Seasonality is the tendency for securities to perform better during some time periods and worse during others. These periods can be days of the week, months of the year, six-month stretches or even multi-year timeframes. For example, Yale Hirsh of the Stock Traders Almanac discovered the six-month seasonal pattern or cycle. Since 1950, the best six-month period for the S&P 500 extends from November to April. By extension, the worst six month period runs from May to October, which is where the phrase “sell in May and go away” comes from. StockCharts offers a seasonality tool that chartists can use to identify monthly seasonal patterns. This article will explain how this tool works and show what chartists should look for when using our Seasonality Charts.
The seasonality tool calculates two numbers: the percentage of time the month is positive and the average gain/loss for the month. The calculations are pretty straightforward. Looking back over 19 years, there will be 228 months in total and 19 data points for each month. If a symbol advances 9 times for one particular month, then the seasonal bar will show 47% (9/19 = .47). If a symbol advances 15 times for a particular month, then the bar will show 74% (15/19 = 74%). The example below shows the seasonality tool in histogram format for the continuous contract for Gold ($GOLD). The number at the top shows the percentage of time $GOLD closed higher for that month. The number at the bottom shows the average gain/loss for those 19 months.
Astute chartists may have noticed that the slider at the bottom shows 20, which implies 20 years. This article was written in January 2014, which marks the beginning of the 20th year, hence the number 20 in the slider. The calculations from February to December are based on 19 years of data. The January calculation is based on 20 years of data and includes month-to-date performance for January 2014. The performance for January 2014 is subject to change until the month ends, which means the numbers in the histogram could change.
Chartists should note the current date when looking at a seasonality chart. If you are looking at a seasonal chart in mid-June, then the seasonal numbers from January to May will be based on the number of years shown in the slider (i.e. 20). The seasonal numbers for June will also be based on the number of years in the slider, but these numbers are subject to change because June is still a work in progress. The numbers from July to December will be based on the number years in the slider less one (i.e. 20 - 1 = 19).
Seasonality tells us what has happened in the past, which is the historical tendency. There is certainly no guarantee that past performance will equal future performance, but traders can look for above average tendencies to complement other signals. On the face of it, a bullish bias is present when a security shows gains more than 50% of the time for a particular month. Conversely, a bearish bias is present when a security rises less than 50% of the time. While 50% represents the exact middle, chartists should look for more extreme readings that suggest a relatively strong tendency. For example, readings above 65% would show an above average bullish bias, while readings below 35% would show an above average bearish bias.
The example above shows Intel (INTC) with a strong bullish bias in April (84%) and October (79%). Also, notice that the average gain is 6.3% in April and 6% in October. On the bearish side, the stock moved higher only 32% of the time in September, which means it moved lower 68% of the time. The average loss in September is 4.5% and traders would have been rewarded for waiting until October 1st to consider buying. Notice that the remaining eight months did not have a strong bias because they range from 42% to 58%. Before leaving this example, notice that Intel was up 58% of the time in June, but the average gain was actually a loss. Even though Intel moved higher more often than it moved lower, the losses during the declines outpaced the gains during the advances.
Chartists can also measure relative seasonality by comparing the performance of one security against another. The option to add a “compare” symbol can be found just under the chart. The example below shows the Russell 2000 relative to the S&P 500 over the last twenty years, which reflects the performance small-caps relative to large-caps.
Chartists can use relative seasonality to find stocks, sectors or groups that outperform the market during certain months. As the chart above shows, the Russell 2000 shows a strong tendency to outperform the S&P 500 in December (74%). In addition, the Russell 2000 outperforms the S&P 500 by an average of two percentage points. Chartists can confirm this by looking at the individual seasonal charts for $RUT and $SPX. As of January 2014, the average gain for the S&P 500 was 1.6% in December and the average gain for the Russell 2000 was 3.6%.
There are three different viewing options available: line, same scale and histogram. The examples above were shown in histogram format, which makes it easy to see aggregate performance for each month. Chartists can switch between viewing options by clicking the icons in the lower left-hand corner. The example below shows a line version for Merck (MRK) relative to the S&P 500 ($SPX). First, notice that January is still a work in progress because the red line has yet to reach the end of the month. This is because the chart was created on January 27th and the month was incomplete. Second, notice that there is no scale because we are only concerned with directional movement. Did the lines move up or down? Notice how the lines tended to decline in May and rise in September.
The middle icon is for viewing with the same scale for all lines. This means all performance lines begin at zero percent and chartists can compare performance month-to-month or year-to-year. Notice that 2013 was the best year, of the four complete years on the chart. The fifth year, 2014, is still a work in progress because this chart was created in January 2014. 2010 was the worst year with the smallest gain, which was barely positive. And finally, note that chartists can click the year icons at the upper left to hide or show the line for a particular year.
Seasonality is a tool that gives chartists a historical perspective on performance tendencies. Even though past performance does not guarantee future performance, chartists can use these seasonal patterns to increase their edge. Chartists can look for bullish setups when the seasonal patterns are strongly bullish, and bearish setups when seasonal patterns are strongly bearish. As with all indicators and technical analysis tools, seasonal charts should be used in conjunction with other analysis techniques.
Chartists can access Seasonality Charts by clicking on “Charts & Tools” at the top of any web page at StockCharts. Simply enter a symbol in the Seasonality section of the Charts & Tools page and click “Go”. Our Seasonality Charts article in the Support Center describes how to use all the controls.