With the beginning of October a couple of days away, here’s our seasonal analysis for the month. We’ve also embedded a table with the scan results that will make it easier to see the results.
A refresher on this scan and some notes about changes.
We have noted on several occasions that we are using 7 years of historical data in the Comparison charts and that is true, however the scan actually looks at 8 years, so in future charts we’ll add the eighth year.
Our study looks at 8 years of historical data and looks at the returns for all optionable stocks 2006 to 2013.
We filter to find two sets of criteria
– Stocks with gains in all 8 years during October
– Stocks with losses in all 8 years in October
We do make an assumption that the month is 21 trading days and work our way back from the last day of the month. If the last day of the month falls on a weekend, then we use the first trading day prior to that date.
We make no assumptions for drawdown, nor do we look at the fundamentals behind such a pattern. We do compare the stock to the market during the same period and look at the average SPY gain/loss vs. the average stock gain/loss. This helps filter out market influence. We are now including group information to identify particular segments that might display a seasonal bias.
Finally we look at the median gain/loss and look for statistical anomalies, like meteoric gains/loss in one year.
WLT – Walter Energy seasonality through October 2013 – 7 year average in black
SWN – Swan Energy seasonality through October 2013
DV – Devry seasonality through October 2013