Understanding ROE

 

SUMMING IT UP

If you want to use fundamental analysis effectively, you must have a clear understanding of your tools.  Currently, this can be hard to do when you are looking at ROE figures because of the inconsistencies in the measurement of ROE.

However, if it’s important to express profit margins as “gross margins,” “operating margins,” or “pre-tax margins,” then it’s equally essential to identify ROE as “net return on beginning equity,”  “EACS return on average equity,” or the like.  At least, in that scenario, anyone using those ratios  will be able to assess their shortcomings and make rational analyses and comparisons.

At a minimum, if you are analyzing a company,  you need to understand how the ROE figures weregenerated, and be aware of distortions that can occur depending on the equity figure used.

Use "normalized income after taxes."

If you are calculating ROE on your own, it should be calculated using the most “stripped-down” income available; i.e., “normalized income after taxes,” which excludes all non-recurring and extraordinary items from the income derived from continuing operations. This is not all that easy to  come by. Either you, or your data provider’s analysts must estimate the tax implications for the non-recurring items excluded, since such items appear above the tax line on the income statement. At this time, the only data providers I  know of that offer normalized earnings are Value Line (www.valueline.com) and Market Guide (www.marketguide.com).

As a practical matter, the second best and probably the “most likely to succeed” would be income available to common shares from continuing operations excluding extraordinary items. And the  net income after taxes would be a near next choice.

Best to divide income by beginning equity

The income source, so long as it’s identified, is not the crucial issue however. Far more important is that, whatever income is used, it should be divided by beginning equity in order to exclude any distortions that arise from changes to equity—especially those from the retained earnings. The concept  of return must be simple. Whatever incremental increases in income may result from the use of  that income should be regarded merely as effective utilization of the beginning equity that  produces the return.

At the very least, for those that use average or ending equity, if it’s worth the trouble to consider return on equity at all, then it’s worth the trouble to back out the items that affect changes in equity  during the period, if not the retained earnings.  The use of average equity is not good, but beats  ending equity.

Probably the least useful calculation of ROE uses earnings per share divided by ending book value, because that figure is not only distorted by all of the issues raised above, but it’s also subject to the additional inaccuracies that come from using different values for outstanding shares—e.g., diluted, average, weighted average, shares at the beginning or end of the period—to arrive at the top and  bottom of the equation.

To be useful, ROE must be calculated using components that clarify, rather than distort, the return picture. Or, at the very least, as an analyst you must be aware of those possible distortions by  being aware of the components used in the calculation. Without this awareness, I would suggest that  ROE is simply not what it’s cracked up to be.