Choose a Site
Share:
Send to FacebookTweet ThisShare on Google+Email to a Friend
Download

Seasonality: Studying Annual Price Cycles

Related Files

Download Files 

By Stanley Dash


Vice President, Applied Technical Analysis
TSLabs@TradeStation.com

 


Features:
  • Focus: Technical
  • Markets: Equities, Futures
  • Time Perspective: Intermediate term,
                                  Long term
Studies/Files Included:
  • Indicators
  • Workspace

 


Summary

 

The study of seasonal tendencies in price movement, also known as seasonals or seasonality, is an area of market analysis intended to identify price patterns and possible turning points within the span of a year.

 

This area of study probably originated in the traditional agricultural commodity markets. In these markets, it would seem logical for prices to rise and fall as the market tracks crop development, supply arrives at harvest time and product is consumed throughout the year. However, this area of study has expanded to include the search for seasonality in other markets, including equities.

 

This Analysis Concepts paper discusses the ideas and terms behind the study of seasonality and provides some tools for monitoring seasonal patterns in TradeStation.

 

Figure 1: Copper, weekly, showing seasonal indicators based on 5 years through 2011
Figure 1

 

Background

 

The study of cycles has long been a part of technical analysis. The goal is to identify a rhythm for price movement and the timing of peaks and troughs. Is there a symmetry to these patterns that provides guidance to upcoming tops and bottoms?

 

Some cycles, such as the Kondratieff Wave, are as long as 50-60 years. Others are shorter, such as the 34 year cycle and the 4-year Presidential Cycle. The existence of some of the more esoteric cycles is debated passionately by technical analysts. And while that debate rages among analysts, traders might scratch their heads over the practical application of a cycle of, say, 50+ years. (If you are fortunate, you get one chance at seeing Halley’s Comet; an octogenarian might get two.)

 

Webster’s entry for seasonal includes this usage note: “Seasonal describes phenomena that occur with or depend upon a season or the seasons.”(1) Seasonal studies are a subset of cycle analysis, specifically the study of cycles that occur within a year and tend to recur in subsequent years.(2)

 

We briefly mentioned the likely origins of this analysis in the traditional commodity markets, as supply changes seasonally while demand is relatively constant throughout the year. Prices move in reaction to the supply and demand imbalances that occur with planting and harvest. It is easy to see how seasonal influences might play into markets such as grains and oilseeds.

 

And it is not a stretch to see how seasonality may become evident in energy markets, too. We know that U.S. consumption of gasoline and heating oil (along with heating oil consumption in Europe) revolves around the summer travel season and the cold of winter in the northern hemisphere. Although refiners have some ability to accommodate this by squeezing more of the needed product from each barrel of crude oil, it is the pricing function of the marketplace that reacts to these seasonal shifts.

 

A few notes are appropriate here.

 

The market is a discounting mechanism, and will anticipate seasonal changes in supply and demand. Therefore the tops and bottoms that you may expect at certain times may occur earlier as the market tries to jump ahead of a harvest or the approach of winter. There is reference to this in the paragraph below on oil markets.

 

Seasonal patterns are not fixed and the timing of seasonal tops and bottoms is by no means fixed. Changes in such factors as global trade, the geographic distribution of supply and demand, and the importance of national economies once considered minor all work to reshape the seasonal curve.

 

Examples include the impact of southern-hemisphere production of crops traditionally dominated by northern-hemisphere production; consumption of agricultural and industrial commodities in emerging economies such as China; and commodities that may offer competition to other established markets, such as corn and natural gas in energy markets.

 

Here is a small selection of charts with some comments. In viewing these or any graphic representation of seasonal patterns, remember that seasonal studies are not intended as maps of price action or forecasts of specific price levels; rather, they are used to identify general trends and turning points that may recur annually.

 

Agriculture

Corn prices may present us with the most classic of agricultural price cycles. Seasonal tops tend to occur around June, the period of new-crop growth during which the crop is most vulnerable to growing conditions. Seasonal lows fall in the August – October period, when the crop is already “made” and into the harvest.

