How to Detect Financial Statement Fraud

It’s a common occurrence. I’m on the hunt for value opportunities. I locate some small-cap company that appears to be a steal. I analyze it, spreadsheet it, fall in love with it, but am still nervous about pulling the trigger. Because despite all my hard work there is that… hintergedanke, as the Germans call it… that says:

How do I know these guys aren’t Enron?

Seriously. I’ve often never heard of the company. I don’t know the management. How do I know the financial statements I’m basing my analysis on aren’t just a pack of lies?

I wish I could tell you that I came up with a way to detect financial statement fraud myself – or that I learned some secret sauce at CFA Level 3 – but once again, it’s Bruce Greenwald to the rescue! I think you’ll really like his approach.

The Master
The Master

Step #1. Who are the auditors?
The auditing firm is our first line of defense, but they can fail us in a couple ways. One is if they don’t detect the company’s falsehoods. Another is if they are helping the company lie!

So first we just examine who they are. Is it a one-man shop you’ve never heard of in a strip mall in suburbia? That was Madoff. Game theory says there may be incentives for both to “cooperate” unbeknownst to us in these situations.

Step #2. Who is the company?
If the auditing firm is reputable, there are still scenarios where it might risk its reputation for the company under audit, or just get sloppy. Ask yourself, when would a company be able to go to its auditors, tell them “I think you should look at things this way”, and have the auditors say “OK, show us!”

That would be when you’re Company of the Year, or a rising star in The New Economy, or paying large auditing fees. If that’s the case, red flag again. That was Enron.

Step #3. Where did the money go?
If we’ve made it this far, there’s still the possibility that the company is fooling both the auditors and general public. How to detect? Greenwald continually hammers home the fact that when doing one’s analysis:

Good Information + Bad Information = Bad Information

Usually when doing his signal + noise = noise spiel, he’s describing the stupidity of valuing a company via DCF. Time-value-of-money has you adding cash flows you can reliably estimate (this year, next year) with others that are basically blind guesses (years 5 to… infinity). Net result? A valuation that might be off by a factor of 2. Or more. Lots of luck!

Here, he advises us to focus on the “good information” to detect fraud. He means, examine specific line items of financial statements that are hard to fake, to see what actually happens to all this money (net income) they’re telling us they’re making.

We need only 4 additional items:

    1. Cash minus debt (balance sheet)
    2. Net buybacks (cash flow statement)
    3. Dividends paid (cash flow statement)
    4. Sales (income statement)

Cash and debt are probably the easiest items for an auditor to verify. And you obviously can’t fake a dividend. You can play some games with sales by jiggering accounts receivable, but you shouldn’t be able to fake large increases in sales over an extended period of time.

And it’s so dirt simple even I could have thought of this. But I didn’t.

1. Sum up all the money (net income) they say they've made
2. Subtract the change in net cash in the bank
3. Subtract what they spent on dividends and buybacks
4. The residue is what was put back into the business

And was it a good reinvestment? Well,

5. See how many $ in new sales they got for each $ put back in.

Ice… cold… logic.

I’ve included an example below for an unnamed small cap with about 1000 employees. You’ll see that they’ve actually paid out slightly more in dividends than the net income they’ve reported over the last 10 years!

Fraud Detection Example

How is that possible? This is one of those companies where FCF is usually slightly larger than net income. Maybe their depreciation is a little overstated. Fine.

This company’s sales are increasing at about the inflation rate. Management either can’t or won’t do more to set sales on fire by retaining earnings. But at least they don’t appear to be stealing. What if they were retaining everything but sales were actually falling? That was Worldcom.

In reality, if you do the exercise on a company and see little to show for a lot retained, you will probably never know that it’s outright due to fraud. But you’ll at least conclude that management is poor at capital allocation, and that will be enough to keep you away.

Leave a Reply

Your email address will not be published.