Over the past week or two, we've seen several traditional and a few esoteric indicators suggest that the market was not as strong as the recent record high in the Dow Jones Industrial Average might suggest. Even with the volatility and net loss over the past few days, the Dow is still within 3% of its high watermark.
 
But the chart pattern the decline has left in its wake was the first indication that something is wrong (see Chart 1). The Dow's mid-September breakout lasted but a few days before the index fell back into its previous pattern. Chart watchers have several names for such head fakes, including the ominous "breakout failure."

Chart 1

Dow Jones Industrial Average

The breakout happened when demand surged on Sept. 18. The failure happened two days later when supply surged to meet and exceed that new demand. Prices sagged, and as breakout failures often do, they led to a return, if not a breakthrough, to the bottom of the pattern from which they emerged.
 
The financial sector made the same move and had the same failure. So, too, did two other key sectors—semiconductors and retail.
 
These are the obvious changes on the charts. Less obvious was a 90% down day, Sept. 25, in which the lion's share of the market's volume and price movement was to the downside. Consider it the first shot across the bow as investors start to sell whatever is not bolted down.
 
Also in the financial news, albeit temporarily, was the moving average death cross as the small-capitalization Russell 2000's 50-day average crossed below its 200-day average. Critics were quick to point out its lack of statistical credibility as a sell signal, and in the postfinancial crisis era they were right. But indicators do not last over the years if they have never had any validity. The short-term trend via its 50-day average proxy is now below the long-term trend, and that is not a good thing.
 
But the Russell's chart has a more sinister feature than just a moving average cross, and it is probably the most important. This index has gone nowhere all year, creating a wide chasm between the performance of big stocks and small stocks. This divergence is not healthy, and tells us that fewer stocks have led the market higher. The same divergence occurred in 2007, and arguably in 2000, at the tops of previous cyclical bull markets.
 
This split between rising and falling stocks is the basis for another strange indicator that fired its warning signal last week. The Hindenburg Omen, named for the ill-fated dirigible that exploded over New Jersey in 1937, looks for high numbers of stocks trading at 52-week highs and 52-week lows at the same time. This split in the market makes it unstable and often leads to declines in the following weeks.
 
Market crashes do not necessarily follow, although they can. The Omen is only meaningful when it fires multiple times in a short time period. That has not happened, but its single occurrence is one of the thousand cuts that should make investors nervous.
 
On the heels of last week's Alibaba Group initial public offering, some market observers noted that record-setting IPO volume often coincides with major market tops. I understand the logic—that companies rush out to raise as much money as possible during a frenzied market advance. The data, however, do not support the idea. It is true that in 2000 and 2007 there were record numbers of IPOs coming to market, but one happened months before the top and the other happened months after. And throughout the 1990s, record-setting IPO numbers did not coincide with tops at all.
 
Also ripped from the headlines, one analyst plotted breakdowns in Bill Gross' former Pimco High Yield Fund as coinciding with major stock tops. Here, the logic that junk bonds break down ahead of stock tops is valid. While U.S. Treasury bonds remain in rising trends, junk bonds have indeed cracked in a "risk off" trade. Money is moving to safety, underscoring the market's mood.
 
Together, all of these factors put the market in a precarious position. And Wednesday, perhaps the most important factor of all was close to signaling a breakdown. The Standard & Poor's 500 traded marginally below its rising trendline from the key November 2012 low (see Chart 2). The New York Stock Exchange composite, which arguably represents the average stock, has already broken its trendline.

Chart 2

Standard & Poor's 500

How much can the market withstand before lapsing into a major decline? That line may have already been crossed.
 
Michael Kahn, a longtime columnist for Barrons.com, comments on technical analysis at www.twitter.com/mnkahn. A former Chief Technical Analyst for BridgeNews and former director for the Market Technicians Association, Kahn has written three books about technical analysis.