How Much Alpha Is in Preliminary Data?

Illuminating the Differences between Prelim and Final Filings

“In theory there is no difference between theory and practice. In practice there is.”
-Yogi Berra

Companies often report financials twice: first, through a preliminary press release and again in their official, i.e., final, SEC filings. In theory, there should be no difference between the numbers reported in a company’s preliminary financial filings and their final filings with the SEC. In practice, often significant difference can occur between the preliminary and final filings. In this month’s research report, we focus on these observed differences within the S&P Capital IQ Point-In-Time database in order to ascertain the nature and exploitability of these differences. We find that:

• A simple event-based trading strategy evaluating the differences between Preliminary and Final data for Diluted EPS and Net Income generates an annualized excess return of +22% for long positions and +24% for short positions.
• Incorporating preliminary data into a simple Earnings Yield (EP) strategy improved the factor’s monthly returns by ~0.2% (20 basis points) over a traditional final only EP factor.
• Diluted EPS shows significant differences between Q4 and non-Q4 preliminaries and finals, i.e., Q1, Q2, and Q3. Additionally, we find that certain sectors have significantly more instances of higher preliminary observations.
• Unlike EPS, we find no discernable patterns in Net Income by sector or size when comparing preliminary and final filings.
• Repeat offenders, or companies that repeatedly report large differences between preliminary and final numbers, come in all shapes and sizes with underlying causes of their offenses being fairly independent across companies.
• We find larger companies, by market capitalization, and Utilities are more likely to be repeat offenders vis-à-vis smaller firms and/or other GICS sectors.

Click here to read the rest of this report.