A look at Stock market Breadth.

The market, true to form, did what it always does--fool the most people, including me.

We have rallied over 10% in a week. I have been somewhat skeptical of the rally, and at the start of the month, began getting into long term put positions. Needless to say, the move has caused me to look closely at my investment thesis that this is a bear market rally.



As you can see from the chart, market breadth is a very reliable indicator at major turns, while the broad market averages can sometimes be misleading. At the top of the market in 2007, market breadth indicators screamed sell 5 months in advance, while the cheerleaders on CNBC were proclaiming even higher highs. At the 2000 peak, market breadth diverged over 2 years before the top. Similar things have occurred at the peak in 1929, as well as the bear markets in the 70's.

Market breadth, particularly the diffusion indicators, which I will show, indicates if the capitalization weighted averages are being pulled up by just a small number of stocks, or if the rally in the indexes is supported by a rise in most individual issues. It is all to easy to "paint the tape" by driving up a few large cap stocks, in order to induce the average speculator that all is well.

The chart has 2 components--the top window is Worden Brothers indicator T2107--percentage of NYSE stocks above the 200 day moving average. The middle window is something of my own design. Worden Brothers also has other diffusion indicators--T2111--percent of stocks 2 deviations above the 200 day moving average, and T2115--percent of stocks 2 deviations below the 200 day moving average.




What I do is plot a relative strength line (smoothed by moving averages) comparing Overbought stocks (green) and oversold stocks (red line) to the percentage of stocks in an uptrend. Then you can easily see how many stocks are in uptrends, downtrends, overbought, and oversold, and how individual issues are moving from the categories.

A strong market should see more and more stocks make it into overbought territory, causing the green line to trend down below the yellow average. Likewise, we should see the red line trend up. The signals should be fairly consistent. Likewise, in a weak market, we should see the red line trend down, while the green line trend up.

When the green line is "too far" away from the yellow average, those are areas to look to buy. When the the red line is "too far" away from its average (purple), those are times to be looking for tops.

When I looked at this today, I was pretty shocked--either we are at the start of a new bull market (hard to believe), or we are setting up for a nasty sell off.

Generally, these signals correlate well, and do not diverge for very long--both are long and short at the same time.

What is strange--at the start of the uptrend in March, they contradicted each other. So few stocks were overbought at the March low, that any uptrend was likely to cause T2107--sustainable uptrends, to outperform, giving a false sell signal. The RS line vs deep oversold stocks was more accurate.

Now, we are going to have another contradiction--the Green line is trending down, giving a buy signal, while the red line is also trending down, and probably going to give a sell signal soon. Almost 85% of NYSE stocks are in uptrends, and 38% of those are in overbought territory. Likewise, the number of deeply oversold stocks is under 1%. Do you want to jump on board an uptrend when a third of all stocks are overbought, and virtually none are oversold?