 

Figure 2: Corn, weekly, showing seasonal based on 5 years through 2011
Figure 2

 

While the discussions and charts presented here use a calendar year, it is common for seasonal studies in agricultural markets to use “crop year” or “crop marketing year.”  The crop year generally runs from harvest to planting and so for corn is considered October to September.

 

Energy

It is also easy to see how seasonal influences may affect energy markets. Europe and the northeastern U.S. are the critical regions for consumption of heating oil during the winter in the northern hemisphere. And Americans are known for their propensity to drive during the summer months.

 

Figure 3: Heating oil, weekly, seasonal based on 5 years (red) and 25 years (blue) through 2011
Figure 3

 

As noted above, seasonal tops and bottoms often occur at times other than those you might anticipate. Certainly, the coldest temperatures of the year in, say, the northeastern U.S. are in January and February. That is a given, so the oil industry refines and ships heating oil, and distributors stock up well before that. The major seasonal top in heating oil tended to occur in September – October. This may be seen in figure 3 above; the blue line is the seasonal based on 25 years of weekly data, through 2011. More recent data shows a shift to an earlier peak in mid- to late summer; the red line is the seasonal based on only the most recent 5 years through 2011.

 

Financial

Compared to agriculture, it may seem counterintuitive to look for seasonal factors in financial futures. Yet, seasonal studies in financials can show some interesting patterns. For example, S&P futures tend to have a seasonal top in the spring and bottom in the fall and late winter.

 

Figure 4: E-mini S&P, weekly, showing seasonal based on 10 years through 2011
Figure 4

 

Seasonal studies are common in other financials, too. Patterns are evident in currencies and interest rates, for example. Theories on seasonal influences in financials include such phenomena as timing of tax payments and receipts, income repatriation and global trade.

 

Equities

In addition to the seasonal influences that may be apparent in stock prices generally, there may also be discernible seasonal patterns in individual equities or sectors. These patterns will vary based on the industry. For example, there may be interesting seasonal patterns in homebuilders, travel and leisure, and retail – not to mention those industries directly connected to the commodity price cycle such as oil and food processors.

 

Figure 5: $SPHOME (S&P Homebuilding Select Industry), weekly, showing seasonal based on 10 years through 2011
Figure 5

 

Calculating Seasonal Patterns

 

Seasonal studies are generally done on daily, weekly or monthly data. In fact, many analysts will restrict their work to weekly or monthly data only. Daily data may be prone to an excess of noise in an area of study intended to see ebb and flow on a higher level. The indicators provided with this paper are for use on weekly and monthly charts only.

 

There are two common types of seasonal studies.

 

One of these is to create a curve for the year based on average price movement historically. This type of indicator is shown in figures 2 through 5 above, and in a sub-graph in figure 1. These types of analyses are done by averaging prices and price movements in each period, such as monthly, over a given number of years. That is, averaging the price action in the last 10 Marchs, Aprils, Mays, etc., to create a 10-year seasonal curve for the year. The monthly average price changes are then summed.

 

Again, keep in mind when looking at these charts that any cyclical analysis attempts to identify turning points, or relative highs and lows, and not absolute price levels per se.

 

There is a wide variety of methods for handling the data and making these calculations. An excellent reference for this is in the work of Perry Kaufman. (See Other References at the end of this paper.)

 

The other common form of seasonal study is an analysis of the frequency of a seasonal tendency. For example, again referring to monthly data, we might examine how many of the last 10 Januarys showed positive returns and how many had negative returns. This data not only is useful on its own, but adds perspective to the seasonal price curve analysis.

 

Figure 6: Copper, monthly, showing seasonal pattern and monthly directional frequency based on 10 years through 2011
Figure 6

 

A seasonal tendency accompanied by a strong historical frequency may be more reliable. And since seasonal curves can be strongly influenced by an exceptional weather year or unusual geopolitical event, this frequency analysis may help identify anomalies.

 

TradeStation Labs provides one of each of these types of indicators for download with this paper.

 

TradeStation Labs Seasonal Indicators

 

The two TradeStation Labs indicators provided with this paper are for use on weekly and monthly charts only. And both of these indicators are displayed beneath the bars for the current year, through the end of the current year. That is, the indicator is plotted for an entire year through December as soon as the year begins. All the calculations are made using data from the years prior to the current year as specified in the input YearsToConsider.

