Return on invested capital is one of my very favorite metrics. In this article, I'll show how you can use it to find stocks to buy, stocks to watch, and stocks to avoid, starting with YRC Worldwide (Nasdaq: YRCW).

Beware ROE
You've probably heard of return on equity, a favorite of Warren Buffett. It measures net income (the "return") relative to the equity capital a business has raised and built. A higher ROE signals a more efficient business.

But ROE can be gamed. Because debt is cheaper than equity financing, a management team whose bonuses depend on ROE targets may be tempted to lever up, increasing risk, just to juice net income and ROE.

Return on invested capital -- which is like a return on debt and equity -- catches this. ("RODE" would have been a catchy acronym, no?) To find ROIC, simply divide a company's after-tax operating profit by the sum of its debt and equity. Because it includes debt, ROIC is harder to fudge than ROE. Studies also indicate that watching ROIC can improve your returns.

Why ROIC reigns supreme
Michael Mauboussin -- the chief smart dude at Legg Mason Capital Management -- divided stocks into quintiles by ROIC in 1997, then tracked them through 2006. The lowest 1997 quintile ended up performing worst, unsurprisingly. But the stocks with the highest starting ROIC didn't perform the best, with annual returns of less than 6%, mainly because they fell out of the top quintile along the way.

Two investing secrets emerge from the nuances of Mauboussin's findings:

1. If you find a rising ROIC, you could have a winner.
Companies that started 1997 in the lowest or second-lowest ROIC buckets, but finished 2006 in the highest or second-highest, delivered returns of 14% annually.

2. While a high ROIC alone doesn't help, consistently high ROIC is a marker of outperformance.
Companies that started in the No. 1 or No. 2 quintile in 1997, and remained there through 2006, delivered a whopping 11% annually.

Will our next contestant come on down?
Let's see how YRC Worldwide, the big-rig owner of Yellow and Roadway trucking companies, stacks up by this measurement. We'll be using numbers from Capital IQ (a division of Standard & Poor's). For most moderate-risk companies, I consider anything greater than 9% to be a decent ROIC; more than 12% is even better. The higher the risk, the higher the ROIC you'll need to be content. Based loosely on Mauboussin's numbers, we'll also assume that annual standard deviations of less than 10% mark the beginning of "consistent" territory.

ROIC for YRC Worldwide













What can we conclude? The trend is clear; we don't need to worry about standard deviation. YRC Worldwide is doing the exact opposite of what Mauboussin's screen would want. And with its shares down nearly 99.5% -- no, really! -- since 2005, investors who took warning from the falling ROIC would have saved much of their wealth.

For some industry perspective, notice that while Con-Way's (NYSE: CNW) ROICs fell during that same time period (though not to the extent YRC's figures fell), fellow competitor J.B. Hunt (Nasdaq: JBHT), which owns the vast majority of its trucks rather than using owner/operator trucks, posted ROICs during the same time period that held relatively steady in the 13% to 26% range since 2005. Heartland Express (NYSE: HTLD), which employs more owner/operators, delivered similarly solid ROIC performance, just a few percentage points lower. And just for educational comparison, Expeditors International (Nasdaq: EXPD), a debt-free export/import business with a slightly different model within the same overall industry, sports a five-year historical ROIC ranging from 16% to 23% -- closely matching its ROE, given its debt-free status.

Bottom line for YRC? Given its slight improvement recently, and the rich rewards associated with rising ROIC, it's possible that YRC could actually improve from here. But I'm not betting on it.

James Early owns no stocks mentioned in this article. You can investigate his Motley Fool Income Investor newsletter free for 30 days. The Motley Fool has a disclosure policy.