Im seeing some varying % Change when using it under different timeframes.  For instance, if you look up AAPL on a daily chart, from 1/2/13 through 6/7/13, it will produce

How is the % Change that appears in the Custom Date Sort dialog box calculated?  For a daily time frame it almost makes sense.  if you change to a weekly then monthly specifying the same date range, it shows different results:

AAPL Close 1/2/13 - 549.03

APPL Close 6/7/13 - 441.81

Daily                Weekly                 Monthly

-19.53               -16.17                    -3.00

Again using the exact same date range.  Daily is what I expect, but the weekly and monthly % chg, does not produce the same result, even though the custom date sort box shows the same date range.

I'm wondering how they are calculated?

Here's some screen shots

Daily:

Weekly:

Monthly:

The calculations in any given time frame are done entirely using the bars in that time frame. Date from other time frames is not considered in those calculations.

So on a Daily Time Frame, you are using the closing values for 1/2/2013 and 6/7/2013:

100 * (441.81 / 549.03 - 1) = -19.53%

On a Weekly Time Frame, you are using the closing values for 1/4/2013 and 6/7/2013:

100 * (441.81 / 527.00 - 1) = -16.17%

And on a Monthly Time Frame, you are using the closing values for 1/31/2013 and 6/7/2013:

100 * (441.81 / 455.49 - 1) = -3.00%

Thank you, after watching the "Finding Top Performers with Custom Date Sort" and "Seasonality Analysis with Custom Date Sort". that reducing your timeframe down to a weekly or monthly timeframe to do a > 2 year study could produce skewed results.  Could I request that they actually show the correct dates I am selecting in the custom date sort dialog?  Also if they could include the number of days between the two dates?

Or allow us to tweak the dates by manually overriding the dates, that way we do not have to flip around in time frames, just give us the % chg for those time periods, and number of days......

Thank you for your suggestion.