Hindenburg Omen Tracker: Today's Signal, McClellan Oscillator & Cluster History
Free Hindenburg Omen tracker — a breadth-based crash-risk indicator that fires when both new 52-week highs AND new 52-week lows are elevated simultaneously, with McClellan Oscillator confirmation. This is a broad-universe variant computed daily from our own ~5,600-symbol US equity database with SPY as the trend filter, not the classic NYSE-only specification — see the methodology comparison for the differences.
Current risk read
- Severity
- Elevated
- As of market close
- 2026-06-05
- Last trigger
- 2026-06-05
- Cluster window
- expires in ~30 trading days without a new trigger
Since prior session:a new trigger fired today; 30-day trigger count 6 → 7; McClellan -4.5 → -123.8 (deteriorating).
Today's reading
As of market close on June 5, 2026, the Hindenburg Omen is ACTIVE — all four trigger conditions were met. 153 stocks made new 52-week highs (2.93% of 5,224 issues) and 163 made new 52-week lows (3.12%). The McClellan Oscillator closed at -123.79 (bearish breadth). SPY closed at $737.55, above its 50-day moving average of $713.51. There have been 7 triggers in the last 30 trading days — a confirmed cluster, the configuration historically associated with elevated crash risk.
Hindenburg signal2026-06-05 · close
IN CLUSTER
Daily triggerACTIVE·7 triggers in last 30d
McClellan
-123.79
bearish
New Highs
153
2.93%
New Lows
163
3.12%
Hindenburg Omen as of market close 2026-06-05: daily trigger ACTIVE with 7 triggers in the last 30 trading days. SPY $737.55 above its 50-day SMA ($714); universe 5,224 issues.
Source
Daily OHLCV for ~5,600 US equities (our own price database)
Methodology
All four Omen conditions evaluated daily; clusters within a 30-trading-day window tracked separately
Updates
Daily after market close (~1pm PT)
Last: 2026-06-06
Range:
Hindenburg Omen
Signal Status
ACTIVE
Cluster Status
CONFIRMED
McClellan Osc.
-123.79
NH% / NL%
2.9% / 3.1%
SPY vs 50d SMA
Above ($714)
New Highs
153
New Lows
163
Cluster Count
7
SPY Price
50-Day SMA
Signal Day
Cluster Period
McClellan Oscillator
Positive
Negative
SPY Price
Zero Reference
01
Methodology — classic spec vs our variant
We implement a variant of the Hindenburg Omen, not the literal classic specification. The differences are explicit and material. If you compare trigger dates between this page and another tracker that uses the NYSE-only methodology (e.g., the StockCharts implementation), you should expect them to differ — sometimes by weeks. This table documents every choice we make.
| Condition | Classic NYSE Omen | The Trading Tools implementation | Practical effect |
|---|---|---|---|
| Universe | NYSE-listed issues only (~3,000 stocks; the original 1995 specification) | Our full symbol universe (~5,600 US equities across exchanges; valid-issue count varies by date) | Different universes produce different daily highs/lows and different trigger dates. |
| New highs / new lows threshold | ≥ 2.2% of total NYSE issues (some modern variants use 2.8%) | ≥ 2.2% of valid issues in our universe | Same percentage, different denominator. |
| Minimum-issue floor | The smaller of new highs / new lows must be ≥ 75 issues | Not enforced | A few historical triggers in our data had a smaller side below 75. Classic Omen would not have flagged them. |
| Trend filter | NYSE Composite 10-week MA rising (or NYSE Composite > level 50 trading days ago) | SPY closing price > SPY 50-day SMA | A meaningful difference. SPY can be above its 50d while a rising-NYSE-Composite-MA test would fail, and vice versa. |
| McClellan Oscillator | Negative McClellan Oscillator (some variants use ratio-adjusted) | Negative raw McClellan = EMA19(net advances) − EMA39(net advances), unadjusted | Raw oscillator can be unstable when the underlying universe is changing in size. |
| ≤ ratio constraint | New highs ≤ 2× new lows | New highs ≤ 2× new lows | Implemented identically. The asymmetry (lows can exceed 2× highs but not the reverse) is intentional in the classic spec. |
| Cluster window | ~36 days (originally 36 calendar days; some modern variants use 30 trading days) | 30 trading days (matches one of the modern variants) | Pick a window — both definitions exist in the literature. Our chart shading and "in cluster" status both use this window. |
| Universe stability & survivorship | Not relevant — NYSE issue count is reasonably stable and includes delisted issues as of each date | Our history contains today’s listings only: ~2,000 of them traded in 2010, growing to ~5,600 now. Signals are suppressed when valid issues < 2,000. | Survivorship caveat: delisted names are absent from early years, which skews historical breadth toward survivors — our variant likely under-counts new lows (and therefore under-triggers) in the early window. |
Classic-spec sources: Jim Miekka's 1995 work as documented and refined by Robert McHugh; the StockCharts/ChartSchool reference; independent technical-analysis publishers; and the McClellan Financial Publications methodology.
