Screening out Earnings Manipulators; Quality of Earnings

budget cuts

Accounting shenanigans have a way of snowballing: Once a company moves earnings from one period to another, operating shortfalls that occur thereafter require to engage in further accounting maneuvers that must be even more “heroic.” These can turn fudging into fraud. (More money, it has been noted, has been stolen with the point of a pen than at the point of a gun.)” –Warren Buffett, Shareholder Letter, 2000.

We pick up from the last lesson and read Chapter 3 in Quantitative Value: Hornswoggled! Eliminating Earnings Manipulators and Outright Frauds.

This is an important chapter for improving as an investor. Your goal might be to understand accounting up to the intermediate level so as to adjust accounting principles into economic reality. What story are the numbers telling you?

Think of this chapter as a way to build an early-warning system for companies with weak accounting.

The authors propose three ways to detect aggressive accounting that lead to poor quality of earnings:

  1. Scaled total accruals (STA), which uncovers early-stage earnings manipulations
  2. Scaled net operating assets (SNOA) which captures a management’s historical attempts at earnings manipulation.
  3. The third is the probability of manipulation, or PROBM, a tool that identifies stocks with a high probability of fraud or manipulation.

When the growth in cumulative accruals (net operating income) outstrips the growth in cumulative free cash flow, the balance sheet becomes “bloated.” Stocks with balance sheets bloated in this way find it difficult to sustain earnings growth.  When managements take the low road in aspects that are visible, it is likely they are following a similar path behind the scenes. There is seldom just one cockroach in the kitchen.

A warning sign is high accruals that show much higher income than cash flow. See pages 64 to 68 in Quantitative Value.

Though you should read this chapter carefully and for the nerds, dig into the research papers below, but I highly suggest studying Chapter 8 in Quality of Earnings, Chapter 8 (sent via email to Deep-Value at Google Groups).   A gem of a book. and Defining Earnings Quality CFA Publication

Earnings Management, Fraud Detection and Adjusting for Accruals

How a group of Cornell Students sold Enron before the collapse

Information in Balance Sheets for Future Stock Returns

Earnings Mgt and LR Stock Performance of Reverse LBOs

Earnings Mgt and LR Performance of IPOs 1998

Can Forensic Accounting Predict Stock Returns

Analysts do not adjust adequately for accruals 2004

After a few days of digesting this post, we will tackle Chapter 4: How to Avoid the Sick Men of the Stock Market, in Quantitative Value.

How can we avoid bad news as investors through our knowledge of financial statement analysis and human motivations (incentives)?

Sound Money


A Banker for All Seasons John Exeter

An American Original: Voodoo Child

Have a Great Weekend!


Tawes Hockey

4 responses to “Screening out Earnings Manipulators; Quality of Earnings

  1. BTW, It sounds ironic the price that appears on the cover of Rueff’s book, especially when your book includes the word inflation on it!

  2. Minor typo errors on p67 “floats between 1 percent…,peaks at 5.2 percent the next year.” # should be 3.6%? (in the table)
    GMI = 2.162 (Red Flag) # should be 1.891 (in the table)

    Maybe I missed something but what is the cut off for a red flag? Is there more information elsewhere?

    It might be a good exercise to use these tools to work on an actual topical problem, Noble. Is it or is it not? Rumours…

    I haven’t looked at the academic papers yet, but QV doesn’t go into detail about calculating PMAN. How do you get PMAN from CDF(PROBM)?
    The formula for PROBM is given along with explanation, but how do you derive PMAN from CDF?

    Thank you.

  3. This is a very interesting topic and one which deserves careful study.

    Let me raise two points about the STA test suggested in Quantitative Value.

    1. I think the authors are on the right track when they suggest that a high level of accruals might be a warning sign. Comparing NPAT with CFO is one way to identify high accruals, but I prefer a slightly different approach.

    Typically, I compare reported EBITDA with adjusted CFO. I work in a modified IFRS environment and it is common to see interest income/expense and taxes reported as part of CFO. I adjust the reported CFO figure to exclude all interest and taxes.

    As a general rule of thumb, if adjusted CFO comes in at say 90% of EBITDA, then I tend to treat that as acceptable. If adjusted CFO is a lot less, then I either start asking a lot of questions or I move on to the next company. However, sometimes the differences are due to seasonal factors (eg, wholesalers usually don’t get paid for pre-Christmas sales until January).

    I find this approach easier to apply (and, frankly, easier for my little brain to understand!) While my approach doesn’t pick-up any legerdemain relating to D&A, there are some separate tests I use to deal with that.

    2. One word of caution when comparing P&L and CFO figures: Sometimes differences can arise because of the way that taxes are booked under the applicable accounting standards. This can get quite technical, but please bear with me for a minute.

    In my country, we have a Goods and Services Tax (GST), which is a kind of comprehensive value added tax imposed at the rate of 10% on nearly all transactions (other than most financial transactions). Our accounting standards take a two-faced approach as to how GST is accounted for. Broadly speaking, income and expense items in the P&L are reported net of GST (ie, GST is excluded); whereas income and expense items in the Cash Flow Statement are reported inclusive of GST. (I am not making this up!)

    This means that, all other things being equal, P&L figures will be approx 9% less than the corresponding CFO items (in respect of non-financial transactions). This can sometimes result in material differences between the reported NPAT and CFO figures, even if there is no monkey business going on.

    So don’t apply the authors’ STA test blindly without first getting to grips with how items like sales taxes and value-added taxes are reported.

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.