Quote from Fierze:
Thank you very much circadian! Are the dates in the sheet correct? In that case i'll rearrange them so they fit together. It's a bit strange though, some dates only show advancers and no decliners etc...
Yeah, unfortunately, my data provider's NYSE a/d data sucks if it's more than 1 year back on the hourly.
Here's what you do, it'll take you about 20 minutes to clean up the data, but it's pretty simple, and user friendly:
1). delete the first month of adv data, to set it equal to the decln data that starts on 7/10/2006
2)create an if/and function that is true if the date and the time are correct. Here's one that will work, just copy it, and paste it into cell O3:
=IF(AND(B3=I3, H3=A3), "true", "xxxx")
Copy it down the entire O column. When you see the xxxx, that means that the time and date of the ADV data is not the same as the DECL data. Then you delete the erroneous data, and update the if/and function to reflect the change. Move on to the next.
3) from there, you can create your ADVN/DECL ratio by dividing column F by column M. Then create yourself a longer period moving average (or a linear regression), and some standard errors and deviations. When the A/D ratio is >1.5-2X deviations from the trend (MA or LR), mind it.
Good luck.