Tracking the Market's Trend
Analyzing the stock market's internal health can play in important role in determining your level of success in the stock market, says technician Tom Aspray and he has one tool that can be used to assess the market's strength.
The advance/decline data is one measure of the market’s internal health that I have found to be the most reliable in determining the stock market’s intermediate-term trend. It has done a good job of keeping one on the right side of the market since March 2009.
The data on the number of stocks making new 52-week highs and new lows is another market internal that I follow regularly. The NYSE high/low data series was one that I started entering into my Apple II before the 1982 market low.
It was also a market indicator that I featured in my early writing for Stocks and Commodities. I did more work with this data two years ago and—given the current state of the market—wanted to readdress it again and discuss how it can be used to provide additional insight on the health of the stock market.
In a bear market, the number of stocks making new lows can give you a good idea of how heavy the selling is, and therefore, the extent of the market’s weakness. As long as the number of stocks making new lows keeps expanding, it indicates that the sellers are still in charge.
This chart covers the period from March 2002 through 2003. Stocks had been declining since the first quarter of 2000 when the Spyder Trust (SPY) made a high of $155.75 and the Nasdaq Composite hit 5132.
The first wave of selling culminated on September 21, 2001, when 762 stocks made new lows. The market then started a significant bear market rally that retraced 38.2% of the previous decline in the S&P 500. By May the chart shows that the decline had clearly resumed and the number of stocks making new lows had begun to expand.
As prices plunged to a low in July 2002 at $77.68 (point 1), there were 912 stocks that made new 52-week lows. During the following four-month rally, the SPY rebounded 25% from its lows before stalling below the 50% Fibonacci retracement resistance.
On October 10, SPY made a new low at $77.07 (point 2), but only 598 stocks made new lows. This positive divergence, line b, was a sign that fewer sellers had pushed the market lower.
When SPY rebounded from these lows, it failed to surpass the prior highs, which suggested that a more complex bottom was forming. In March 2003, SPY dropped to a low of $79.38 (point 3) but only 319 stocks made new lows. This was another sign that the market had bottomed and the A/D line completed its bottom in January 2003.
Now let’s fast forward to 2008-2009. By June of 2008 it was clear that the bear market rally was over, and on July 15, the number of new lows hit a high of 1144 (see circle).
By September, the sellers were clearly back in charge as Lehman Brothers had filed for bankruptcy, and by October, the panic in the financial markets had reached a fever pitch. On October 10, with the SPY making a low of $83.58 (point 1), there were 2476 stocks making new lows.
NEXT PAGE: Identifying Market Tops and Bottoms|pagebreak|
After a brief rally into the Presidential election the stock market again turned lower making a new low at $74.34 on November 20. On this new price low, there were only 1404 stocks making new 52-week lows (point 2). This was the first positive divergence that had occurred in the bear market.
The number of stocks making new lows dropped to a low level in December and January but started to increase as the market started to decline again in February. In late February the number of new lows spiked to 553, but as I noted in my column New Lows Decreasing, this was an encouraging sign.
As the SPY made further new lows at $67.10 on March 6, the number of new lows rose to 805 (point 3), which was still well below the previous two major peaks. This suggested that the market might be sold out.
As these examples from the new low data illustrates, this is not a trading tool but another piece of evidence that can used in conjunction with other indicators like the A/D line to help identify a market bottom.
In bull markets, I have always kept an eye on the number of stocks making new highs and what happened in 2007 and 2008 is typical of what I have observed in the past. In June 2006, as the SPY was correcting to the $122.50 level, the number of stocks making new highs had dropped to under 50.
The daily chart shows that by the end of July 2006, a double bottom (see circle) had been completed and the number of new highs began to expand once more. By early fall a solid uptrend was underway and the number of new highs continued to expand.
On December 5, the new highs surged to 568 (point 1), which was the highest number since December 2003. For the next month, SPY consolidated, dropping about 2% from its highs.
As SPY resumed its uptrend and moved well above its prior highs, the number of stocks making new highs was declining. When the SPY was rallying to new highs (point 2), there were only 452 stocks making new 52-week highs.
This was a sign that the market had lost upside momentum and the SPY dropped from a high of $146.20 down to a low in early March of $138.04, which was a 5.6% correction.
By the latter part of March, the uptrend had resumed and the number of stocks making new highs had also started to increase. By early April, the number of new highs was back to 430 but this was still below the early February high of 452. On June 1, the SPY made a new high of $154.40 (point 3) but just 506 stocks made new highs. The divergence from the peak in late 2006 is noted by line a.
The SPY corrected for most of June before again turning higher as SPY peaked at $156 on July 17. Three days before the SPY made its high, the new highs were only able to reach 386 (point 4), which was well below the June reading of 506. This was similar to the divergence that was observed in February, but this time the selling was much heavier.