02
Trigger history
Recent single-day triggers
| Date | SPY | McClellan | NH% | NL% |
|---|---|---|---|---|
| 2026-06-05 | $738 | -123.8 | 2.93% | 3.12% |
| 2026-06-03 | $754 | -123.1 | 3.48% | 3.08% |
| 2026-05-20 | $741 | -75.9 | 2.70% | 2.74% |
| 2026-05-18 | $739 | -180.1 | 2.89% | 3.00% |
| 2026-05-14 | $748 | -90.5 | 4.37% | 2.40% |
| 2026-05-13 | $742 | -142.8 | 4.19% | 4.63% |
| 2026-05-11 | $739 | -38.2 | 5.75% | 3.76% |
| 2026-02-04 | $686 | -42.4 | 9.77% | 4.97% |
| 2026-02-03 | $690 | -47.4 | 8.74% | 5.72% |
| 2026-02-02 | $695 | -49.1 | 5.60% | 3.31% |
Cluster periods (2+ triggers in 30 trading days)
| Cluster start | Cluster end | Peak 30d count |
|---|---|---|
| 2026-05-13 | 2026-06-05 | 7 |
| 2026-01-29 | 2026-03-18 | 5 |
| 2025-10-13 | 2025-12-29 | 11 |
| 2025-08-11 | 2025-09-12 | 2 |
| 2025-02-12 | 2025-03-24 | 2 |
03
Reported historical episodes
How the Omen is reported to have behaved across the past four decades — now cross-checked against our own 15-year trigger record where the windows overlap (our data begins Dec 2010). The “Our variant” column shows what we independently computed from raw price data: 6 of the 7 in-window episodes are confirmed by our own triggers, including the Nov 2019 – Jan 2020 cluster ahead of COVID. Some episodes are widely cited as correct calls (1987, 2000, 2007, 2019–20); others as false positives (2010, 2014, 2017, 2021) — and our data reproduces the 2021 false positives too. The pattern is a roughly 50/50 hit rate, consistent with the Omen being informative but unreliable as a standalone signal.
| Period | Trigger pattern | SPY outcome | Read | Our variant | Source | Confidence |
|---|---|---|---|---|---|---|
| Aug 1987 (pre-Black Monday) | Reported cluster activity | –22% on Oct 19, 1987 (Black Monday) | Widely cited as a successful call — Omen reportedly flagged the regime weeks before the crash. | Pre-dates our window | McClellan Pubs | Widely cited |
| Oct 1999 – Mar 2000 (dot-com top) | Multiple reported clusters | S&P –49% over the next 2.5 years | Persistent breadth divergence as mega-cap tech pulled the index higher while many stocks made new lows. | Pre-dates our window | StockCharts | Widely cited |
| Jun – Aug 2007 (pre-GFC) | Reported cluster in June, more in August | S&P –56% peak-to-trough by Mar 2009 | Among the most-cited successful calls — reportedly fired well before the credit crisis became visible in equities. | Pre-dates our window | McHugh (reported) | Widely cited |
| Jul – Aug 2010 | Reported cluster | No major drawdown — market resumed uptrend | A widely-discussed false positive. The cluster appeared during a mid-bull-market chop; markets rallied through 2011. | Pre-dates our window (data begins Dec 2010) | StockCharts | Widely cited |
| Aug 2013 | Reported single triggers | Brief –4% pullback then continued uptrend | Apparent false positive during a normal correction. | Confirmed — 2 triggers (Aug 6–7, 2013) | McHugh (reported) | Reported |
| Jul 2014 | Reported cluster | No major drawdown | False positive in an extended bull market. | Not confirmed — zero triggers in our variant (universe/trend-filter divergence) | McHugh (reported) | Reported |
| Aug 2015 | Reported cluster around the China-devaluation drop | S&P –12% (sharp but brief) | Concurrent with rather than clearly leading the move. Coincident signal. | Confirmed — 3 triggers in late July, preceding the Aug 24 crash | StockCharts | Reported |
| Aug 2017 | Reported cluster | No major drawdown | False positive. | Confirmed — 2 triggers (Aug 15–16, 2017) | McHugh (reported) | Reported |
| Sep 2018 – Dec 2018 | Multiple reported clusters | S&P –19% peak-to-trough | Reportedly fired through the Q4 selloff — partial credit (no crash, but a meaningful correction). | Confirmed — 12 triggers Sep–Dec 2018, a sustained cluster through the selloff | StockCharts | Reported |
| Sep 2019 – Feb 2020 (pre-COVID) | Sustained reported cluster activity | S&P –34% in 33 days starting Feb 19, 2020 | The Omen was reportedly active in the months before COVID. Correct call in retrospect, though no one could have known the trigger. | Confirmed — 9 triggers Nov 2019 – Jan 27, 2020, ahead of the Feb 19 top | McClellan Pubs | Widely cited |
| 2021 | Several reported single triggers | Continued bull through year-end 2021 | False positives during melt-up phase. | Confirmed — 10 triggers Jul–Nov 2021 (and no major drawdown followed: false positives in our data too) | McHugh (reported) | Reported |
Episodes from 2013 onward are cross-checked against our own computed trigger record (broad-universe variant, not the literal NYSE spec); pre-2011 episodes remain externally-reported claims that we have not re-computed from raw NYSE data. The Omen has been reportedly most predictive when triggers appear alongside other deteriorating conditions — credit spreads widening, VIX rising, sector rotation accelerating. False positives tend to cluster in the middle of bull markets when other risk indicators are calm.
