Negative Volume Index (NVI): Tracking Smart Money on Quiet Volume Days

Paul L. Dysart Jr. was not your typical Wall Street analyst. A native of Iowa who worked on Chicago's LaSalle Street during the roaring 1920s, Dysart gave up his Chicago Board of Trade membership and turned to studying market breadth data — specifically the advance-decline statistics that most traders at the time ignored entirely. In 1933, he launched the Trendway weekly stock market letter, which he published until his death in 1969.

Starting in 1936, Dysart began sorting market days into two buckets based on a simple question: was today's volume higher or lower than yesterday's? He tracked the cumulative sum of advancing minus declining stocks on each type of day separately. The low-volume series became the Negative Volume Index. The high-volume series became the Positive Volume Index.

Dysart's original calculation worked like this: if today's NYSE volume was less than yesterday's, subtract the number of declining stocks from advancing stocks and add the result to a running cumulative total (starting from zero). On days when volume increased, NVI simply held its previous value — completely frozen, as if nothing happened. PVI followed the identical logic but tracked the high-volume days instead.

Here's where things got interesting. Dysart initially expected PVI — the high-volume index — to be the more valuable tool. High volume means action, drama, and presumably useful information, right? But after three decades of observation, he changed his mind completely. In 1967, Dysart wrote that NVI had "proved to be the most valuable of all the breadth indexes." He nicknamed it AMOMET — an acronym for "A Measure Of Major Economic Trend" — signaling that he saw it as something far more significant than just another technical indicator.

Why did low-volume days carry more information than high-volume days? Dysart's reasoning was rooted in crowd psychology. High-volume sessions are dominated by reactive, emotional trading — news-driven surges, panic selling, FOMO-fueled buying. The noise-to-signal ratio is terrible. But on quiet days, when the general public isn't piling in, the remaining participants are disproportionately professionals and institutions who trade based on research, fundamentals, and long-term positioning rather than headlines.

The core insight remains powerful even today: volume tells you who is participating, and who is participating tells you how much to trust the price movement. A rally on declining volume isn't weak — according to NVI logic, it's potentially the most trustworthy kind of move because it's driven by participants who don't need the crowd's validation to act.

Dysart's work predated modern portfolio theory, efficient market hypothesis, and algorithmic trading by decades. Yet his observation that informed money behaves differently from uninformed money on the volume dimension has been validated repeatedly. It's one of those rare ideas from the pre-computer era that didn't just survive the transition to modern markets — it got stronger as institutional trading became more sophisticated and deliberate about minimizing market impact.

The modern version of NVI, refined by Norman Fosback in his 1976 book Stock Market Logic, translates Dysart's breadth-based concept into a price-based calculation that works on any single instrument — not just the broad market.

The rule is almost comically simple:

If today's volume < yesterday's volume: NVI(today) = NVI(yesterday) + [(Close(today) - Close(yesterday)) / Close(yesterday) x NVI(yesterday)]

If today's volume >= yesterday's volume: NVI(today) = NVI(yesterday)

That's it. Two conditions. The starting value is conventionally set to 1,000.

Let's walk through a five-day example to see how NVI actually builds:

Notice Day 3: price dropped, but volume increased. NVI completely ignores this session. The drop happened on a high-volume day, which NVI attributes to crowd behavior — presumably less informed and less indicative of the real trend. Day 5 also shows a price decline, but again volume rose compared to Day 4, so NVI stays frozen at 1,009.98.

This selective updating is what makes NVI fundamentally different from most indicators. A standard moving average processes every single bar. RSI processes every bar. MACD processes every bar. NVI says "I don't care what happened today" roughly half the time. On a typical stock, volume decreases compared to the prior day about 45-55% of sessions, so NVI updates on slightly less than half of all trading days.

The practical consequence of this selective approach is that NVI moves slowly and deliberately. It doesn't whipsaw on volatile news days because those tend to come with volume spikes — which NVI ignores. It doesn't react to earnings announcements (massive volume), Fed decisions (massive volume), or geopolitical shocks (you guessed it — massive volume). NVI only cares about the quiet days in between, when the market digests and repositions.

What NVI's direction tells you:

One thing worth clarifying: NVI doesn't literally track institutional order flow. No indicator can do that without actual order book data. What NVI does is use volume as a proxy to filter for sessions where institutional activity is more likely to dominate. It's an approximation — a pretty good one historically, but still an approximation. Don't treat NVI signals as confirmed insider knowledge. Treat them as a probabilistic filter that tilts the odds in your favor by focusing on the quieter, potentially more informative sessions.

Norman Fosback didn't just modernize the NVI formula — he built a complete trading signal around it that became one of the most cited market-timing rules in technical analysis literature.

The system is elegant in its simplicity: calculate a 255-day exponential moving average of NVI, then check whether NVI is above or below that line. That single comparison gives you a bull/bear market probability reading.