NEXT PAGE: Signs Foretelling Market Turns|pagebreak|
By August 16, the SPY had dropped over 12% and was back to the $137 level. The number of new highs had dropped to just 7. By the middle of September, the fears of a financial crisis had subsided and stocks began to rally more sharply. By October 10, the SPY made a new high at $157.52, line A. This new price high was accompanied by only 345 stocks making new highs (point 5).
This divergence from the early June highs, line b, was the second significant divergence that had formed since the late 2006 high. This has indicated that fewer and fewer stocks were pushing the stock market to new highs and was a sign of a technically weak market. At the time, similar long-term divergences were evident in the NYSE Advance/Decline line.
So what about the market just two years ago? As of April 2013, though not included on the chart, the greatest number of new highs since the 2009 low was 693, which was recorded on April 26, 2010. This was prior to a 17% decline that ended on July 1, 2010.
From these lows, the new highs expanded to 569 on November 4, 2010 (line a). The NYSE Composite bottomed at the end of November and continued to rally until April of 2011 as it gained almost 18%. The number of new highs declined as the market moved higher. As the NYSE was making a new high at 8718, only 361 stocks were making new highs (line b) as they were diverging sharply, line b.
The following five-month correction dropped the NYSE Composite 26.4% to a low of 6414 on October 10, 2011. By September, many were convinced that we were in another bear market. This was contrary to the positive intermediate-term signals from the NYSE Advance/Decline line and the high level of negative sentiment, Can Doom and Gloom Save the Market?, that suggested the market was closer to a bottom.
By the middle of October, there were clear signs that the market had bottomed and once again the number of new highs began to expand into the end of the year. The new highs peaked at 304 on February 1 and then started to decline even though the NYSE Composite continued to move higher.
The new highs made a lower high (line c) at 208 on May 1, indicating that fewer stocks were moving the market higher. But by early June, there were signs from the Advance/Decline data that the decline was over. By the end of July, the number of new highs had surged back to 296. They continued to expand into the September highs, line d, peaking at 495. On January 2, 2013, there were 456 new highs but just 446 on Tuesday, January 22.
So, is there a better way to utilize the new high, new low data? Some analysts just follow the differential of the new highs minus the new lows. This is plotted in the middle of the figure above.
Though its divergences occasionally give some good signals, the pattern is too irregular most of the time for me to find it beneficial. One technique that I have been looking at is to run a five-period WMA (5WMA) and a 13-period SMA (13SMA) of this difference.
NEXT PAGE: A Review of the Last 4 Years|pagebreak|
Let’s take a look at how it has worked over the past four years to see whether it might be a useful addition to the market timing tool box. Though the crossings of the two MAs should not be ignored, I also recommend that trend line and divergence analysis also be used as they are for the A/D line.
The week ending July 24, 2010, (line 1) the 5WMA moved above the 13SWMA after the 5WMA had formed a slight positive divergence earlier in the month. The 5WMA made higher highs in early November and then on November 20, (line 2) it dropped below the 13SMA.
The 5WMA stayed below the 13SMA until early February even though the SPY continued to move higher. But in February, the 5WMA was able to then exceed the November peak.
The decline into the middle of March was fueled in part by the Japanese nuclear disaster. By early May, the SPY was again making new highs but the 5WMA had formed a lower peak and therefore a negative divergence, line a.
By the end of May, the moving averages had also turned and on June 11, (line 3), the 5WMA dropped below its support at line b. This confirmed the bearish divergence. The 5WMA rallied back to the declining 13SWMA in July, but did not move above it before the wave of debt ceiling selling pushed the market sharply lower.
As SPY was making lower lows in early October, the 5WMA was forming a bullish divergence, line c, as it only barely dropped below the 13SMA. On October 22, the positive divergence was confirmed (line 4) as the resistance at line d, was overcome.
The MAs stayed positive and no divergences developed until March 17, when the 5WMA dropped below the 13SMA (line 5). SPY closed that week at $140.30 and managed to edge higher for the next three weeks before it also turned lower.
The MAs stayed negative until June 30, (line 6) when the rising 5WMA finally moved above the 13 SMA. On October 20, this signal was reversed (line 7) and the SPY dropped for the next four weeks.
On December 29, the 5WMA crossed back above the 13WMA just in time to catch the sharp rally in the first week of the New Year. By April of 2013, the 5WMA had risen strongly and was well above the 13SMA but had not exceeded the September highs.
In my opinion, the moving average analysis of the high-low differential is an improvement over the raw data but the signals are not as clear as those obtained from the Advance/Decline line. I left out the trend lines in Figure 5 as the crossing on the MAs coincided nicely with the trend line breaks.
I think the new highs and new lows are something that traders and investors should watch and you can find charts of the data on many free Web sites. I will start to include it more often in my market commentary and I will continue to explore new ways that this data can be used.