Navigating a correction is similar to flying through a storm cloud. Everyone on board has confidence they’ll get through it, even if they have no idea how it will happen.
In complex corrections, most traders are like the passengers on that plane. They know the correction will eventually end even though they have no idea when it will happen.
The time element of technical analysis works the same way as the instrument approach that a seasoned pilot uses to land a plane in zero visibility. “Finding patterns with the Lucas Time Series” (September 2006) highlighted how the time dimension can be used as a pattern recognition system. Here, we will look at ways to navigate through a correction until market direction becomes clear again.
Corrections in all financial markets offer unique challenges. To the trend follower, a whipsaw that depletes the bankroll isn’t a distinct possibility; it’s a fact of life. A momentum player has to decide where to bail out in the event a pullback retraces into the prior trend.
One discipline uses moving averages while another depends on common Fibonacci retracements. The dilemma is that price action can spike beyond the moving average or Fibonacci point, taking the stop and leaving the trader on the sidelines. Traders also use different moving averages. Which retracement or moving average dominates is a constant challenge even to the experienced trader. Price action usually lands near a cluster of common Fibonacci points or between a set of moving averages.
The other challenge in recognizing corrections is whether they will be what Elliotticians call a sharp or flat pattern. A sharp correction will move counter to the main trend and will generally retrace 38%, 50%, 61% or 78% of the main trend. The problem with sharp corrections is the difficulty in determining which one of these levels will end the pullback.
A flat correction will take on the shape of a sideways consolidation and sometimes turn into a triangle.
The challenge for traders when working with sideways consolidations is they seldom retrace beyond 38% and sometimes only 23%. Moves that retrace between 23% and 38% are fairly simple to recognize, but the trick is knowing when they will end.
All markets cycle according to Fibonacci price and time movements. Of course, it’s not perfect; the Fibonacci time sequence works about 60% to 70% of the time. However, when it doesn’t work, the Lucas series fills most of the gap. Most traders who have become familiar with the Fibonacci time dimension are only now becoming aware of how often the Lucas series reveals itself in the markets.
Three case studies demonstrate this condition with different degrees of complexity.
Exchange-traded funds (ETFs) have become popular for traders and investors who want to play a certain sector of the market without having to rely on any particular stock.
“Crazy like a Vox” (above) is a daily chart of the Vanguard Telecommunications Services (VOX) ETF, which represents the telecom industry. This example highlights both challenges for traders even though it is a fairly common pattern. The first problem is to determine if the correction will be deep or wide. In real time the pattern develops into a complex sideways wave that is tricky to play.
Fibonacci retracements can guide the trader in determining the depth of the pullback. When drawing retracement lines, the challenge is determining which major pivot is going to dominate. The solution is simple: Draw retracements from both pivots.
There are two major pivots in this example, one on Oct. 19, 2005, and the other on Jan. 3, 2006. At the June low, price action goes beyond the 38% retracement level of the January secondary pivot but stops just short of the 38% level of the lower October pivot.
There are actually two opportunities to buy a dip. The first one came on a gap up after a 24-day pullback at the June low with the large white candle. That particular opportunity would have netted a small profit on a swing trade but longer-term players would have been stopped out. The 47-bar cycle offers a higher probability of success. The retest of the low in July completed a 47-day running correction where the final leg takes off without the benefit of having made a new price low. The final retracement for this leg touches the 50% retracement off the secondary low and when it is combined with the 47-day cycle, this was the higher probability pattern recognition signal.
“Down with Morgan” (below) demonstrates a classic sideways or descending triangle consolidation. After a healthy rally off the bear market bottom, JP Morgan Chase (JPM) tops the week of March 1, 2004, but doesn’t break out again until early October 2005. The first leg down bottoms on the 38% retracement level off the March 2003 pivot (retracement lines not shown). However, as time goes by, price action flattens out.
This suggests traders have the price retracement solved, but it still doesn’t break out. The reason is that a flat correction can make up for the lack of steep price retracement by working off the overbought condition through a longer period. Many traders follow oscillators like the Relative Strength Index (RSI) to determine overbought and oversold conditions. These are not trigger signals but give traders an idea of overall sentiment. As you follow the progression of the RSI, observe how long it takes to work off the overbought condition.
Luckily, the time dimension, like the instruments on the plane, provide a guide as to where price action is in the progression of the pattern. The initial low bottoms and forms a bullish candle formation in “week 11.” As it turned out, it was only a swing trade as secondary highs formed in the seventh and 21st weeks off the low. In each case, a working knowledge of candlestick methodology gave advance warning of a reversal.
