Combining Candlestick Charts and the Demand Index

The objective of this series of articles is to show you several different trading methods so that the individual trader can select the ones that “feel right” to them, and build their own unique trading method. One question I have often received when presenting the Demand Index (DI) is whether candlestick formations can reinforce the signals from the DI. The question is very astute as they definitely can. I have found that the more information or data that you have to support your trading decisions, the higher your probability of success will be, and the higher your level of confidence as a trader will be, as well.

I was first exposed to candle charts in Tokyo in 1989 (just before the Nikkei’s top) when the sponsors of my presentation gave me The Japanese Chart of Charts, by Seiki Shimizu, one of the early books on the subject. I should make it clear that I am not an expert on candle charts, though I have found them to be a useful tool when combined with other indicators, such as the DI. In the past few years, there have been several excellent books that combine candlesticks and other technical analysis methods. One of my favorite books is written by an old friend, Greg Morris, and is called Candlestick Charting Explained: Timeless Techniques for Trading Stocks and Futures.

Figure 1 - Click to Enlarge

Candlesticks show the period’s price with emphasis on the opening and closing price relative to the high and low price. If the closing price is above the opening price, the box is white, and when the close is below the opening price, it is red. The lengths of line outside the boxes, also known as shadows, indicate the highest and lowest prices throughout the day. These are also important, because a small box with a large shadow can tell you whether buyers or sellers are dominating the market.

Figure 2 - Click to Enlarge

In many years of looking at the Demand Index (DI), I have noticed that trend-change signals often correspond with reversal formations on the price charts that are more easily identified with candle charts. The weekly chart of Amgen (AMGN) covers the period from June 2002 through November of 2003. The DI made its low in June (point 1) and then five weeks later, even though the stock was over 10% lower, the DI was higher, forming a positive divergence (point 2). This was the basis for the uptrend in the DI (line A), which stayed intact until May of 2003 as the stock rose from a low of $30.57 to over $65 per share.

As you may recall from this previous lesson, the breaking of an up- or downtrend in the DI is usually the first warning sign of a change in trend if no weekly divergences are evident. The week of July 18, AMGN hit a high of $72.37, but then closed at $70.77. The long shadow relative to the box is an indication that as the market made further new highs, the sellers took over (see first circle). The DI had also formed a negative divergence, point 4, which indicated that even though AMGN had made a new high in price, volume and buying pressure were lower than they were in July. After declining for six weeks, AMGN once again turned higher and moved briefly back above the $70 level. But as the candle formation in May 2005 indicates (see circle), the sellers dominated once again. You should also note that at this point, the DI was significantly lower (point 5). The ensuing correction lasted until October of 2004.

NEXT: Analysis of Oil Futures and AAPL

Figure 3 - Click to Enlarge

In the crude oil market, analysis of the volume is especially important, and I have found the Demand Index to be quite useful here as well. On the daily chart above, you will note that the DI formed its initial peak in the latter part of March (point 1), and then, after a six-day correction, crude oil made new rally highs just above $58 a barrel. However, here the DI was clearly lower (point 2), forming a negative divergence. The daily candle formation shows a tight range between the open and close, a relatively higher upper shadow, and a lower close. This formation (see first circle) is often called an “evening star” and suggests that the sellers had taken over.

It is very important to realize that this divergence was formed only over two weeks of trading, and therefore, was indicative of a short-term correction, not a change in the intermediate trend. The length of time that it takes for the divergence to form is directly related to the significance of the signal. In other words, a divergence on the weekly chart that takes eight to 15 weeks to develop has much more intermediate-term significance than one that forms over only three to five weeks.

Crude oil prices corrected for several weeks into early April, where the DI formed its initial low. Then, after a five-day rally, crude again turned lower, but the DI held above the previous lows. An even stronger positive divergence was formed by the DI in late May (line B) as crude dropped down to the $46-per-barrel area. The small real body, or dojiformation (see second circle) at the lows on May 20 suggested that the market was sold out and the up-gap opening the following day, along with the strong close, was very positive. Several days later, the resistance in the DI (line C) was overcome, confirming the daily bottom formation.

Figure 4 - Click to Enlarge

This weekly analysis of Apple (AAPL) leading into the December 2007 highs is a good example of how multiple weekly divergences in the Demand Index can help you identify major tend changes. The DI moved above the zero line in late-July 2006 and held above it on the pullbacks in late 2006 and in early 2007. The rally into June 2007 took AAPL above $127 and set the stage for a three-week sideways period, dropping the DI back to 33. This then became a key level of support to monitor. Apple surged to a new high of $148.92 the week ending July 28, 2007, and the candle chart shows a star formation. The DI did not confirm these highs, point 2, and the engulfing pattern the following week took the DI below support at line a. AAPL then dropped to just below $112 over the next few weeks before the uptrend resumed. This seven-week divergence was consistent with a correction, not an intermediate-term top. Incidentally, the weekly OBV did confirm the July highs.

APPL again started to rally sharply, and by the end of September, the stock was at new highs. In early November, a new high (point 3) was followed by another bearish engulfing pattern, and the failure of the DI (point 3) to confirm these highs gave a stronger warning as there was important DI support at line b. On December 27, 2007, AAPL made a high of $202.92, which again was not confirmed by a new weekly high in the DI, point 4. The following week, another bearish engulfing pattern was formed (see circle), and this, combined with the six-month pattern of negative divergence, indicated that an intermediate top was in place. The next week, the Demand Index violated support at line b, confirming the divergence, and AAPL eventually dropped to the $115 area.

As I mentioned earlier, if you would like to read more about the Demand Index, you should read this previous lesson. The DI remains one of my favorite volume-based indicators, along with on-balance volume (OBV).

Tom Aspray, professional trader and analyst, serves as video content editor for MoneyShow.com. The views expressed here are his own.