How to calculate Return on Equity?

[Halebop]

Dupont Formula isn't much in vogue but I think the Return on Equity version is easier to work with if expressed as:

(Net Income / Sales) * (Sales / Total Assets) * (Total Assets / Avg Equity)

So based on your numbers above (Sales I think are $5.562b) but adjusting out extraordinary movements in earnings (Earnings back down to $955m)...

=(955/5562) * (5562m/8176) * (8276/3604)
=0.1717 * 0.6803 * 2.296
=0.2681
=26.81%

Personally, I'd just go for ROE as Net Income / Avg Equity (or more simply End of Year Equity)...

= (955/3604)
=0.265
=26.5%

In this example they are virtually the same in any case. The DuPont measure is injecting a qualitative stance by supposing that high sales margins and high return on assets are attributes that should go hand in hand with high return on equity. Probably works for most arguments but more difficult to apply to a Bank on the asset measure or a large format discount retailers on the sales margin one - you'd want to be comparing against peers rather than other industries and that might make you lose sight of return on equity if other industries provide better prospects.

[steve fleming]

Dupont Formula isn't much in vogue but I think the Return on Equity version is easier to work with if expressed as:

(Net Income / Sales) * (Sales / Total Assets) * (Total Assets / Avg Equity)

So based on your numbers above (Sales I think are $5.562b) but adjusting out extraordinary movements in earnings (Earnings back down to $955m)...

=(955/5562) * (5562m/8176) * (8276/3604)
=0.1717 * 0.6803 * 2.296
=0.2681
=26.81%

Personally, I'd just go for ROE as Net Income / Avg Equity (or more simply End of Year Equity)...

Halebop,

The two formulae you quoted are exactly the same??

The sales & total assets being both a denominator & numerator simply cancel out?

In your example total assets are calculated as $8,176b and then $8,276b. What is the $100m adjustment?

Cheers

[Halebop]

Oops. The $100m was a typo in my spreadsheet transposed to post. Only excuse I can offer is that I was eating squid and typing with one hand.

Du Pont is not intended as a ROE measure, the last part of the equation turns it into an ROE measure but it gets you back to where you should be anyway If the user stuck with a simple ROE calc.

Du Pont is really a qualitative measure used to weed out low margin, high Capex businesses or identify high margin / low capex stars. I suspect without intending to DuPont also created a benchmarking tool because to my mind it is really most effective when measuring peers rather than a grab-bag of targets.

Du Pont alone (without trying to get back to ROE) is (Profit / Sales) * (Sales / net Assets). So I'm not sure if it tells you much of anything useful to compare a low margin / low asset supermaket with a high margin / high asset property company for instance. Particularly on a trending series, I think it is more useful to compare a basket of super market operators or a basket of property companies.

It's one measure though and I suspect not popular for some good reasons. Not least of which is you can't tell what you are seeing from the results - did Assets or Margins contribute most to the outcome? Buffett's simple tenets are much easier by keeping them on separate lines.