 

Please be sure to load the chart with one year of historical data more than your YearsToConsider input. Also, the chart must have an adequate Space to the Right setting on the Format Window – General tab to allow for a full year of weeks or months, per the data interval.

 

TSL:Seasonal Performance

 

This indicator plots the seasonal price curve as displayed in all the charts above. The indicator may be applied to the same chart more than once using different values for the YearsToConsider inputs. This was done in figure 3, with two instances of the indicator in the same sub-graph.

 

Plot Description
Seasonal Plots the seasonal price over one year

 

Input Default Description
YearsToConsider 5 Number of years of data, prior to the current year, to use in calculating the seasonal pattern

 

TSL:Seasonal UpDn Performance

 

This indicator plots the frequency of positive or negative returns in each period during the preceding years. Figure 7 is the same indicator on the same data as shown in figure 6. In this example, the return of each of the 12 months is calculated for each of the last 10 years. The indicator then calculates how many of each of those periods were up or down in the last 10 years.

 

Figure 7: Copper, monthly, showing monthly directional frequency based on 10 years through 2011
Figure 7

 

For the sake of graphic clarity, the histogram displays a normalized number, making it easy to spot positive and negative values that may be important. Referring to figure 7, we see a value of +30 for July, indicating that copper has had positive returns in July in most of the last 10 years. More specifically, 80% of the last 10 Julys had positive returns. May and June have a value of 0, so there are no visible histogram bars. That indicates no statistical edge in the frequency of returns for that month: 50-50.

 

The horizontal reference lines noted in the lists of plots and inputs below have both been set to zero in figures 1, 6 and 7 to create a single base line. The user can adjust the inputs as desired.

 

Plot Description
UpDnPerf Plots the frequency of higher or lower returns for each month
UpperBand Horizontal reference line
LowerBand Horizontal reference line

 

This indicator also draws a vertical line on the price bars to delineate the years.

 

Input Default Description
YearsToConsider 5 Number of years of data, prior to the current year, to use in calculating the frequency of returns
UpperBand 20 Location of horizontal reference line
Lowerband -20 Location of horizontal reference line
UpColor DarkCyan Plot color for positive histogram value
DnColor Red Plot color for negative histogram value

 

Conclusion

 

Seasonality is a vibrant area of study among technical analysts and traders. This subset of cycle analysis provides perspective on price action driven by the passing of the seasons as well as by the financial infrastructure that society has built around the calendar.

 

Monitoring seasonal tendencies should be a complement to other tools that generate specific trading signals. For example, a trader may adjust trade size on a given signal based on a seasonal bias. Or signals could be taken only when they agree with the general seasonal tendency.

 

Seasonal studies are generally not used in short-term trading; maintaining the proper time perspective for a seasonal tendency, and considering how it relates to the time horizon of a trading signal, is an important aspect of integrating seasonality into trading decisions.

 

 

 

References

1. Webster’s Encyclopedic Unabridged Dictionary of the English Language, Random House, 2001.
2. Murphy, John, Technical Analysis of the Financial Markets, NYIF, 1999, page 359.

 

 

Useful References

1. Hirsch, Jeffrey, ed. Commodity Traders Almanac. New Jersey: John Wiley & Sons, Inc.
2. Kaufman, Perry J. (1998). New Trading Systems and Methods (4th ed.). Hoboken, NJ: John Wiley & Sons, Inc.

 

 

 

 

To use the files provided with this issue of Analysis Concepts:

  • Files with extension ".eld" – These contain EasyLanguage® documents: analysis techniques and strategies. Double-clicking on this file will start the EasyLanguage Import Wizard. Follow the prompts to completion. The analysis techniques or strategies will automatically be placed in the correct locations for your use in TradeStation. This should be done before opening any workspaces provided.
  • Files with extension ".tsw" – These are TradeStation workspaces. These may be stored in any folder where you choose to save TradeStation workspaces.
  • Files with extension ".txt" – These are text versions of the EasyLanguage documents and are generally used only by advanced EasyLanguage users.
  • Other supporting documents or files may also be attached to the report.

Questions or comments? Contact us at TSLabs@TradeStation.com