Not to belabor the point, but we should again warn viewers that big gains or losses in our charts don't necessarily have any correlation with future performance.  We're not saying this only as the standard "anti-nincompoop" spiel, but because we've found that strong historical gains in a particular period can actually be a contrary indicator, depending on the month or quarter that you're looking at.  For more information on this issue, take a look at our monthly and quarterly tables of best long and short indicators.

Our procedure can result in slightly exaggerated or dampened gains or losses over the duration of the period examined.  Don't look at the y-axis as a precise indication of average percentage gains gotten over history.

To fool the plotting software into plotting correctly, we needed to have dates in a particular format (i.e. we couldn't output "1rst day of year").  If you see "1/27/03", just mentally substitute "sometime in late January".

We toss massive (greater than 20%)  one day gains and losses from our data, considering them to be anomalies.  In the case of individual stocks (as opposed to indices), the data is not adjusted for dividends. 

It seems that at least one other web site has defined "quarterly" and "monthly" in a different way than we do.  To be perfectly clear:  for our purposes, a "quarterly" chart is generated by averaging the percentage gains and losses for the first day of every quarter for which we have data, and then proceeding to the second day, third day (etc.), until we come to the end of the quarter.  The first day of a quarter would be the first trading day in January, April, July, and October.  The same thinking applies to monthly charts.  The idea is to simply try to determine whether a particular stock exhibits repetitive behavior with respect to quarters or months.

One final caveat that users should be aware of...we assume 251 trading days in a year, 62 days in a quarter, and 20 days in a month.  But, of course, not every month has 20 trading days.  In fact, months can have as few as 18 trading days, and as many as 23.  Every month, however, has a first trading day, so you can expect that the first day's average will be accurate over the course of history.  But when we get to the 19th or 20th day of a month, we have less data to work with...accuracy declines.  The main point...be particularly careful about making inferences based on the tail end of our charts.

Copyright © 2010 MarketSynopsis.com. All rights reserved.