Much like the popular assumption that you should “sell in May and go away,” ある、 popular theory that US equities almost always move higher during the month of January. People start talking about this “1 月の影響” in mid-December and will argue that any year end price moves are “in anticipation” of the coming rally in January.
One of the popular explanations for the January Effect is that stocks who have benefited from a Santa Claus Rally will begin reporting fourth quarter earnings. This is an interesting theory, しかし as quantitative traders, we would like to see some actual data to back up the argument. Does this supporting data exist?
CXO Advisory Group published an article that pulled all of the data available to analyze the legitimacy of the January Effect. In order to acquire the largest possible sample of data that would still be manageable to work with, they elected to use Robert Shiller’s S&P Composite Stock Index. This index calculates monthly levels of the S&P Composite Index by taking the average of the daily closes during a given month. They were able to use data dating back to 1871, which gave them 143 年 to analyze.
Average Return By Calendar Month
The first chart that the article looks at plots the average return for each individual month over the entire sample period. The chart displays the highest, lowest, and average return for each of the twelve months.
The average return for all of the months in the sample period was 0.43%. The average return for January was 1.57%. January also had the lowest standard deviation of any of the months. この evidence seems to support the January Effect theory, but the authors were interested in digging much deeper.
Average Monthly Return in Subperiods
The next chart they produced showed the sample data split into three equal time periods. The chart plots the monthly returns for the entire data period, the data from 1871 を通じて 1918, the data from 1919 を通じて 1966, and the data from 1967 を通じて 2013.
This chart further supports the idea of the January Effect. January is by far the most consistent performing month. The article also notes that it is the strongest month in two of the three subperiods.
Breaking The Data Down Further
The article continues by breaking the data down into even smaller samples. 今度こそです, the authors measured how the month of January performed relative to the average of all of the months in each data point. They first looked at the data grouped into decades, and then broke it down further into individual years.
Breaking the data down into decades reveals that the performance of January relative to the average of all months has been very poor in recent years. The best-fit line that is drawn on the chart also shows that the impact of the January Effect has been declining over time.
Breaking the data down even further into individual years shows us that there is no rhyme or reason to what happens in any individual year, but it also supports the idea that the January Effect has been decreasing over the course of the data sample.
Changing the Data
The next step that the authors take is to repeat the first two charts using different data. Instead of using the Shiller data, which calculates based on the average daily closing price, they simply used the monthly closing prices of the S&P 500. Using this approach, they were able to collect data dating back to 1950.
S を使用してください。&P 500 データ, the average monthly return for all months is 0.72%. The average monthly return in January is 1.2%, which makes January the fifth best month. この場合, January has the second highest standard deviation in its returns.
The conclusion of the article suggests that while there is some evidence that can be found to support the January Effect, しかし that evidence is likely skewed by results from a long time ago. It would be very risky to blindly trade based on the January Effect in today’s markets.
メッセージを残してください