# How to Detect Earnings Manipulation Using the Beneish M Score

Pick the best value stocks with our Stock Ranks, screening and valuation tool. Try the live demo today.

Written by

Jae Jun

To get this kind of information and other exclusive articles before regular readers, get on the VIP Mailing List today.

********

What You Will Learn

• How to detect earnings manipulation using Beneish M Score
• How to use the 5 variables of Beneish M Score for flexible valuation

### Introducing the Beneish M Score

Calculate the Beneish M Score | Flickr: photosteve101

#### What is the Beneish M Score?

I was recently introduced to an article by a reader on the M Score and have found it quite interesting and wanted to share it with you as well. I’ve summarized and edited parts of the original article.

The Beneish M score was created by Professor Messod Beneish. In many ways it is similar to the Altman Z score, but optimized to detect earnings manipulation rather than bankruptcy. This is the link to the original M score for earnings manipulation paper.

Beneish used all the companies in the Compustat database between 1982-1992.

#### The Beneish M Score Variables

The M score is based on a combination of the following eight different indices:

DSRI = Days’ Sales in Receivables Index

• Measured as the ratio of days’ sales in receivables in year t to year t-1. A large increase in DSR could be indicative of revenue inflation.

GMI = Gross Margin Index

• Measured as the ratio of gross margin in year t-1 to gross margin in year t.
• Gross margin has deteriorated when this index is above 1. A firm with poorer prospects is more likely to manipulate earnings.

AQI = Asset Quality Index

• Asset quality is measured as the ratio of non-current assets other than plant, property and equipment to total assets.
• AQI is the ratio of asset quality in year t to year t-1.

SGI = Sales Growth Index

• Ratio of sales in year t to sales in year t-1.
• Sales growth is not itself a measure of manipulation. However, growth companies are likely to find themselves under pressure to manipulate in order to keep up appearances.

DEPI = Depreciation Index

• Measured as the ratio of the rate of depreciation in year t-1 to the corresponding rate in year t.
• DEPI greater than 1 indicates that assets are being depreciated at a slower rate. This suggests that the firm might be revising useful asset life assumptions upwards, or adopting a new method that is income friendly.

SGAI = Sales, General and Administrative expenses Index

• The ratio of SGA expenses in year t relative to year t -1.

LVGI = Leverage Index

• The ratio of total debt to total assets in year t relative to yeat t-1.
• An LVGI >1 indicates an increase in leverage

TATA – Total Accruals to Total Assets

• Total accruals calculated as the change in working capital accounts other than cash less depreciation.

But before I get into the analysis, click on the image below to get the investing red flags that will help you detect potential blow ups. You’ll also get exclusive content and resources that we don’t share anywhere else.

#### The Beneish M Score Formula

The eight variables are then weighted together according to the following:

M = -4.84 + 0.92*DSRI + 0.528*GMI + 0.404*AQI + 0.892*SGI + 0.115*DEPI – 0.172*SGAI + 4.679*TATA – 0.327*LVGI

A score greater than -1.78 indicates a strong likelihood of a firm being a manipulator. In his out of sample tests, Beneish found that he could correctly identify 76% of manipulators, whilst only incorrectly identifying 17.5% of non-manipulators.

#### The 5 Variable Version of the Beneish Model

The five variable version excludes SGAI, DEPI and LEVI which were not significant in the original Beneish model.

M  = -6.065 + 0.823*DSRI + 0.906*GMI + 0.593*AQI + 0.717*SGI + 0.107*DEPI

#### Using the Beneish Score to Select Stocks

In 2008, Beneish goes into more detail in another paper that he published titled “Identifying Overvalued Equity” which seeks to use the M score to select stocks.

Beneish examines portfolio deciles based around his M score over the period 1993-2003 with annual rebalancing done four months after the financial year end.

The results produce 14% for the 8 variable model and 14.8% for the 5 variable M score version where the top M score stocks were held long while the lowest M score stocks were shorted.

This article wouldn’t be complete without a supporting spreadsheet of course.

You can download a spreadsheet that will automatically pull in the required financial data to calculate the Beneish score for the prior 2 years and TTM. Both the 5 variable and 8 variable version of the M score is included.

You will receive a list of all the other free spreadsheets, the excel add-in from the SMF group required to download data from the internet as well as all the installation help material.

And that is how you calculate the Beneish M Score.

Feel free to check out the free version and then when ready, go to the stock valuation software page and review what you will get with the premium.

The premium version includes several valuation models as well as fundamental analysis data, historical data, charts and competitor comparison features. Just by entering one ticker, you can immediately get all that information on your favorite stock which will save you hours in your analysis.