How Hindenburg Omen Tracker Works
- 1Evaluate Hindenburg Omen conditions daily across a ~5,600-stock universeEach trading day after the close we check whether all conditions are simultaneously true: (1) both new 52-week highs and new 52-week lows are at least 2.2% of valid issues; (2) the SPY closing price is above its 50-day simple moving average (our trend filter — the classic Omen uses the NYSE Composite's 10-week MA rising); (3) the McClellan Oscillator is negative; (4) new highs do not exceed 2× new lows. The universe of valid issues is computed from our own daily price database of ~5,600 US equities across exchanges, not the literal NYSE listing — see the methodology comparison on the page.
- 2Flag single-day triggers and clustersA single day meeting the conditions is a "trigger". Our implementation uses a 30-trading-day window to count clusters; the page chart highlights periods where 2+ triggers fall within that window. Clusters historically carry more interpretive weight than isolated single-day triggers, though both have a high false-positive rate.
- 3Surface the signal alongside SPY and McClellan context, with universe filteringThe chart on this page shows SPY price with the 50-day moving average overlay, vertical red lines marking each trigger day, and shaded regions marking cluster periods. A second chart shows the McClellan Oscillator. The data window runs from late 2010 to the present — 15+ years. To keep the 2.2% thresholds statistically meaningful, signals are suppressed on days when the valid-issue count is below 2,000 (the universe grows from ~2,000 issues in 2010 to ~5,600 today, since it contains today's listings only — see the survivorship note in the methodology table).
Who Uses Hindenburg Omen Tracker
Long-Term Investors
Use the Omen as a regime warning, not a sell signal. Confirmed Omen clusters historically have appeared during transitional regimes when breadth is mixed and risk is rising — a signal to review portfolio risk, not necessarily to liquidate.
Risk Managers
One input among several in a multi-factor risk dashboard. The Omen is most useful when it confirms what credit spreads, VIX, and other breadth indicators are already saying. Standalone Omen signals are noisy.
Tactical Traders
Watch for confirmed clusters as one of multiple signals before reducing equity exposure. Combining an Omen cluster with rising VIX and widening credit spreads has historically been a stronger combined signal than any one alone.
Researchers and Educators
The Omen is a textbook example of a breadth-based market indicator. Its historical track record (mix of correct calls and false positives) makes it a useful case study for teaching the limits of single-indicator forecasting.
Pro Tips
01
Confirmed clusters matter more than single-day triggers
A single Omen trigger is statistically noisy and easily produced by a few news-driven days where new highs and new lows both expand. Two or more triggers within a few weeks (a "cluster") is the version most analysts treat as the real signal.
02
False positives are common — don't sell on the Omen alone
The Omen has fired multiple times during extended bull markets (notably 2010, 2014, 2017, 2021) without a major drawdown following. Acting only on the Omen would have meant missing many continuing rallies.
03
The signal is most reliable in transitional regimes
Look at when the Omen has fired historically: late 2007, mid-2019/early-2020, occasional clusters in choppy periods. The common thread is regimes where breadth is mixed — bull-market clarity tends not to produce Omen triggers.
