Hindenburg Omen (Occurred Twice this Month)

I would tend to agree, although the Hindenburg omen might easily be explained by publicity like this submission on HN. Investors read about the Hindenburg omen and the fact that it has occurred twice this month. They assume that a crash is imminent, so they start selling stocks, which causes the market to collapse.

In this way what is drivel could be made reality because too many people don't believe it to be drivel. There are some really interesting analysis of stock market dynamics such as this in the book "The Wisdom of Crowds" by James Surowiecki

Doesn't it say in the article that in the past, 24% of the Hindenburg Omen were followed by a stock market crash? From the article:

Looking back at historical data, the probability of a move greater than 5% to the downside after a confirmed Hindenburg Omen was 77%, and usually takes place within the next forty-days. The probability of a panic sellout was 41% and the probability of a major stock market crash was 24%.

While the past performance is quite compelling, the criteria seem to be extremely arbitrary.

There's a word for that:

http://en.wikipedia.org/wiki/Paredolia

No, it is generally complete BS. It survives on human gullibility, and the observation by stock brokers that promoting technical analysis is good for trading volume. A good book that addresses this A Random Walk Down Wall St.

That said the particular "omen" they are talking about does have some merit. Normally there is a broad correlation between different stocks, so you don't typically get both lots of highs and lows at the same time. (This correlation is why the DOW, which just tracks a small number of indicator stocks, is correlated with much broader sets of stocks.) But you can get lots of highs and lows at the same time if the market is very volatile. Volatile markets sometimes go down very sharply. Of course causation goes the other way. Fears that the market could drop cause volatility, and sometimes those fears are proven right.

That said we can and do measure market volatility directly. (For instance every options pricing model does that.) And that gives a much more fine-tuned predictor of risk than this "omen".

In an ideal world, technical analysis tries to model psychological effects that are statistically significant in a population of humans. Don't buy just because you see a W, or crap like that. But if you understand why a W is an interesting pattern, and what it's based on, it can be a valuable signal to digest.

I don't know enough to trade purely based on technicals, and I'm not sure anybody should. But when I want to buy a stock I wait until it's oversold (http://clearstation.etrade.com/cgi-bin/details?Symbol=goog) That usually saves a few bucks.

It kind of works because everyone else thinks it works. Self fulfilling prophecy. It's like looking at traffic lights and thinking traffic should move the way they dictate it - it works because, almost, everyone participating in traffic follows the same signs.

Technical analysis has always seemed like nonsense to me, too. I've never once seen an example that seemed to have a firm logical or mathematical foundation.

Has an authoritative source ever debunked it?

It sounds like the beginning to one of those 80ies barbarian movies (I'm thinking of The Beastmaster, which is a must-see):

"On the 10th day, the Omen foretold the coming Prophecy that the Evil Emperor would be slain by a Child born under the Star of the Omen"