The notion that earnings drive stock prices became conventional wisdom for very good reason: Over the long run, it's true. Over the short term, however, fear and greed play a big part in setting prices, too. That terrible twosome often prompts companies to massage or manipulate their numbers in hopes of currying the market's favor. When it comes to earnings, my motto is "trust, but verify."

Check the cash
As fellow Fool Sham Gad has pointed out: "Creative accounting decisions, such as changing depreciation schedules or including non-recurring gains at certain points in time, can manipulate [a company's per-share net income]. We all know that management has a considerable incentive to meet earnings expectations, so you should always examine earnings with a healthy dose of skepticism."

One of the best ways to verify earnings is to compare them to a company's cash flow. Earnings and cash flow are two different ways of measuring the same thing: the money a company earns and spends. Over the long term, they should be roughly equal. The difference is in the timing.

A key concept in accounting is the "matching principle," meaning that associated revenues and expenses should be "matched," or recognized at the same time. For example, on its income statement, a company records an inventory expense (cost of goods sold) not when it buys the inventory, but when it sells it -- matching the cost with the revenue generated.

The statement of cash flows takes the opposite approach, recognizing expenses when the cash is paid. That's exactly what makes it a good way to verify the reported earnings number.

Timing is everything
Since the two financial statements differ mainly in the timing of entries, big changes in one measure often predict similar changes in the other. The trick is figuring out which measure is the leading indicator in a given case. Let's look at a couple of examples.

In 2007, Amazon.com (NASDAQ:AMZN) reported $476 million in net income, up from $190 million in 2006. Cash flow from operations went up by a larger amount -- rising to $1.4 billion from $700 million -- but the magnitude of the change was similar. When I look at the details, I see that most items roughly doubled in size -- except inventory, which grew at a much slower rate.

That tells me a couple of things. First, Amazon was able to control inventory, shipping its products more efficiently. That's a good thing, and it bodes well for future earnings and cash flow. Second, the inventory growth didn't match up with accounts payable growth. Since most of the payables are likely due to suppliers of inventory, I'd expect the two metrics to move back in line over time. In short, while its cash flow in 2007 was a bit higher than "normal," earnings and cash flow both appear to confirm that Amazon became significantly more profitable in 2007.

Fuel Tech (NASDAQ:FTEK) tells a different story. Its net income also grew in 2007, increasing to $7.2 million from $6.8 million the prior year. Meanwhile, its cash flow from operating activity fell from $8.2 million to $4.1 million. A quick glance at the cash flow statement reveals the culprit: Accounts receivable grew by $15 million. This figure represents money already recognized as sales on the income statement, but not yet collected from customers.

Looking further into the 10-K, I find this note: "Accounts receivable includes unbilled receivables, representing costs and estimated earnings in excess of billings on uncompleted contracts under the percentage of completion method. At December 31, 2007 and 2006, unbilled receivables were approximately $16,813,000 and $3,615,000, respectively."

Unbilled receivables arise because Fuel Tech uses the "percentage of completion" method to recognize sales related to long-term contracts. Under this method, the company estimates how much of the work has been completed, then recognizes associated revenue and expense. That's a bit too much estimating for my taste. If the company had reported the sales at the time cash was collected, rather than going by its estimates, year-over-year revenue and earnings would both have fallen in 2007.

Does the decline in cash flow mean that Fuel Tech's earnings will fall in the future, or does the rise in earnings signal that cash flows will also increase? It depends on whether the company ever collects on those receivables. Its clients, "utility and industrial customers worldwide," probably won't skip town, so perhaps the company will indeed collect. On the other hand, if the work really is as far along as estimated, why haven't the customers been billed?

According to the company, "Fuel Tech's construction contracts are typically six to twelve months in length. A typical contract will have three or four critical milestones that serve as the basis for Fuel Tech to invoice the customer." The company's rise in unbilled receivables could suggest that such milestones are not being met.

How is an investor supposed to know the difference? Sometimes the company will explain the change in the 10-K or on a conference call. Other times, it won't. In those cases, "caveat emptor" may be your best advice.

Further Foolishness: