Sohu.com Beneish M-Score (Annual)

About Beneish M Score

The Beneish M Score helps to uncover companies who are likely to be manipulating their reported earnings. Companies with a higher score are more likely to be manipulators. This is a probabilistic model, so it will not detect manipulators with 100% accuracy.

The best cut-off point depends on the costs mistakenly classifying in one of two ways:

1) Classifying firm that is manipulating earnings as a non-manipulator (Type I error), and
2) Classifying a firm as a manipulator when it actually was not manipulating (Type II Error).

Here are optimal cut-offs according to Beneish, presented as the score followed by the cost of Type I error relative to cost of Type II error):

M Score HTML Table:

Score	Relative Error Costs (Type I:Type II)
M Score > -1.49	(10:1)
M Score > -1.78	(20:1)
M Score > -1.89	(40+:1)

Beneish excluded financial institutions from his sample when calculating the M-Score, so extreme care should be taken when looking at M-Scores of financial firms - their business models are different from the manufacturing and other service firms that Beneish used in his study.

If you want more details, here is the original Beneish M Score Paper, or you can learn about our calculation by clicking "Learn More" below.

Learn More