Hindenburg Omen Flashes Again as U.S. Stock-Market Breadth Sends Mixed Signals

The Hindenburg Omen, a market-breadth warning that looks for an unusual mix of stocks making new 52-week highs and new 52-week lows while the broader market is still rising, triggered again on Monday under a widely used version of the indicator’s rules. The reading adds to a series of similar signals seen this summer, but it is not a stand-alone forecast of a crash. By Friday, one of the signal’s key breadth components had turned positive again.

The latest trigger is notable because it appeared with major U.S. indexes still near record territory. The S&P 500 ended Friday at 7,785.76, down 0.17% for the session but only about 0.4% below its 52-week high, while the NYSE Composite finished at 24,821.68, up 0.05% on the day. That combination of strong index levels and uneven participation beneath the surface is exactly the type of condition the Hindenburg Omen is designed to flag.

Monday’s breadth data produced another Hindenburg Omen after a cluster of similar signals in May and June. The indicator remains sensitive to the underlying market-breadth dataset, however, and different data providers can reach different conclusions about whether a particular session meets every qualifying threshold.

Monday’s breadth numbers met the classic warning tests

The logic behind the Hindenburg Omen is that a healthy rising market should usually produce a broad set of new highs and relatively few new lows. When large numbers of stocks are simultaneously reaching both extremes, the market can be interpreted as internally divided even though the headline indexes remain elevated.

On Monday, Reuters counted 297 new 52-week highs and 149 new lows on the New York Stock Exchange. McClellan Financial’s daily oscillator spreadsheet recorded 1,086 advancing issues and 1,646 declining issues that day, putting 2.8% of advances plus declines at roughly 76.5 issues. Both the new-high and new-low counts were comfortably above that threshold. The 297 new highs were also just below twice the 149 new lows, satisfying another version of the traditional test by a margin of only one issue.

The trend and oscillator conditions also lined up. The NYSE Composite closed Monday at 24,667.89, well above its May 28 close of 23,302.26, the level from 50 trading sessions earlier. McClellan’s spreadsheet showed its NYSE McClellan Oscillator at about minus 6.36 on Monday. Under the rules Jim Miekka supplied to market technicians, as described in McClellan Financial’s explanation of the indicator, the signal looks for a rising NYSE Composite, unusually high numbers of both new highs and new lows, and a negative McClellan Oscillator.

Those calculations help explain why Monday qualified in the SentimenTrader-based analysis. They also show why the indicator can be sensitive to data choices. The new-high versus new-low test was extremely close on the two-to-one rule, and McClellan Financial notes that its final breadth figures use WSJ/Barron’s data and can differ from calculations made with other feeds. In an indicator built from hard thresholds, even a small difference in issue counts can change a yes-or-no result.

There is another methodological wrinkle. McClellan’s summary of Miekka’s later rules says that once an Omen is initiated, the warning remains valid for 30 trading days and additional signals during that window should be ignored. SentimenTrader’s historical work, by contrast, studies repeated occurrences in clusters. The cluster statistics are therefore best understood as one research convention for measuring repeated breadth stress, not as a universal official Hindenburg Omen series.

History shows weaker returns after signals, but plenty of false alarms

The Hindenburg Omen has a memorable name, but its historical record is less dramatic than the label suggests. Dow Jones Market Data, using Hindenburg Omen data supplied by SentimenTrader, found that the S&P 500’s average return during the three months after signals since 1970 was negative 2%, compared with an average gain of 2.2% across rolling three-month periods. Over one year, the average post-signal return was 7.4%, below the broader rolling average of 9.1%, according to the analysis reported by MarketWatch.

Clusters have produced more cautionary historical results than isolated readings. Jason Goepfert, founder of SentimenTrader, told MarketWatch that he considers 11 or more Omens within a three-month period more informative than a single occurrence. His analysis found seven such clusters since 1965, with an average three-month return of negative 3.5% and an average one-year return of 2.6%.

That sample is small, however, and the current episode has not followed the historical average so far. MarketWatch said the latest cluster of at least 11 signals was completed on June 29. The S&P 500 closed that day at 7,440.43 and ended Friday at 7,785.76, a gain of about 4.6% over that span. A cluster can therefore identify a period of unusual market dispersion without producing an immediate decline.

The indicator also has a structural limitation: it is intended as a warning about market conditions, not as a timing tool that predicts the exact day a correction begins. A market can remain internally uneven for weeks while major indexes continue to rise. That is one reason technicians generally treat the signal as a prompt to examine other breadth, momentum and volatility measures rather than as an automatic instruction to sell stocks.

Friday’s data show why the warning is not a crash call

By the end of the week, the breadth picture had improved from Monday. McClellan Financial’s final NYSE data for Friday showed 1,444 advancing issues against 1,291 decliners, and its McClellan Oscillator stood at positive 39.31. Under Miekka’s formulation described by McClellan, an Omen’s 30-trading-day warning window is activated when the oscillator is negative and deactivated when it turns positive. That means the oscillator component was no longer in a bearish state at Friday’s close, even though the earlier warning remains relevant as evidence of recent market dispersion.

The broader tape was also mixed rather than uniformly weak. The Dow Jones Industrial Average slipped 0.20% Friday, the Nasdaq Composite fell 0.28% and the S&P 500 lost 0.17%, while the Russell 2000 gained 0.51%. The NYSE Composite edged higher. That is not the kind of synchronized deterioration that would by itself confirm a major market break.

Investors still have reasons to watch breadth closely. Strong headline indexes can conceal sharp differences among sectors and individual stocks, and repeated appearances of new highs and new lows at the same time can signal that leadership is becoming less cohesive. The Hindenburg Omen is designed to make that split visible before a broad decline is obvious in the major averages.

For now, the most defensible reading is narrower than the indicator’s ominous name. Monday’s data met the traditional warning criteria, and the recent cluster has historically been associated with weaker forward returns. But the S&P 500 has continued to advance since the June cluster, Friday’s McClellan Oscillator was positive, and the signal itself varies with the breadth dataset used. The Omen is flashing a caution about market internals, not establishing that a stock-market crash is imminent.