04
Cross-confirm with VIX, credit, and other breadth
When an Omen cluster coincides with VIX backwardation, credit spreads widening, and overall MA breadth deteriorating, the combined signal is far more meaningful than any one alone. When the Omen fires alone with all other indicators calm, it's usually noise.
05
McClellan Oscillator below zero is a regime cue
Even without the full Omen trigger, a sustained negative McClellan Oscillator (the chart on this page) signals weakening internals. Watch for the Oscillator to flip back positive as a sign the deterioration may be ending.
06
The cluster window matters
Some implementations use 30 days, others 36. Triggers within roughly six trading weeks indicate persistent rather than fleeting deterioration. Single-trigger noise versus persistent regime change is the central question; our implementation uses 30 trading days.
07
Position-size your reaction proportional to confirmation
A useful framework: single trigger = no action (still in the noise). Confirmed cluster = reduce equity beta by 10–20%. Confirmed cluster + VIX backwardation + credit spreads widening = larger reduction. The Omen alone rarely justifies a major position change.
08
Watch for the Omen during melt-ups
Counterintuitively, the Omen often fires near euphoric tops because some stocks are still making new highs while sector rotation is producing new lows in lagging names. The 1999 dot-com top and 2007 cycle peak both showed Omen activity in advance.
Common Issues & Solutions
I see no Omen triggers in the data window▾
Stretches without triggers are normal. The Omen requires very specific simultaneous conditions and rarely fires. Historical context: the Omen has fired meaningful clusters roughly once every 1–3 years on average since 2000.
A trigger fired but nothing happened — was it wrong?▾
Yes, this happens often. The Omen has a high false-positive rate. Roughly half of historical triggers have NOT been followed by major drawdowns. The signal is more useful when triggers cluster (2+ within a few weeks) and when it confirms what other indicators are saying.
Why is the McClellan Oscillator near zero?▾
The McClellan Oscillator is the difference between 19-day and 39-day EMAs of net advances. It's designed to oscillate around zero, so most days it's near zero. Sustained excursions away from zero (especially below zero, which is one of the four Omen conditions) are the signal.
Why do your trigger dates differ from other Omen trackers?▾
Universe and trend filter. The original 1995 specification used NYSE-listed issues and the NYSE Composite trend filter; our variant uses a ~5,600-stock cross-exchange universe and SPY vs its 50-day SMA. Different universes produce different daily highs/lows counts and therefore different trigger dates — sometimes by weeks. Always check the methodology of any Omen tracker you compare against; ours is documented in full in the comparison table on this page.
Frequently Asked Questions
What is the Hindenburg Omen?▾
The Hindenburg Omen is a technical-analysis indicator developed by Jim Miekka that attempts to flag elevated crash risk by identifying simultaneous expansion of both new 52-week highs AND new 52-week lows. Named after the 1937 Hindenburg airship disaster, the Omen reflects the idea that when stocks are pulling apart into both extremes at the same time, the market is showing signs of internal disagreement. The signal has a high false-positive rate — it has flagged some major events (1987, 2000, 2007, 2019–20) but also fired during many extended bull markets without a major drawdown following.
How is the classic Hindenburg Omen calculated?▾
The classic definition (as popularized by Jim Miekka and Robert McHugh) has four conditions, all on the same trading day: (1) both new 52-week highs and new 52-week lows are at least 2.2% of total NYSE-listed issues; (2) the smaller of the two counts is at least 75 issues; (3) the NYSE Composite 10-week moving average is rising; (4) the McClellan Oscillator is negative. A "confirmed" Omen requires 2+ triggers within roughly 36 days.
How does The Trading Tools implementation differ from the classic Omen?▾
Our implementation is a variant — we make several explicit choices that differ from the classic NYSE-only specification. (1) Our universe is all symbols in our price database (~5,600 US stocks across exchanges), not the NYSE-listed subset. (2) Our trend filter is SPY > 50-day SMA, not NYSE Composite 10-week MA rising. (3) We do not enforce the 75-issue minimum. (4) Our cluster window is 30 trading days. (5) We use the unadjusted McClellan Oscillator (raw EMA difference rather than ratio-adjusted). The full comparison table is on the page above. Trigger dates from this page may not match other Omen trackers that hew closer to the classic NYSE methodology.
Who created the Hindenburg Omen?▾
The indicator was developed in the 1990s by Jim Miekka, a blind market technician. It was popularized by Robert McHugh, who refined the definition and tracked its triggers publicly. The name was suggested by Miekka's friend Kennedy Gammage and is intended to evoke the disaster's sudden, catastrophic nature.
