As discussed earlier, the Japanese commonly refer to a gap as a window. Whereas the Western expression is “filling in the gap”, the Japanese would say “closing the window.” In this section I will explain the basic concepts of windows and then explore other patterns containing windows (gaps). These include the tasuki gaps, the gapping plays, and side-by-side white lines.
A window is a gap between the prior and the current session’s price extremes. Exhibit 7.1 shows an open window that forms in an uptrend. There is a gap between the prior upper shadow and this session’s lower shadow. A window in a downtrend is displayed in Exhibit 7.2. It shows no price activity between the previous day’s lower shadow and the current day’s upper shadow. It is said by Japanese technicians to go in the direction of the window. Windows also become support and resistance areas. Thus a window in a rally implies a further price rise. This window should also be a floor on pullbacks. If the pullback closes the window and selling pressure continues after the closing of this window, the prior uptrend is voided. Likewise, a window in a declining price environment implies still lower levels. Any price rebounds should run into resistance at this window. If the window is closed, and the rally that closed the window persists, the downtrend is done.
Traditional Japanese technical analysis (that is, candlesticks) asserts that corrections go back to the window. In other words, a test of an open window is likely. Thus, in an uptrend one can use pullbacks to the window as a buying zone. Longs should be vacated and even shorts could be considered if the selling pressure continues after the window closes. The opposite strategy would be warranted with a window in a downtrend.
We see windows 1 and 2 in Exhibit 7.3 amid a rally which originated with the bullish engulfing pattern. A bearish shooting star arose after window 2. The day after this shooting star, the market opened lower and closed the window (that is, filled in the gap). Remembering the concept that corrections go back to the window, this pullback to the window should not be a surprise. If the window is closed and the selling pressure were to continue, the end of the uptrend would be flagged. This did not happen. The selling force evaporated once the window closed. In addition, the support set up at window 1 held.
During the week of February 20 the market shuffled its feet. It then retested support at window 2. After this successful test, the market pushed ahead and opened window 3. This was a significant window because it represented a gap above the $1.10 resistance level. This old $1.10 resistance area, once broken, should become support. Add to this the support at the window near this $1.10 area and you have two reasons to expect $1.10 to provide a solid floor. Throughout March, this area did indeed provide a firm footing for the bulls.
The Japanese believe windows from a congestion zone, or from a new high, deserve special attention. See Exhibit 7.4. The early March window above $.I5 was a significant break above a month-long congestion zone. Thus there was dual support in the window near $.15. The first was because of the window, the second was because the old resistance area had become support. Notice the solid support this window provided for the next few months.
April 2 and 3 comprised a harami. This indicates that the prior trend (in this case, a downtrend) had run out of steam. A bullish engulfing pattern formed a few days later. On April 16 an inverted hammer appeared. Each of these patterns appeared near the window’s support at $.15.
In March 1988, a bullish engulfing pattern presaged a rally. (See Exhibit 7.5.) A window opened during the rally. The rally progressed until the bearish counter attack line. The window held as support for five weeks but the persistence of selling after the window closed annulled the uptrend.
Thus far, the focus of this section has been on the use of the window as support or resistance and as a continuation indication. There is another use. (See Exhibit 7.6.) A window, especially if it is made with a small black candlestick from a low price congestion area, can mean a meaningful upside breakout. Exhibit 7.6 illustrates this principle. Throughout February prices were locked in a relatively tight congestion band. Between February 24 and 25, a small upside window opened via a diminutive black candlestick. This window was confirmed as support the next session. On that session (February 26), the market not only held the window as a support but produced the strongest type of candlestick line, a long white candlestick that opens on its low (that is, a bullish belt-hold line) and closes on its high.
A large window appeared in mid-January as indicated by Exhibit 7.7. From late January to late February, there were numerous return moves to this window (corrections go back to the window). Each of these rallies was short circuited when they got near the resistance level created by the window.
Look at the Dow in Exhibit 7.8. The “Crash of ’87” created a window in the 2,150 to 2,200 zone. Two criteria were needed to tell us the downtrend was over. The first was to close the large window. The second was for a continuation of the buying force once the window closed. These criteria were met in early 1989.
Exhibit 7.9 is another example of a window creating resistance. The narrow window 1 in late May implied continuation of the decline. It also became resistance within the next few weeks. Interpreting window 2 offers a chance to underscore the importance of the trend. Real estate agents say the three most important factors about a property are location, location, and location. To paraphrase this, the three most important aspects of the market are trend, trend, and trend. Here, in Exhibit 7.9, we see a market whose major trend is southward.
In this environment, a bullish morning star emerges. Do you buy? No, because the major trend is down! Covering some shorts would be more appropriate. When should nibbling on the long side in this market begin? In this case, it is if the market pushes above $.I164 and buying continued after this level. This is because in mid-July the market formed a window (window 2). The top of the window was $.1164. Until the bulls were able to prove their vigor by pushing prices above this window, going long should be viewed as a high-risk strategy in spite of the morning star. The morning star did act as support on a test of its low a few days after it formed. Yet, after trying for a week the bulls failed in their attempt to close window 2. This told us a new rally was not likely. The moral of this story is that candlesticks, or any other technical tool, should be considered in the context of the prevailing trend.
In Exhibit 7.10, we see that the market headed south after September’s hanging man and the black line which engulfed it. The window in late September signaled a continuation of the decline. The window was closed, but buying steam quickly dissipated as proven by the shooting star.
There are three windows to discuss in Exhibit 7.11. Window 1 is a downside window from March 1989. It became a resistance level over the following few weeks. Window 2 is another downside window which means more selling pressure will be felt. A long white candlestick a week after the window formed a bullish engulfing pattern. This was the first sign of a bottom. The next week, the market closed sharply above the window. This generated another reason to believe the selling pressure had dried up. Window 3 is a window in a rally which means to expect more strength. This window closed in the second week of October, but not for long as buying pushed prices higher and in the process fashioned a hammer. Normally hammers are important only during a downtrend (since they are bottom reversal signals). In this case, it became important because it reflected a test of the window’s support level. If the market had continued lower after this hammer, the uptrend would have been nullified.
Exhibit 7.12 shows a series of three windows. Window 1 became support when the market sold off a few days after the window opened. Window 2 stopped the rally a month later. Window 3 kept a ceiling on all the rally attempts during the week following its opening. Another interesting aspect about the September rally that stalled at window 2 is that the rally, as shown by the numbered days 1 through 8, made eight new higher highs. Candlestick theory states that after about eight to ten new highs or lows, without a meaningful correction, the odds are strong that a significant correction will unfold. Each new high or new low for the move is called a “new record high” or “new record low” by the Japanese.
Thus the Japanese will say there are ten record highs or lows, meaning there were a series of ten higher highs or lower lows. If there are, say, eight new highs without a meaningful correction, the Japanese refer to the market by using the expression “the stomach is 80% full.” What is interesting about this gold chart is that there were eight record highs. This gave a warning that a top might be near. The fact that after these eight record highs the market was at a resistance area set by window 2 was an extra strong signal to be cautious about the long side of this market.
The enchanted number three makes yet another appearance in Exhibit 7.13. Traditional Japanese technical analysis posits that after three up or down windows, the chances are- strong that a top (in the case of three windows in an uptrend) or a bottom (in the case of three windows in a downtrend) is near, especially if a turning point candlestick pattern or line appears (such as a doji, piercing pattern, or dark-cloud cover) after the third gap. Here we see hanging-man lines after the third window.