recent research in information systems using XBRL shows that there are huge differences between what are reported in GAAP financial statements and what Compustat captures even after making adjustments. See some interesting forthcoming research in JIS that exploits the fact that mandated XBRL filings have existed for US public companies since 2011!
Yet an entire mini- industry has grown up around comparing non- GAAP numbers in earnings press releases and GAAP as captured by Computstat! Yet investors are comparing these non- GAAP to real GAAP numbers NOT Computstat GAAP! Indeed most capital markets research is based not on the GAAP numbers investors have but rather what we are now learning are pretty different C- GAAP numbers.
What does all this mean for market efficiency studies? functional fixation studies? accruals?
I have no clue but it seems to me that a lot of things that we think we know about how investors use GAAP numbers need to be revisited given what we have been researching for a long time are C-GAAP numbers which we now have clear evidence are NOT the same thing! Indeed, I expect that the vast majority of accounting studies since the early 1970’s are about c-GAAP not GAAP!
I recall being laughed at when I raised this question 20 years ago and was told that the adults knew better! Maybe they didn’t! Wouldn’t that be funny!