Why 255 days specifically? In the era when Fosback conducted his research, the US stock market had approximately 255 trading days per year. So the 255-day EMA smooths exactly one year of NVI data, filtering out seasonal patterns and short-term noise to capture the underlying trend in smart money behavior.

Fosback's findings, published in Stock Market Logic (1976), were striking:

That asymmetry is the most important thing to understand about NVI signals. The bullish reading is remarkably reliable — when smart money is trending positive (NVI above its long-term average), the broader market is almost certainly in an uptrend. But the bearish reading? It's barely informative. A 53% chance of a bear market means there's still a 47% chance it's actually a bull market. You wouldn't bet your portfolio on those odds.

This asymmetry has a logical explanation. Bull markets tend to be long and steady, with institutional investors gradually accumulating positions over months and years. NVI captures this accumulation pattern well because it builds slowly during sustained quiet-day buying. Bear markets, by contrast, tend to be shorter and more volatile — driven by sudden liquidation events, margin calls, and panic. These events happen on high-volume days that NVI ignores entirely. So NVI is structurally better at identifying bulls than bears.

How to use the signal in practice:

1. NVI crosses above the 255-day EMA — This is your primary buy signal. Historically, entering the market when this crossover occurs and holding until the reverse crossover produced strong risk-adjusted returns on major indices. It doesn't tell you exactly when the bull market started — the signal often lags the actual bottom by weeks or months — but it confirms that the trend has shifted.

2. NVI stays above the 255-day EMA — As long as this condition holds, treat the market as bullish. Dips in price are buying opportunities, not reasons to panic. The 96% reliability means that only in very rare circumstances (roughly 1 in 25 signals historically) will NVI be above its EMA during an actual bear market.

3. NVI crosses below the 255-day EMA — This is where it gets tricky. You might be tempted to go short or sell everything, but the 53% probability means you're barely better than guessing. The more prudent approach: treat a downward crossover as a reason to reduce exposure and tighten risk management, not as a definitive sell signal. Move to cash partially, tighten stop losses on existing positions, and wait for confirmation from other indicators before getting aggressively bearish.

4. NVI stays below the 255-day EMA — The longer NVI remains below, the more concerning the signal becomes. A brief dip below and quick recovery is common and not alarming. A sustained period of 3+ months below the EMA deserves attention and likely warrants defensive positioning.

Some platforms default to a 255-day simple moving average (SMA) rather than an EMA. Both work, but the EMA gives slightly more weight to recent NVI values, making crossover signals a touch faster. The difference is marginal. If your charting platform only offers SMA smoothing for NVI, that's perfectly fine — Fosback's original research used both approaches and found similar results.

A practical note for MetaTrader 5 users: NVI is not a default built-in indicator on MT5, but several quality implementations are available in the MQL5 marketplace. Look for versions that include the 255-period signal line and display both NVI and its EMA in the indicator window. TradingView has NVI available in its community scripts with the EMA overlay.

NVI and PVI are two halves of the same analytical framework. They split every trading session into two categories based on volume direction, and each index tracks what price does in its respective category. Understanding both — and especially the relationship between them — gives you a far more complete picture than either one alone.

Here's the side-by-side comparison:

The reliability numbers tell an interesting story. NVI is excellent at calling bull markets but mediocre at calling bears. PVI is somewhat better at calling bears (~74% below its EMA) but less reliable for bulls (~67% above its EMA). They complement each other almost perfectly — NVI handles the upside, PVI handles the downside.

This complementary nature leads to four possible market states when you combine both indicators:

State 1: NVI above EMA + PVI above EMA (Both Bullish) Both smart money and the crowd are trending positive. This is the strongest bullish configuration. Broad participation means the rally has legs. Think of sustained bull markets where every dip gets bought — both institutions on quiet days and retail on busy days are pushing prices higher.

State 2: NVI above EMA + PVI below EMA (Smart Money Bullish, Crowd Bearish) This is the most interesting state — and arguably the most profitable to trade. Smart money is accumulating on quiet days while the crowd is selling on high-volume days. This divergence often occurs during the early stages of a new bull market, when the public is still traumatized from the previous decline but institutional buyers are already positioning. Fosback's research showed this state often precedes the most explosive rallies.

State 3: NVI below EMA + PVI above EMA (Smart Money Bearish, Crowd Bullish) Dangerous territory. The crowd is enthusiastically buying on high-volume days, but smart money positioning on quiet days is deteriorating. This pattern frequently appears near market tops, when retail excitement is at its peak but institutional investors are quietly distributing. Late 2021 into early 2022 showed elements of this pattern — retail traders were still piling into growth stocks while NVI was weakening.

State 4: NVI below EMA + PVI below EMA (Both Bearish) Both smart money and the crowd are trending negative. This is the clearest bearish signal available from the NVI/PVI framework. Everyone is selling, regardless of volume conditions. Full risk-off mode.

Using both indicators together also creates what you might call a "confidence meter" for your trades. When NVI and PVI agree — both above or both below their EMAs — you can trade with higher conviction and potentially size up. When they disagree, you're in a transitional zone where caution and reduced position sizes are warranted.

