Cycle Analysis Primer

| March 10, 2011

The Basics

Cycle analysis (CA) is an excellent complement to technical analysis (TA) because its focus is time, whereas TA is concerned primarily with price. The idea behind CA is that markets have internal rhythms, and if you are able to accurately identify those cycle periods, you can then forecast when and where tops, bottoms, support and resistance are likely to develop.

The first step in applying CA to a given market is identifying its cyclical properties. Every market is unique and thus has its own unique rhythms, so a large amount of historical data should be analyzed in order to gauge the periodic tendencies of a given index. Once you have identified the temporal properties of the market under study, it is possible to develop reliable forecasts of when cycle boundaries are likely to occur. The following chart displays the identified short-term cycle lows (STCLs) of the S&P 500 during a six-month period in late 2010 and early 2011.

Each cycle low denotes the end of one short-term cycle and the start of the next. Notice that the STCLs are relatively evenly distributed over time, although time and price are not always closely synchronized. Cycles that trend higher are said to be “right translated,” while cycles that trend lower are “left translated.” Each short-term cycle also has two to four component trends, so there are cycles within cycles. Cycles that have four component trends are divided into two halves, an alpha phase followed by a beta phase. The inflection points that occur within the short-term cycle are denoted by the alpha high (AH), beta low (BL) and beta high (BH) labels. While these intracycle inflection points will also tend to occur at relatively regular intervals, there will often be significant variation in both phase duration and price behavior.

Imagine a rubber ball bouncing at regular intervals up and down a series of staircases with varying degrees of steepness and duration. CA attempts to identify the points at which the ball peaks and makes contact with the ground.

Cycles within Cycles

As you may have already guessed, there are also cycles across longer time frames. In fact, the short-term cycle is one of four that we monitor. The other three, in order of increasing duration, are the intermediate-term cycle, the annual cycle and the long-term cycle.

The Intermediate-term Cycle

The intermediate-term cycle typically lasts from 15 to 25 weeks, depending upon the market. The following chart displays several intermediate-term cycles in the gold market during 2009 and 2010.

Like the short-term cycle, a given intermediate-term cycle will have either two or four component trends. If a cycle has only two component trends, it will have one intermediate-term cycle low (ITCL) and one intermediate-term cycle high (ITCH). However, if a cycle has four component trends, it will have two phases and form two half cycle highs (HCHs) and one half cycle low (HCL).

The Annual Cycle

As you might expect, the annual cycle, also known as the seasonal cycle, typically lasts about 12 months. The following chart displays two recent annual cycle lows in the US dollar market.

Here is an example of an annual cycle that lasted much longer than one year, as there were 20 months between consecutive annual cycle lows (ACLs). The annual cycle plays an important role in long-term cycle analysis, as annual cycle inflection points nearly always lead to long-term cycle trend changes.

The Long-term Cycle

Finally, the long-term cycle typically lasts anywhere from 4 to 9 years, again, depending upon the market. The following chart displays the last several long-term cycles in the oil market.

The long-term cycle is the dominant cycle that dictates trend direction within the overall secular trend.

Setups, Triggers and Signals

As with all such chart analysis techniques, the application of CA is inherently subjective. There are many different methods that can be used to categorize past cycle lows and forecast the development of future ones. We use a set of custom momentum and price oscillator technical indicators to perform our identification and forecasting processes, but there are other ways to employ CA.

Our CA methodology uses a two-step process to predict cycle inflection points. The first step is a signal setup in the form of oscillator crossovers, and the second step is a trigger in the form of an engulfment of a predetermined number of previous session closes. Once a signal has been generated by a setup followed by a trigger, a stop level is usually set at the extreme of the previous trend, and a move beyond that stop level causes the signal to be invalidated. The following chart displays a short-term cycle low signal in the gold market that occurred in July 2010.

A cycle inflection point is subsequently confirmed when price action exhibits follow-through in the direction of the turn. Our custom indicators correctly identify 95 percent of all market inflection points across all time frames while issuing false signals only 7 percent of the time.

Using Cycle Analysis to Improve Trading Performance

Because it facilitates the identification and reliable forecasting of financial market cyclic characteristics, properly executed CA assists in the selection of optimal entry and exit points. In other words, CA enables you to open long positions and close shorts positions near meaningful lows, as well as close long positions and open short positions near meaningful highs. Let’s design a simple trading system based upon CA in order to demonstrate its application and usefulness.

The first step in designing our simple trading strategy is to pick a time frame. The most basic form of swing trading targets market moves that last from several weeks to several months and those time periods align nicely with intermediate-term market cycles, so we will start with that foundation. Using a simple trend following strategy, our system will trade in the direction of the long-term trend using intermediate-term cycle inflection points as entry and exit points. In order to see this simple strategy in action, let’s apply it to the gold market, which has been trending higher since the last LTCL in 2008.

Since the long-term cycle is right translated, our trading system will enter long trades at ITCLs and close them at ITCHs, thereby capturing a majority of the intermediate-term moves. Had the long-term cycle been left translated, we would have entered short trades at ITCHs and then covered them at ITCLs. Again, this is a simple trend following strategy, so our objective is to stay aligned with the long-term trend.

The sessions during which the trades are opened and closed would then be chosen using short-term cycle analysis. For example, the latest long trade would have been opened at one of the last two STCLs.

Given the extremely long duration of the previous intermediate-term cycle, the STCL signal that was generated on February 3 would have made an excellent, low-risk entry point. However, waiting for the next STCL to occur at a higher level, thus confirming that the short-term cycle had transitioned from left to right translation, would have also been an effective strategy. Of course, in both cases, stop loss levels would have been set below the entry points in order to practice sound risk management.

The simple trading methodology outlined above effectively demonstrates the usefulness of CA. Knowing the approximate time frame during which lows and highs are likely to develop in a given market facilitates the selection of optimal entry and exit points, thereby enabling you to both maximize profits and minimize risk.

Another Market Analysis Tool

Like TA, CA enables us to better understand price behavior and develop more reliable forecasts across all time frames. However, also like TA, CA is absolutely not a crystal ball. It is simply another tool that helps us to identify the most likely possibilities, which is ultimately the key to long-term success as both an investor and a trader.

Share on FacebookShare on TwitterShare on LinkedInShare via email

Category: Articles, Commentary, Video


Comments are closed.