Using Market Internals to Time the Market

Since the March 2009 lows, I have frequently featured charts of the market internals as the Advance/Decline line, in particular, has been useful in staying on the right side of the trend. Whether you are trading the S&P futures, ETFs, or options that correlate with the US stock market, I think you will find that the A/D line can improve your market timing even if you are a very short-term trader. I follow the A/D line and up/down volume numbers throughout the market day and they are often very helpful in predicting how the market will close. There have been quite a few times when the market has been flat or up a bit going into the last hour, but the A/D numbers were positive by 2:1 or better. The result was often that the market surged into the close. Even for investors, staying with the trend of the A/D line should significantly improve your performance.

In analyzing the A/D line, you should approach it like any other indicator or market. By that I mean identify support and resistance levels through trend lines while also comparing it to a market average such as the NYSE Composite or S&P 500. Be sure to keep in mind that when the A/D line breaks above resistance, this level then becomes support, or if it drops below support, this level is the new resistance.

Figure 1 - Click to Enlarge

For the first chart, I went back to the middle part of the last bull market. For some background, stocks made their lows in early 2003, and then in May, the A/D line overcame more significant resistance. Throughout 2004, the NYSE A/D line confirmed each new market high. In March 2005, it made another higher high but then reversed to break its short-term uptrend (line a) as noted by vertical line 1. This resulted in about a 7% correction as the A/D line started to bottom out in April and May. This allowed one to clearly identify resistance in the A/D at line b, which was overcome in May (line 2). By the middle of June, the A/D line had moved above the March highs even though the NYSE Composite was significantly below its highs. This was a very positive sign. The A/D line stayed in a strong uptrend until early August when the short-term uptrend, line d, was broken (line 3). This gave an early warning as the A/D line then formed a short-term negative divergence at the next highs (see box), which set the stage for a two-week correction.

Both the A/D line and the NYSE made marginal new highs in September, but in early October, the A/D line was back to support that went back to early August, line 4. The A/D line was now acting weaker than prices and formed lower highs, line f. The 5% correction lasted until early November when the downtrend in the A/D line was broken and it then moved above resistance. This is pretty typical bull market behavior and the A/D line continued to confirm the major highs during 2006.

Figure 2 - Click to Enlarge

The same type of methodology can be applied to the A/D line during bear markets, and these examples should further reinforce that the length of any divergence can provide valuable clues as to how deep or long a correction should be expected. Figure 2 starts with the sharp drop in late February over fears about the Chinese market, and while the NYSE violated the previous lows, the A/D line was much stronger. About a week after the lows, line A, the A/D line was already making new highs as resistance, line b, was overcome. For the next few months, the market and A/D line were in sync as the June 4 highs were confirmed. The decline from these highs was quite sharp with the A/D line finding support at line c. The NYSE Composite pushed to new highs in July, point 2, but the A/D line did not confirm. Six days after the highs, the A/D’s intermediate-term uptrend, line a, was broken. The divergence was confirmed (line B) when the A/D line violated support at line c. The six-week divergence resulted in a fairly sharp decline as the March lows in both the A/D line and NYSE were tested.

The decline was so deep that it was very unlikely that the A/D line would be able to confirm new price highs if they occurred in the next few months. The A/D line completed a short-term bottom formation in early September and then started to trend higher, line d. The NYSE rallied sharply into the October 11 highs (point 3), but the A/D line was much weaker as it just rallied up to former support, line c, which was now resistance. The uptrend in the A/D, line d, was broken three days after the highs (line c) and the break of support at line e generated a stronger sell signal. The three-month divergence in the A/D line was consistent with a more significant top than the divergence that was formed in July.

Figure 3 - Click to Enlarge

The analysis of the NYSE high/low data is less straightforward than that of the A/D line, and I use it more to confirm the analysis of the A/D data and do not trade based solely on its signals. For example, in 2006, the A/D line was confirming the higher prices and the new highs also show a strong uptrend, line a. The new highs made a sharp new high at 568 in early December, but even though the NYSE was higher in early February, line b, the new highs were only 452, point 2. The new highs came back nicely in April as the trend of lower highs was broken. At the June highs, the NYSE was 6.4% above the February peak, while the new highs just reached 489 (point 3). In July, as the A/D line was forming a negative divergence, the number of new highs peaked at 389and the number of new highs dropped sharply going into the August lows. By September the number of new highs began to increase once more, but as the major averages were making new price highs on October 10, there were only 202 stocks making new highs. The new highs improved a bit to 272 as the NYSE made a new closing high on October 31 but this was a significant divergence when compared to 489 in June. Of course, the number of new lows can be used in a similar manner to confirm new lows in a downtrend. For example, the number of new lows peaked in October 2008 at 2476, but then diverged in November when there were 1404. At the March 2009 lows, the number of new lows was just 805. This was a sign of internal strength as we noted in this February 27 chart study.

Earlier this week, I wrote a Tips for Traders column in which I gave a current analysis of the market internals. That analysis was written as of the close on November 13, and I wanted to update the analysis (but not the charts) through the close on Tuesday, November 17.

Figure 4 - Click to Enlarge

Before looking at the current readings, one should know that since the March lows, the A/D line has acted stronger than prices and it surpassed first key resistance on March 17. This started a pattern of higher highs and higher lows. The chart above (updated through November 13) shows that the A/D line confirmed the price highs in October before both turned lower. The decline in the A/D line was sharper than expected as it retested the previous lows and started to act weaker than prices. The A/D line has not yet confirmed the new closing high for the major averages on Monday, November 16. A failure of the A/D line to confirm prices over the next week or so would suggest a short-term top had been completed. This deterioration needs to be confirmed by a significant shift in downside momentum to signal a deeper correction. A break in the NYSE A/D line below the key support level would be negative. The A/D line on the NASDAQ is even weaker (see this recent chart study) as it is not far above its key support, and well below the October highs.

Figure 5 - Click to Enlarge

In October, the number of NYSE stocks making new highs was 433, well above the September reading of 340, therefore confirming the price action. However, on Monday’s new high, the number of stocks making new highs was just 307. This is another indication that the rally in the stock market is not as strong as it was in October. To reassert the uptrend, we need to see the number of NYSE new highs exceed 433.

I would like to emphasize again that the key thing to remember is that a divergence in the market internals, either bullish or bearish, needs a significant change in momentum to generate a buy or sell signal. The formation of the divergences should be viewed as warning that the current trend is weakening, which makes a change in market direction more likely.

I hope these examples clearly illustrate that divergences that form over a period of weeks are consistent with a correction in the intermediate-term uptrend, not a change in the intermediate term. Those divergences that develop over several months are warning of a change in the intermediate trend. Most stock market professionals do follow this data as they have found that monitoring of the market internals is important. I think that this type of analysis on the market internals has stood the test of time and hope you will start using it in your trading.