Go now and see for yourself why people rave about the spreadsheets.

To get this kind of information and other exclusive articles before regular readers, get on the VIP Mailing List today.

********

### What is Old School Value?

Old School Value is a suite of value investing tools designed to fatten your portfolio by identifying what stocks to buy and sell.

It is a stock grader, value screener, and valuation tools for the busy investor designed to help you pick stocks 4x faster.

Check out the live preview of AMZN, MSFT, BAC, AAPL and FB.

### 19 responses to “How to Detect Earnings Manipulation Using the Beneish M Score”

1. Charlie says:

Hey Jae
When you say a score greater then -2.22 you mean more negative right?

2. Graham Jervis says:

i doubt it, i think he means more positive, MSFT has a score of -3.14 and -2.95, i doubt they are manipulating their numbers.

3. Jae Jun says:

@ Charlie,

By greater than -2.22, a score of -2.21, -1.3 or 3.2 show that there is a high chance of earnings being manipulated.

4. kubrick says:

i can’t get my M score sheet to work. i have the 10 yr and 5 yr spreadsheets which i paid for and the altman z and they all work excpet for this one. when i open the sheet it gives me the option to enable the content and i said yes but still doesnt work

5. Ranajit says:

Good article Jae. You can try the Beneish model on Satyam Computers Limited(SAY)..now Mahindra Satyam..whose ex-CEO admitted to major accounting fraud in 2009. The M model seems to have raised a red flag every alternate year from 2002. Check it out.

Can you think of a few more examples of such companies so that we can try them out?

6. Fraser says:

Hi,
In the post it says “The five variable version excludes SGAI, DEPI and LEVI which…”. However, the five variable formula includes DEPI but does not includes TATA.
I assume this is a typo? If so, is the typo in the comment or the formula?
Thanks.

7. Jae Jun says:

@ Fraser,
Yes that is a typo. The 5 variable version excludes TATA.

8. dodge1664 says:

Hi Jae,

good effort with the spreadsheet, but I’m struggling to figure out how the total accruals inside TATA is calculated. Could you spell that out in more detail?

Roger

9. Jae Jun says:

@ dodge1664,

Total Accruals/ Total Assets (TATA) is calculated as
(Net Income – CFO)/Total Assets

10. Dodge1664 says:

@Jae Jun

Thanks for the quick reply! That seems to be the formula Beneish used in his 2005 paper \The Relation between Accruals and the Probability of Earnings Manipulation\, however his 1999 paper \The Detection of Earnings Manipulation\ seems to use a different one involving current assets, current liabilities, income tax due and depreciation. Other authors on the web refer to your formula as \net operating accurals\ rather than \total accruals\, e.g.

http://www.investopedia.com/university/accounting-earnings-quality/earnings9.asp

I also found this formula for total accruals which also subtracts
Capex and dividends:
http://www.portfolio123.com/blog.jsp?postid=53&topic=idea

I’m pretty confused, but I guess I’ll go with your suggestion, thanks!

11. Jae Jun says:

Thanks for those links. I’ll look at it and see how easy it is to calculate.
I don’t believe the number has to be so precise, but the original formula should suffice quite fine.

12. BillyBobJoe says:

I got the download link, but dont see the Mfactor benoit spread sheet. Is it in one of the others somewhere. I am hoping to use it as another filter in my options trading methodology.

Thanks, BBJ

PS have followed Randy Harmelink’s stuff for years and also created my own versions

13. BillyBobJoe says:

Beneish M Score spreadsheet, sorry.

BBJ

14. Jovica says:

I’ve tried M score from your workbook for MSFT,…didn’t check all the numbers … I’ve noticed that COGS is 59.160 for 2012 (in your sheet) but the number is actually 17.530? Please check if you can and thank you for these great sheets.

15. good catch. I will have to reupload a fixed version.

16. Arco Verolme MBA says:

Software is very good and also transparant about the calculations. One point is missing when you put in your own statements on StatementsData . There are no formulas. Also the Cash flow statement is not calculated when you type the data of Balance sheet and Profit and Loss statements in case you dont have the cash flow statement.

17. RPC says:

Hi Jae
Unable to download spreadsheets when I put my email address in yellow pop-up on the right hand side on this page. I get a Message this mailing list is not active.
Pls look into it
Thanks

18. should work now. Thanks

19. ucicelos says:

Hi JJ this thread is pretty old but for the 5 variable model, every where on the web I find the same definition as yours while the text says “The five variable version excludes SGAI, DEPI and LEVI which were not significant in the original Beneish model.” but in the actual formula DEPI is still there and TATA is gone…..is the statement correct or the formula??