Thin Slices From The Top Of A Bubble – John Hussman 

In the shortest possible form, here are the Iron Laws:

To fully encourage your sense of deja-vu, the following is clipped from my October 3, 2000 commentary in Hussman Investment Research & Insight, a few weeks after our measures of market action turned negative. It details a lesson that I admittedly had to learn twice in my investment career as a result of that stress-testing stumble. I strongly encourage investors to learn it once, learn it permanently, and teach it to your children.

Lest we forget the setup to the worst financial collapse since the Great Depression, it will also be useful to revisit similar considerations I emphasized at the 2007 market peak. What’s fascinating to me is how unwilling investors will be to recognize any of this until the third speculative bubble in 15 years collapses again over the completion of this cycle.

When Hindenburg Omens are Ominous

I’ve frequently noted that Hindenburg “Omens” in their commonly presented form (NYSE new highs and new lows both greater than 2.5% of issues traded) appear so frequently that they have very little practical use, especially when they occur as single instances. While a large number of simultaneous new highs and new lows is indicative of some amount of internal dispersion across individual stocks, this situation often occurs in markets that have been somewhat range-bound.

Still, when we think of market “internals,” the number of new highs and new lows can contribute useful information. To expand on the vocabulary we use to talk about internals, “leadership” typically refers to the number of stocks achieving new highs and new lows; “breadth” typically refers to the number of stocks advancing versus declining in a given day or week; and “participation” typically refers to the percentage of stocks that are advancing or declining in tandem with the major indices.

The original basis for the Hindenburg signal traces back to the “high-low logic index” that Norm Fosback created in the 1970’s. Jim Miekka introduced the Hindenburg as a daily rather than weekly measure, Kennedy Gammage gave it the ominous name, and Peter Eliades later added several criteria to reduce the noise of one-off signals, requiring additional confirmation that amounts to a requirement that more than one signal must emerge in the context of an advancing market with weakening breadth.

Those refinements substantially increase the usefulness of Hindenburg Omens, but they still emerge too frequently to identify decisive breakdowns in market internals. However, one could reasonably infer a very unfavorable signal about market internals if leadership, breadth, and participation were all uniformly negative at a point where the major indices were still holding up. Indeed, that’s exactly the situation in which a Hindenburg Omen becomes ominous. The chart below identifies the small handful of instances in the past two decades when this has been true.

Full article here