Custom Search

daily seasonal tendency

Figure 5-3 shows the daily seasonal tendency for July soybean futures during the month of June. While the soybean market is not a market currently recommended for day trading, I chose it for inclusion, since the data history here dates back to the 1960s. As you can see from the chart, there are a number of high-percentage moves, as well as two pronounced seasonal tendencies during the month (denoted by heavy black lines on the "all years" line plot).

Further studies on key date seasonals can be performed by examining the data in a variety of different ways. We could compare the opening and closing price each day in given markets in order to determine if there has been a tendency for the close to be less than or greater than the open on certain calendar dates. When we perform such an examination of the data, this is exactly what we find. Again, the main consideration here is not whether we can isolate such tendencies, but rather, whether these tendencies are merely artifacts of the data (i.e., random events) or underlying characteristics of the markets themselves. I'm inclined to think that they do indeed represent patterns that are innate to each market; however, more research is required.

One way of improving the probability of key date patterns is to combine them with timing. In so doing we can hopefully capture the best of both worlds, allowing a timing indicator to validate or negate a pattern. This is discussed in greater detail later in this chapter.

Using Day-of-Week and Specific Date Price Patterns for Day Trading

Let's take a look at some day-of-week and specific date price patterns with an emphasis on their applicability to day trading. We'll begin with preholiday behavior. The premier researcher on specific date patterns in price behavior was Art Merrill. His claim that the price of the Dow Jones Industrial Average tends to close higher on the day before certain major holidays was supported with historical data back to the late 1800s. Merrill found an astonishing tendency for the Dow Jones Industrial Average to close the day higher than the previous day on the day before Christmas, Labor Day, Independence Day, Thanksgiving Day, and New Year's Day. His statistical analyses firmly supported his claim that preholiday behavior was not a random event. In other words, the probability of such patterns occurring by chance is about one in ten thousand!

Although the Merrill work was updated only through 1984, the pattern has remained valid. Consider the list shown in Figure 5-4.

How can the day trader take advantage of these statistics? There are two possibilities. First, a long position could be entered on the close of trading the day prior to the target day. The position would be closed out the next day. This, however, would not be a true day trade, since the position would need to be carried overnight. Another way would be to buy on the opening of the target day and exit on the close. While this would be a true day trade, it would not be a trade made exactly in accordance with the Merrill research;

however, it would be fairly close in most cases. I have updated these findings for S&P 500 futures in tabular form (Figure 5-5) for the Christmas, New Year's, and Independence Day holidays from 1984 to 1997. What do you conclude?

I believe that there are other days that have shown a high probability of closing up or down. The task of finding such dates is not a difficult one provided you have the data as well as a computer to analyze it. My work has pinpointed many dates across all active markets that have shown a high probability of closing higher or lower than the opening price. In this case, these are pure day trades, since they would be entered on the open and exited on the close. A sample of such dates in several markets appears in Figure 5-6.