Early-stage crash for S&P?
A Hindenburg Omen is where more than 2.8 per cent of stocks on the New York Stock Exchange are making new annual highs, with at least the same proportion making new annual lows. The number of new highs should not be more than double the lows, however. Meanwhile, the NYSE Composite index must also be above its level of 50 days ago. The signal occurs when the McClellan Oscillator is negative and the signal then lasts 30 days, unless that oscillator turns positive.
Negative McOscillator
If you think this sounds horribly convoluted, join the club. For me, the best indicators are usually straight-forward ones. But I'm willing to use something more involved if its record is good. So, how has the Hindenburg Omen performed over time? Eyeballing this chart (http://bit.ly/1934hZv) of its record, you'll see that clusters of Hindenburg signals occurred around the big tops of 2000 and 2007, as well as ahead of the 1987 and 1998 crashes.
I have two quibbles here. First, it is clear that the Omen throws up a lot of false signals. It has often cropped up in the middle of strong uptrends and with no subsequent ill-effects. Second, I can't really understand the theory behind it. A good signal should work for a logical reason. The criteria of the Hindenburg Omen make me think someone simply went through a lot of trouble to find something that worked on certain occasions.
Dow's new yearly highs, less lows
The Titanic Syndrome also looks at the number of NYSE stocks making new highs and new lows. If more stocks are making new annual lows than annual highs within seven days of the index reaching a major high, a preliminary alert is created. The creator apparently then looked for new lows to outstrip new highs on four out of five days, and for the market itself to decline on four out of five days.
Once again, the Titanic Syndrome does indeed appear to have flashed up ahead of some major peaks in the S&P, with especially strong clusters of signals before the two savage bear markets of the 21st century. However, this alert too tends to go off for no good reason, thereby risking getting you out of a perfectly good bull market, such as that of the early- and mid-1990s, or around the middle of the last decade.
I don't really understand why signals like this get so much publicity. Tom McClellan - whose site, www.mcoscillator.com, drew them to my attention - reckons their sensationalist names make them headline grabbers. As one who writes trading analysis in an investment publication, I know all too well there is often a conflict between lively copy and sound advice.
Erratic DAX
The two most successful pieces of advice I ever give are pretty dull: keep doing what I've been recommending for several weeks or to do as little as possible. Doggedly following long uptrends is how money is made, while staying out in choppy periods is how it is kept. Given the erratic action of the past few days, my advice at the moment is to do as little as possible. And to ignore anything named after doomed zeppelins and cruise liners.
