Do 38.2 percent and 50.0 percent Fibonacci correction levels require different rules from the ones described thus far in this chapter? This question is important because nothing is more frustrating for a trader than to wait for the Fibonacci retracement level 61.8 percent, and then shortly before it is reached, to have the market turn with no trade done.
The dilemma of making the right decision at the right time based on Fibonacci price corrections becomes obvious when we look at the Dax 30 Index in Figure 6.6.
A strong rally is followed by a price correction to the first Fibonacci retracement level of 38.2 percent. At this point—when almost all information services, market letters, and media representatives call for new lows—the trader must decide whether to wait for the 61.8 percent retracement level or to risk executing a long trade in the Dax 30.
What we need and are trying to find here is additional support that makes the decision a little easier. A tool that could help us consider the final strength of the price correction would be ideal.
One tool that has proven reliable already when used in combination with the 61.8 percent Fibonacci correction level is candlestick charting. If we change the charting technique and add candlesticks to the chart of the Dax 30 Index, we see a clear picture that makes our decision convincing and valid (see Figure 6.7).
From the chart, we can draw the following rules for a solid short-term long trade in the Dax 30 index:
• The Fibonacci Level 38.2 percent is reached.
• After the correction level is reached, we identify a doji candlestick pattern, which is a strong indication that the market price will at least not drop further from this price level. The following day, we identify an engulfing candlestick pattern that confirms the market price is ready for a correction.
• To receive still more confirmation of the potential trend reversal to the upside, we can wait two more days; the market price forms a double bottom.
At the latest on the fifth day after the price movement touching the 38.2 percent retracement line, a harami pattern assures us of a possible market entry long at controlled risk. We set the stop-loss at the low of the day that breaks the 38.2 percent correction line.
A similar simulation can be conducted for the 50.0 percent retracement level, which is also well recognized by traders. For traders who plan to keep stocks longer in a portfolio, this is a decent correction level, especially because it is reached much more often than the conservative 61.8 percent Fibonacci retracement level.
Figure 6.8 shows a chart of the S&P 500 Index between May and September 2002.
Once again, we are caught in a situation where waiting for the 61.8 percent correction level might lead to a dead end with no trade at all. Using plain correction levels without additional indicators has little value, especially for traders with a short-term orientation. As before, this is where candlestick pattern recognition becomes part of the game.
Integrating the candlestick charting technique into the general picture of Fibonacci price corrections changes the S&P 500 Index chart, as shown in Figure 6.9.
The candlestick pattern analysis runs as follows for a long signal in the S&P 500 Index once the price move breaks the 50.0 percent retracement line:
• The harami candlestick pattern is a first indication that the market correction might not continue.
• The harami pattern is followed by a hammer candlestick pattern. The hammer is a very strong confirmation that a trend change to the upside is about to occur.
• Finally, and in addition to the candlestick patterns, the resistance line is significantly broken to the upside.
We get a buy signal at the high of the day with the hammer candlestick pattern and on the breakout of the resistance line. The most recent valley before the long entry determines the stop-loss.
Instead of combining Fibonacci price corrections with candlestick patterns or 3-point chart patterns, traders can also merge them with Fibonacci extensions, support and resistance lines, and PHI-ellipses in an integrated Fibonacci-related approach.
Was this article helpful?