Board-Ansicht
Mix-Ansicht- Mark Hulbert: Hindenburg Omen - Amanito, 05.10.2005, 10:40
Mark Hulbert: Hindenburg Omen
-->Watch out for the Hindenburg Omen
Should we pay any attention to this indicator?
"Yes" is the answer from quite a few of the investment newsletters I monitor. Indeed, in recent days so many advisers have referred to the warnings that the indicator is emitting that investing blogs are all abuzz.
And, naturally, more than a few of you emailed me to ask that I devote a column to it.
Let me start by reviewing the Hindenburg Omen. The core idea behind it is that it's bearish whenever there are a large number of both new 52-weeks highs and new 52-week lows on the New York Stock Exchange.
From what I can tell, it was created in the 1970s by a fellow named Jim Miekka, who was editor of a newsletter called the Sudbury Report. Credit for christening this indicator the"Hindenburg Omen" goes to Kennedy Gammage, who used to edit a newsletter called the Richland Report.
Why would a large number of both new highs and new lows be bearish? Peter Eliades, editor of the Stockmarket Cycles newsletter, recently provided an answer:"Under normal conditions, either a substantial number of stocks establish new annual highs or a large number set new lows - but not both." When there are high levels of both,"it indicates that the market is undergoing a period of extreme divergence... Such divergence is not usually conducive to future rising stock prices. A healthy market requires some semblance of internal uniformity, and it doesn't matter what direction that uniformity takes. Many new highs and very few lows is obviously bullish, but so is a great many new lows accompanied by few or no new highs. This is the condition that leads to important market bottoms."
In terms of genealogy, the Hindenburg Omen is a descendant of an indicator called the High Low Logic Index, which Norman Fosback, editor of Fosback's Fund Forecaster, devised in the early 1970s. According to Gammage, the Hindenburg Omen is also"derived from a New High - New Low indicator developed by Gerald Appel many years ago." Appel, of course, is the editor of the Systems & Forecasts newsletter.
Both Fosback and Appel have good track records, so I tried to contact both of them Tuesday to see what reactions they might have and whether they believe that the market will soon crash.
Fosback had not responded by the time I filed this column. But Appel told me that while the simultaneous appearance of many new highs and new lows is not a positive development for the market, one should not exaggerate its bearishness. The market's prospects following such a development are"sub-par," but not"calamitous."
Appel also pointed out that there have been some noteworthy cases in which a large number of both new highs and new lows have not presaged a market decline. He mentioned in this regard the situation prevailing prior to the bull market taking off in 1982.
To be sure, Miekka, Gammage and other devotees of the Hindenburg Omen have always insisted that the new high/new low data must be interpreted in the light of several other indicators, and that in conjunction with them the false signals are kept to a minimum.
According to Robert McHugh, editor of McHugh's Financial Forecast & Analysis, the two other indicators that have traditionally been looked at when interpreting the new high/new low data are the NYSE 10 Week Moving Average (it must be rising), and the McClellan Oscillator (it must be negative). (The McClellan Oscillator is a measure of market breadth.)
However, even in conjunction with these other preconditions, according to McHugh's calculations, a simultaneous large number of new highs and new lows has not always presaged a market decline. McHugh has proposed adding yet two more preconditions:
"New 52 Week NYSE Highs cannot be more than twice New 52 Week Lows; however it is okay for New 52 Week Lows to be more than double New 52 Week Highs."
"There must be more than one signal within a 36 day period, i.e., there must be a cluster of Hindenburg Omens (defined as two or more)"
According to McHugh, the market historically has declined 87% of the time once all five of these preconditions have been met. And the market has actually crashed 27% of the time.
All of these preconditions were met in September, which is why so many advisers are worried.
The proliferation of these preconditions makes it very difficult for a statistician to offer much insight, for several reasons. First, with each new precondition, the number of historical precedents shrinks. McHugh, for example, counts just 22 cases since 1985 in which all of his preconditions were met. Unfortunately, with a sample size this small, it's impossible to draw firm statistical conclusions.
The second problem with piling precondition on top of precondition is that it runs the risk being ad hoc and arbitrary. After the fact, of course, with the benefit of 20-20 hindsight, it's always possible to pinpoint one or more factors that, if they had been taken into account, would have led to a better prediction. But rarely does such an approach lead to a successful market timing system.
The proof of the pudding is in the eating, of course. But it is a statistical sleight of hand to back-test a system's performance over the same period that was used to jerry rig it in the first place. The only real proof of the pudding comes when testing a system over a different period than the one used to devise it.
This is why statisticians usually insist on real-time tests of any system that is jerry-rigged to fit past data. And given how rare it is for all five preconditions associated with the Hindenburg Omen to occur, it will take years for us to know whether these five preconditions, taken together, truly represent an increased probability of a market crash.
So the best a statistician can say is that we should wait and see. This response no doubt is profoundly unsatisfying, given the prospect of an imminent crash.
But think of it this way. There are other newsletters among the 180 or so that I monitor that currently are as bullish as the devotees of the Hindenburg Omen are bearish. And they have devised market timing systems that have just as much superficial plausibility as the Hindenburg Omen, but which currently are quite bullish.
<ul> ~ http://www.marketwatch.com/news/story.asp?guid={DA451070-3B84-4BCE-ADAD-11B3D26F</ul>
Thread gesperrtgesamter Thread:
