Estimates are still handled independently of fundamentals. So things won’t always coincide.

The biggest thing that I have to get my hear wrapped around is the how to deal with the difference between the company fiscal year and the delay in quarterly reports.

I have a question for you. Are the analysts’ estimates synced with the annual reports? Can I assume that current and next years sales rolls over at the time the annual report is published? Or is this a bit of a gray area? I know that P123 was struggling with this a bit I think before you arrived and with previous data vendor. They were blending the current and next year for some applications.

The major flaw in my study is that I’m not dealing with stocks that have disappeared through M&A activity or bankruptcy. So there’s clear survivorship bias, which I simply didn’t know how to avoid.

The chance of look-ahead bias comes in because sales growth is measured as trailing twelve-month earnings divided by previous twelve-month earnings. If I had not incorporated a six-month delay, there would have been a chance that companies with late statements would have already begun their TTM growth by the time I measured the factors. I wanted to err on the safe side. A six-month delay is probably unnecessary, but I think at least a three-month delay would be a good idea. Your suggested approach–blending quarterly sales depending on how far into the quarter the company is–may also avoid look-ahead bias if it’s done carefully.

My desire is to predict sales 52 weeks ahead. Companies will generally be somewhere in the middle of a quarter, so I intend to blend the quarterly sales, based on how far into the quarter the company is. This is an attempt to harmonize stocks that have different year-end dates.