The secondary high to this pattern came on the 21st week, which led to a retest of the low 56 weeks off the top. The May 2004 was eventually taken out, to the tune of pennies in April 2005, but the pattern still did not break out. There was yet another retest of the low in October 2005. It was only after another 55 weeks off the secondary high from October 2004 that price action took off for good. This low also coincided with market conditions as the Nasdaq broke out of an 89-week triangle the same week. This breakout led to a whopping 43% gain by April 2006.
The key to recognizing when this pattern would break depends on several factors. It’s important to have a working knowledge of candlestick reversal methodology together with an understanding of the time relationships. Neither of these corrections retraced sharply into the overall trend.
When that happens it’s easy for traders to get lost in the winding nature of these patterns, especially when moving averages provide little comfort in these type of market conditions. The good news is these flat corrections often follow repeatable time tendencies, which are easy to recognize.
THE TIME FACTOR
In many instances, a sharp correction will terminate on a Fibonacci retracement of time and close to the price retracement. From the February 2003 low, the Nasdaq powered higher to the January 2004 high in 234 trading days. It then retraced for 143 days to the August 2004 low for an almost perfect 61% time retracement as it stopped right at an intermediate level 38% price retracement. That was a straightforward cluster of time and price. Many situations traders deal with everyday are more complex.
“When technology breaks” (below) shows a sharp correction that exhibits a more common complex time calculation. Luckily, high probability pattern recognition tendencies enable the advanced calculation to be made simple. The Nasdaq traced an intermediate level high in April 2006.
What makes this complicated is most Fibonacci cycle analysts and traders have come to expect turns on important windows, such as 55, 89, 144 or 233 days. Few realized it at the time but the high came in on the 423-day (to be precise on the +1 day) trading window off the August 2004 pivot low. The 423 calculation is not a Fibonacci number, but it is derived from the 4.23 extension of waves, which is a common Fibonacci relationship.
The 2006 correction was steep as well as swift. There were not many indications of a bottom in July. The challenge to traders was the choppy nature of the rally on the way up. There were five larger degree pivots from the October 2002 low. For simplicity only the lines from August 2004 are shown.
Most traders will draw retracement lines from only the closest pivots, and the October 2005 low was already taken out. The July low at 2012 was close to the 61% retracement line of August 2004 pivot and 50% of a smaller pivot from August 2003. It’s not practical to be concerned about so many pivots. Sentiment at the time suggested most traders were expecting more downside action. It never came.
As it turned out, the Nasdaq leg off the August 2004 low expired in 424 (+/-1) trading days. The pullback was 62 days or a 0.146 time retracement. The 0.146 number has a variety of Fibonacci relationships and is a common yet lower probability retracement level. These less-obvious time relationships materialize daily in all time frames.
Here is where it is necessary to make the complex simple. You don’t need to be too concerned about the 423-day leg or the 0.146 time retracement.
If you are aware of that relationship, it is fine but there is an easier way to play the situation. There was a set of high probability tendencies traders could easily recognize. At the July low, the Dow was 47 days off its May high and the Nasdaq was 62 days off its April high. These easy-to-recognize tendencies came with bullish candlestick patterns off the lows. Nobody really knew how far the ensuing rally would go, but shorts that were aware of these tendencies should have been ready to cover in the face of a counter-trend bounce. Those who went long could at least anticipate a test of technical retracements at higher levels.
In many cases, traders familiar with Fibonacci price retracements can rely upon price action validating common retracement lines by stopping right on them. When that is the case, the task at hand is much easier. The problem comes when price action doesn’t land as anticipated.
But if you wait exclusively for such perfect opportunities you’ll miss many setups. An understanding of these tendencies solves this problem. As was the case with the summer low, by the time many traders realized the bottom was in, risk was much higher too. Only the most aggressive and seasoned traders were able to take advantage of the July low and lower risk/reward levels. That these calculations contributed to a leg that not only took out the 2006 high but created all-time highs in the Dow demonstrates how powerful they are.
In each example shown here, the time dimension helped overcome traditional challenges most traders have in navigating through flat and sharp corrections. These patterns are more progressively complex but when seen through the lens of the time dimension they are simple and practical to use. They work in all markets and time frames.
Jeff Greenblatt is the editor of the Fibonacci Forecaster, a forecasting service that covers the stock market, gold, bonds and crude oil. He is the author of a book on market timing using Fibonacci and Lucas relationships to be released this summer by Marketplace Books. He can be reached at email@example.com.