How accurate is the Hindenburg Omen?▾
Its track record is mixed. The Omen correctly fired before the 1987 crash, the 2000 dot-com top, late 2007 (before the 2008 crash), and late 2019 / early 2020 (before COVID). But it also triggered in 2010, 2014, 2017, and 2021 with no major drawdown following. Various academic studies have found accuracy rates between 25% and 75% depending on how "crash" is defined and whether you require confirmed clusters or accept single-day triggers.
When was the last time the Hindenburg Omen triggered?▾
The signal status card at the top of this page shows whether today is currently in an active trigger or cluster state, and the plain-English daily reading below it summarizes the current conditions. Each red vertical line on the chart marks a trigger day in our data window; confirmed clusters (shaded red regions) are rarer than single-day triggers and historically more meaningful.
Should I sell stocks if the Hindenburg Omen triggers?▾
Probably not on the Omen alone. The historical false-positive rate is high enough that selling on every trigger would have meant missing many continuing rallies. A more nuanced approach: a confirmed Omen cluster should make you more attentive to other risk signals — VIX, credit spreads, breadth divergences, news flow. If multiple risk signals align with the Omen, that combined message is much stronger than the Omen alone.
What is a "confirmed" Hindenburg Omen cluster?▾
A cluster is when 2 or more single-day triggers occur within a defined window — the classic Omen uses ~36 days; our implementation uses 30 trading days. A single trigger is often noise (one news-driven day where both highs and lows expanded), but multiple triggers within a few weeks indicate persistent breadth deterioration rather than a one-off event. The "Cluster Status" indicator at the top of this page shows whether the current period contains a cluster.
What is the McClellan Oscillator?▾
The McClellan Oscillator is a market-breadth indicator built from the difference between 19-day and 39-day exponential moving averages of net advances (advances minus declines). It oscillates around zero. Positive values indicate broadly improving breadth; negative values indicate deteriorating breadth. The Hindenburg Omen requires the McClellan Oscillator to be negative as one of its four trigger conditions, ensuring the broader breadth tape is weakening when the Omen fires.
Why does the Hindenburg Omen require both new highs AND new lows?▾
The simultaneous expansion of both is unusual and signals internal disagreement. In a normal bull market, you see lots of new highs and few new lows. In a normal bear market, the reverse. When you see both simultaneously, sectors and individual stocks are diverging — some are still trending up while others are breaking down. This kind of fragmented market is historically associated with regime instability and elevated risk.
How long does a Hindenburg Omen warning last?▾
The standard interpretation is that a confirmed Omen cluster signals elevated crash risk for approximately 30 days from the most recent trigger date. After that window, the warning is considered "expired". In practice, multiple analysts extend the watch period through any continuing breadth deterioration, since the underlying conditions often persist beyond a strict 30-day window.
What is the difference between the Hindenburg Omen and the Titanic Syndrome?▾
Both are breadth-based crash indicators. The Titanic Syndrome (developed by Bill Ohama) fires when an index makes a new 52-week high and within 7 trading days has more new 52-week lows than new highs — a much narrower trigger. The Hindenburg Omen is broader and requires simultaneous extreme breadth conditions. Both signals have similar underlying logic (breadth disagreement) but different specific triggers.
Does the Hindenburg Omen work on Nasdaq or just NYSE?▾
The classic 1995 specification was developed using NYSE-listed issues. Some modern implementations use the S&P 500 or combined NYSE + Nasdaq universes. Trigger dates and historical accuracy vary depending on which universe you use. Our implementation uses our full symbol universe (~5,600 US equities across exchanges) — see the methodology comparison block on this page. If you compare against another tracker and see different trigger dates, the underlying universe and trend filter are usually the reasons.
How often does the Hindenburg Omen trigger?▾
Historically, single-day triggers have occurred a few times per year on average, with periods of activity (typically 1–3 in a month) followed by quiet stretches of months. Confirmed clusters (2+ triggers within a few weeks) are much rarer — roughly once every 1–3 years on average since 2000. Long stretches without a trigger are common and normal.
Is the Hindenburg Omen a leading or lagging indicator?▾
When it works, the Omen is a leading indicator — its forward window is intended to anticipate drawdowns rather than confirm them. When it produces false positives, of course, it leads nothing. Treat it as a leading indicator candidate that requires confirmation from other signals before being acted on.
Can I combine the Hindenburg Omen with other indicators?▾
Yes — and most practitioners do. The most common combinations: Omen + VIX backwardation (volatility confirmation), Omen + high-yield credit spreads widening (credit confirmation), Omen + falling MA breadth (broader breadth confirmation). A confirmed Omen with all three confirmations is a much stronger combined signal than any one alone.
Related Tools
Last updated: 2026-06-05