If you're just an average investor, how can you pick up on whether or not a company is cooking its books?
Craig Newmark recently pointed to one of the neater ways in which someone might be able to determine if the numbers they're examining are following a natural pattern as opposed to an artificially-contrived one using Benford's Law, but unless you have access to reams and reams of internal company data, it's pretty unlikely that you as an individual without that kind of access could find out if something shady might be going on.
But for publicly-traded companies, you can get access to a company's publicly-reported financial statements, such as their annual reports or the 10-K statements they file with the United States Securities and Exchange Commission (SEC). And with that information, you can calculate a company's "fraud score" or "F-Score", which can provide a pretty good indication of whether or not the people inside the company might be manipulating their accounting.
Using math originally developed by Patricia M. DeChow, Weili Ge, Chad R. Larson and Richard G. Sloan in 2007 using data from 1982 through 2002, and updated in 2010 to include all years from 1982 through 2005, we've updated our tool for calculating the F-Score for any publicly-traded company for which you can obtain the indicated data here!
For that, you'll need consecutive years worth of the company's annual data - our tool- right here- provides that using Enron's data for the years 1998 (two years prior), 1999 (one year prior) and 2000 (year of interest).
a result greater than a value of 1 indicates a statistically higher than expected likelihood that the numbers the company in question has published have been misstated, which is "accountingese" for suggesting that the company's books may have been cooked! The following guide, developed by the F-score's creators, may be used to interpret the tool's results ("F-Score 1" corresponds to the specific model used in our tool):
We should note that the math is somewhat sensitive - the formula's creators indicate it will produce a high frequency of false positives, which means that an F-Score greater than one should be taken as an indication that an average investor should be much more diligent in reviewing a company's business before making investing decisions related to it.
The tool above provides different results from our original version of the F-Score formula, which was based upon the original 2007 math.
This update to our original tool is the result of a collaborative project with Pasi Havia, who was seeking to implement a Finnish-language version of the tool. We owe our thanks to Pasi for his detective work in finding that the formula for calculating a company's F-score had changed from 2007 and for developing the new and improved code to calculate the F-Score!
If you compare the results between Pasi's version and ours above, you'll find that the results between the two tools are nearly identical - the difference comes down to how the rounding for the value of the mathematical constant e in the formula was done (Pasi rounded it to 8 decimal places for the sake of matching the authors' results in their paper, while we just let it run!)
Related Tools at Political Calculations
- Predicting Bankruptcy
How likely is it that a publicly traded company will declare bankruptcy in the next year? Our tool for calculating the company's Altman Z-Score can answer! It predicted General Motors failure years before the company failed!
Does that publicly-traded company have enough money flowing through its veins to keep operating?