PDA

View Full Version : 9-20-2014 Comments



Mike
09-20-2014, 11:20 AM
I was asked via private message whether I am still following Didier Sornette's work. I am, the log-periodic power law confidence level is off the charts on US Tech stocks indicating bubble territory.

We had a Hindenburg Omen yesterday (Google the term if this is unfamiliar). This is a condition where we have high numbers of NYSE new highs and new lows on the same day, plus some other criteria. Normally these precede a topping market before a significant correction. We had numerous warnings prior to the 2007 top. They don't always lead to a bear market, the last series of Hindenburg Omen warnings came before the 10% NASDAQ correction in 2013. High numbers of new highs at the same time as high numbers of new lows points out a possible bifurcated market where a narrowing number of stocks are making new highs while a widening number of stocks are making new lows. Normally a Hindenburg Omen is confirmed by additional warnings.

The following chart supports what I am seeing. It shows that only 45% of stocks trading on the NASDAQ are above the 200-day moving average. With the index near new highs, having more than half of the stocks below the 200-day moving average is significant. I note that on each rally that more and more stocks are being left behind below the 200-day. A strong rally would show the percent of stocks above the 200-day continuously in the range of 50%-75%.

26034

Yesterday showed distribution on the NASDAQ bringing us to a total of five within a trailing 25-day window. This level of distribution is becoming significant. In periods prior to 2009 having five days of distribution would indicate a time to pull some money off the table. IBD put the rally back into confirmed rally (from market under pressure) because the S&P500 index closed at a new high on 9/18/2014 (two days ago). Yesterday's reversal closed below that level that turned us back to confirmed rally status. This implies that with further weakening that we could revert back.