Have the horses left the barn? One of the major tenants of Dow Theory is that the averages must confirm each other. The key to this basic tenant is that no important bull or bear market can take place unless both averages agree with each other. Originally, Dow Theory was referring to the Industrial and Rail averages. Today the theory refers to the Industrials and the Transports. It is not necessary for the two to confirm a signal simultaneously but they should be at least relatively close together.

So have the Industrials and the Transports confirmed each other? Of late no. The Dow Jones Transports (DJT) made its high back on November 28, 2014. That high has not been seen since. For the next five months the DJT chopped around while the Dow Jones Industrials (DJI) made new highs in late December, again in March and most recently on May 19, 2015. By that time the DJT was roughly 500 points below its late November high. Three new highs by the DJI, lower highs or just considerably lower by the DJT. Three strikes and one is supposed to be out. Maybe it is.

The very next day (May 20, 2015) the markets plunged. The DJT was down sharply on high volume. The DJI? Well it barely moved. Another divergence. The DJT has given a number of intermediate sell signals and on May 26, 2015, the index was at its lowest level since October 2014. The DJI Fell over 200 points on May 26, 2015 but is now only testing its 50-day MA.

The chart of the DJI and the DJT shows this growing divergence between the two Dow indices. The DJI like the S&P 500 appears to be forming a large ascending wedge triangle. A breakdown under 17,900 could suggest a decline back to the October 2014 lows. As for the DJT well it may be trying to form a descending wedge triangle (the opposite of an ascending wedge triangle). But it is inconclusive at this stage.

What is important, however, is that these divergences between the DJI and the DJT do not occur with regularity. Not surprisingly these divergences usually only occur at significant tops and bottoms. A review of some of them follows.

Source: www.stockcharts.com

March 2009 - Both the DJI and the DJT made new lows. It is not often that both the DJI and the DJT confirm each other at what turned out to be a significant low.

July to October 2007 - The DJT made its high in July 2007. When the DJI made its final high in October 2007 the DJT was already sharply lower.

October 2002 to March 2003 - Both the DJI and the DJT made new lows together in October 2002. In March 2003 the DJT fell to new lows while the DJI made a higher low. That low turned out to be the final low before the bull market that topped in 2007 got underway.

May 1999 to January 2000 - The DJT made a high in May 1999. Over the next several months the DJT was trending lower. The DJI also made a high in May 1999 then continued to new highs in July and August 1999. Following a period where the markets were generally down the DJI moved to another new high in January 2000. That turned out to be the final high for the DJI as the markets collapsed into the 2000-2002 high tech/internet collapse. The DJI made a lower high in March 2000 while the S&P 500 and the NASDAQ were both making new highs.

April 1998 to July 1998 - The DJT made a high in April 1998. The DJI made a higher high in July 1998 while the DJT was making a lower high.

November/December 1994 - Both the DJI and the DJT made important lows in late November 1994. In December 1994, the DJT plunged to new lows while the DJI made a higher low. That low proved to be an important low prior to the big bull market of 1995-2000.

1990 - The DJI and the DJT both made highs in June 1990 but one month later the DJI made new highs while the DJT was not even close. That proved to be the top prior to a sharp drop in 1990. The DJI made its final low in October 1990. Over the next two months, the DJI started to move higher but the DJT kept testing the October low. Nonetheless, it proved to the final low before the 1990's bull market got underway.

The current divergence between the DJI and the DJT is one of the most pronounced seen over the past fifteen years. These type of divergences normally only take place at important tops and bottoms. The advance that has been seen over the past year has been constructed mostly by the companies in the indices buying back their own stock. In many instances, the purchases have been financed by issuing new debt. Goldman Sachs had noted in late 2014 that buybacks constituted one of the largest sources of US equity demand over the past few years.

Buying back one's own stock is a non-constructive activity and while it helps push up the value of the stock it has little if any economic value. Rising stock values can translate into increased compensation for management of the companies. It is estimated by some that buybacks could hit as high as $450 billion in 2015. Previous highs in share buybacks were seen leading into both the 2000 market high and the 2007 market high.

Over the past year numerous indicators have been deteriorating. Sentiment indicators remain at or near record highs. An interesting indicator is the McClellan Oscillator. The McClellan Oscillator is an oscillator that takes the differential between two exponential moving averages of daily market breadth. A rising oscillator is an indication of money flowing in while a falling oscillator is more indicative of money flowing out. Extreme readings are an indication of overbought or oversold conditions. Values below zero are an indication that money is leaving the market. The McClellan Oscillator turned negative on May 22, 2015 even as the market continued higher in price.

Source: www.stockcharts.com

Warning signs have been creeping into the market for months even as it moves higher. Continued divergences between the indices are a sign that something is brewing. In the past these divergences signalled market collapses of between 20% to 55%. This is a brew that may not be palatable to many investors.