a decent confirmation of an uptrend, while an NVI below it says much less about a coming fall. Wikipedia also lists Fosback’s PVI figures for comparison: 79% for bull markets when PVI is above its one-year average and 67% for bear markets when below.

Whipsaws and lag

StockCharts is honest about the weaknesses. Even with such a long moving average, its Dow Industrials example shows plenty of whipsaws, with the NVI crossing its 255-day EMA several times during choppy phases. And because a 255-day EMA is slow by design, signals arrive late. That is the price you pay for filtering out noise.

Dysart himself was cautious. Per Wikipedia, he wrote that between 1946 and 1967 the NVI gave 17 significant signals, of which 14 proved right and 3 wrong, but he also admitted it “seriously erred” in 1963-64 and 1968. By 1969 he had reduced the weight he gave it for primary trends, while still crediting it with good early warnings of short-term reversals.

Using the NVI on Indian stocks and indices

StockCharts notes the indicator was designed with a major stock index in mind but can be applied to ETFs, single stocks or anything with volume. It adds a fair warning: on individual stocks, NVI sometimes produces excellent bullish or bearish divergences and sometimes looks completely out of sync with price. Keep that in mind before you build a whole system on it.

Here is a practical way we would use it on an Indian chart:

• Trend filter for the index. On the Nifty 50, check whether the NVI is above its 255-day EMA. If it is, and price is also above its own long average, you have two independent reads pointing the same way. If price makes new highs while the NVI is flat or falling below its average, the rally is being carried by high-volume days alone, which deserves a closer look.
• Correction inside an uptrend. StockCharts shows a two-average setup: the general trend is up while the NVI stays above its 255-day EMA, and a drop below a shorter 100-day EMA marks a correction within that trend. A swing trader can use the return above the 100-day line as a cue that the pullback may be ending.
• Trendlines on the indicator. You can draw trendlines on the NVI itself. In the StockCharts example, breaks above falling trendlines on the NVI marked the end of corrective periods and came before significant price moves in the underlying ETF.
• Divergence with price. A stock that drifts lower on quiet days while bouncing only on heavy-volume days will show a falling NVI even if price looks stable. That kind of divergence is worth noting, though, as StockCharts warns, it will not work on every stock.

A second, hypothetical stock example

Imagine a mid-cap that trades between ₹480 and ₹520 for three months. Price is going nowhere. But on most low-volume sessions it closes slightly higher, and the big drops all come on days of heavy selling that then fade. Because the NVI only counts the quiet days, it grinds upward through the range and eventually crosses above its 255-day EMA, while price is still stuck. That is the pattern Dysart described in 1969: Wikipedia quotes him saying signals are “most authentic when the NVI has moved sideways for a number of months in a relatively narrow range”, which is essentially a base forming on the indicator.

Would we buy on that alone? No. But it is a reason to put the stock on a watchlist and wait for price to confirm with a breakout above ₹520 on decent volume. The NVI gives you the hint; price has to give you the trade.

NVI is a hint, not a trigger. Let the indicator put a stock on your radar, then make price do the confirming.

Where the negative volume index goes wrong

Every indicator has blind spots, and this one has a few specific to how it is built.

Structural volume changes. Wikipedia points out that Dysart’s NVI lost much of its edge in the mid-1960s, probably because NYSE daily volume kept rising as more issues were listed, so prices rose on days that still registered as higher volume. The same risk applies to any market where turnover trends higher for years: a long uptrend in volume can quietly starve the NVI of updates.

Event days get ignored. Budget day, an RBI policy surprise or a big results reaction can easily come with a volume spike. The NVI skips all of it. That is intentional, but it means the indicator says nothing about how the market handled the biggest news of the month.

Single-stock noise. In a thinly traded stock, volume can swing wildly from day to day for reasons that have nothing to do with informed buying, like a block deal or index rebalancing. The day-to-day comparison then becomes almost random.

Lag. A 255-day signal line is slow. By the time the NVI crosses it, a good part of the move can be over.

This is why StockCharts’ bottom line is that NVI should not be used on its own. Pair it with price structure, support and resistance, and a momentum or money-flow reading. If you already use the Money Flow Index, the two make a sensible pair, since MFI weighs volume on every day while NVI looks only at the quiet ones. A momentum oscillator such as the Percentage Price Oscillator can add timing on top.

Setting it up in your charting tool

If your charting platform offers NVI, it will usually sit with the volume indicators. On StockCharts it is set as a cumulative indicator that cannot be adjusted, but you can change the length of its moving average; setting that parameter to 1 effectively removes the average. If your platform uses a base of 100 rather than 1,000, nothing changes in how you read it.

A simple starting template:

• NVI with a 255-day EMA for the primary trend.
• A 100-day EMA on the NVI to spot corrections inside that trend.
• Price with its own 200-day moving average, so you can see whether price and NVI agree.

Test it on the Nifty first, where volume data is cleaner and the indicator was designed to work, before trying it on individual mid-caps and small-caps.

Quick recap

The negative volume index updates only on days when volume falls, on the idea that informed money is more visible on quiet sessions. Dysart built it on net advances in the 1930s; Fosback rebuilt it on index price changes and added the one-year moving-average rule. Above its long average, it has historically been a strong confirmation of a bull market in US data; below, it has been far less reliable as a bear warning. It lags, it whipsaws, and on single stocks it can be hit or miss. Used as one input alongside price, it can tell you something the busy days hide. For more on building a complete process, see our guide to technical analysis for Indian stocks.

Disclaimer: This article is for educational purposes only and is not investment advice. The examples use hypothetical figures. Trading in stocks and derivatives carries risk; please do your own research or consult a SEBI-registered investment adviser before making any investment decision.