A practical workflow: check NVI/PVI status weekly (they move slowly, so daily checks are overkill). If both are bullish, focus your trading on long setups and be aggressive with winners. If they disagree, reduce overall exposure and keep trades smaller. If both are bearish, consider moving to the sidelines or trading only short setups. This framework won't give you entries or exits, but it provides a background filter that keeps you aligned with — or at least aware of — the institutional versus retail dynamic.

One caveat: both NVI and PVI were designed and tested primarily on broad market indices. On individual stocks, the smart money / crowd money distinction becomes blurrier because a single stock's volume can spike due to company-specific events (earnings, FDA approvals, M&A rumors) that involve both informed and uninformed participants simultaneously. The NVI/PVI framework works best on indices, ETFs, and highly liquid instruments where the crowd vs. smart money distinction holds more cleanly.

Let's address the skepticism head-on: an indicator designed for 1930s NYSE breadth data, refined in the 1970s for stock indices, applied to a 24-hour decentralized forex market using tick volume instead of real exchange volume. On paper, it sounds like you're trying to play jazz with a harpsichord. But the picture is more nuanced than the skeptics suggest — and more limited than the enthusiasts claim.

The Volume Problem

The biggest theoretical objection to NVI on forex is the volume data itself. Forex is an over-the-counter market with no centralized exchange. The "volume" your platform displays is tick volume — the number of price changes per period — from your specific broker. Different brokers see different tick volumes because they process different subsets of global order flow.

Does tick volume work as a proxy for NVI's purposes? The answer is: reasonably well on major pairs during liquid hours, and poorly everywhere else.

The saving grace is that NVI doesn't need perfectly accurate volume — it only needs to know whether volume went up or down relative to yesterday. Even with noisy tick volume data, the directional call (higher vs. lower than yesterday) tends to be consistent across brokers on major pairs. NVI is comparing relative volume changes, not absolute numbers, which makes it more robust to data quality issues than you might expect.

Timeframe Considerations

NVI was designed for daily charts, and that's where it works best on forex too. On D1, the volume comparison is between complete 24-hour periods, which smooths out within-session noise. On W1, NVI becomes extremely slow — updating only on weeks with lower total volume — but produces very clean long-term signals for position traders.

Sub-daily timeframes are not recommended. The smart money vs. crowd money premise doesn't scale well to intraday forex, where volume fluctuations within a session are driven by timezone overlaps (London/New York overlap = high volume, Asian session = low volume) rather than by differences in who is trading.

Practical Applications for Forex

Despite the limitations, NVI has legitimate uses in modern forex trading when applied correctly:

1. Weekly Trend Filter for Major Pairs Apply NVI with a 52-week (equivalent to 255-day) EMA on the weekly chart of EUR/USD or GBP/USD. When NVI is above the EMA, maintain a bullish bias for the pair — look for long setups on lower timeframes. When NVI crosses below, shift to neutral or slightly bearish. Because NVI moves so slowly on W1, you'll only get a signal change every few months, making it a background filter rather than an active trading tool.

2. Divergence Detection on D1 Watch for NVI divergences on daily charts. If EUR/USD price makes new highs but NVI makes lower highs, it suggests that the advances happening on low-volume days are weakening — potential early warning of a trend reversal. These divergences are rare (maybe 2-4 per year on a major pair), but when they appear, they deserve attention.

3. Combining NVI with Forex-Native Indicators NVI works best on forex as part of a multi-indicator approach. Use NVI's directional bias as a filter, then apply forex-native tools for timing:

• NVI bullish + RSI pullback to 40-50 — Higher-probability long entry
• NVI bullish + price at daily VWAP or session S/R — Context-rich entry zone
• NVI bearish divergence + bearish candlestick pattern — Potential short setup

4. Confirming Risk-Off / Risk-On Regimes On USD/JPY and other risk-sensitive pairs, NVI can help confirm whether quiet market sessions are showing accumulation (risk-on positioning) or distribution (risk-off). Since USD/JPY is heavily influenced by institutional carry trade flows, NVI's smart money premise has a clearer theoretical basis on this pair than on most.

The Honest Assessment

NVI is not a primary indicator for forex. Its theoretical foundation (exchange volume, centralized markets, clear retail vs. institutional volume patterns) maps better to stocks and indices than to decentralized FX. But as a secondary filter — a way to check whether quiet-session behavior aligns with your trade thesis — it adds a dimension that most forex indicators completely ignore. Nobody is going to build a forex strategy around NVI alone, and nobody should. But as one voice in a committee of indicators, the old professor from the 1930s still has something worth hearing.

For traders who want the strongest NVI signals while trading currencies, consider running NVI analysis on stock index futures (ES, NQ) or major index ETFs (SPY, QQQ) and using the resulting bull/bear regime as a background filter for your forex trades. If NVI says the equity market is in a high-confidence bull phase, that broader risk appetite context applies to risk-on currency pairs (AUD, NZD, emerging market currencies) as well.