Nelson Peltz, chief executive officer of Trian Fund Management LP, speaks during the WSJDLive Global Technology Conference in Laguna Beach, California, U.S., on Tuesday, Oct. 25, 2016.
Patrick T. Fallon/Bloomberg/Getty Images

Its attack on Nelson Peltz doesn't add up.

By Shawn Tully
September 22, 2017

To bolster its attack on Nelson Peltz, P&G has unveiled a 20-page analysis alleging that companies show little or no benefit from having Peltz’s Trian as an investor. Peltz, who’s mounted a proxy battle seeking a board seat at P&G’s October 10 annual meeting, has fired back. P&G’s newest broadside, charges Trian, is riddled with “material factual errors, data errors, calculation errors and misrepresentations of Trian’s and Nelson Peltz’s track records.”

While assessing P&G’s entire critique would be a lengthy undertaking, this writer was intrigued by one of the major points in dispute. P&G claims that this comparison with prominent rivals demonstrates Trian’s mediocre performance, and Trian counters that the P&G’s calculations are marred by glaring errors, and show no such thing.

After checking the numbers, my conclusion is that Trian is basically correct. P&G needs to work on its math.

On page 83 of its web presentation defending management’s performance and criticizing Peltz’s fitness as a director, P&G displays a bar chart comparing the record of Trian holding Mondelez to rivals Hershey and Nestle from 2013 to 2016, based on “Adjusted Operating Margin.” That measure is apparently EBIT, or earnings before interest and taxes, because the 15.3% margin shown for Mondelez, and its 470 basis point improvement over the past three years, are the EBIT numbers reported in its 10K.

According to the chart, Mondelez trailed Hershey in raising its margins, 590 bps to 470 bps, and barely beat Nestle’s 440 bps improvement. It also shows that Mondelez’s 2016 margins of 15.3% stand far below Hershey’s 24.4% and Nestle’s 19.5%.

In its rebuttal, Trian states that a basic error skews the comparison in favor of Hershey and Nestle. P&G, says Trian, is using EBIT for the earlier periods, and EBITDA for 2016 figures. Since EBITDA adds depreciation and amortization back to earnings, it gives a big boost to last year’s margins, and hugely inflates, and misstates, the improvement from 2013.

To see who’s right, I ran the numbers. Let’s start with Nestle. P&G states that it’s excluding coffee from its calculations, apparently because Mondelez recently sold its own coffee franchise. From Nestle’s 10Ks, I was unable to determine coffee’s precise contribution to sales and profits; however, even including coffee, the overall numbers demonstrate P&G’s faulty math. In 2013, Nestle posted an EBIT margin of 15.27%. In 2016, P&G claims that Nestle’s margins showed an improvement of over 400 bps, rising to 19.5%. But it appears that Nestle is using EBIT for 2013, and EBITDA for 2016. By my calculations, it’s only by adding back depreciation and amortization, raising EBIT by almost one-quarter, that Nestle’s record “improves” at all.

Trian claims that if EBIT in 2013 is compared with EBIT in 2016, Nestle added only 100 bps to its margins. That’s what my calculations show as well.

For Hershey, P&G’s numbers are even more misleading. They show EBIT for 2013 at 18.5%; that’s actually the figure for 2012. The correct number for 2013 is 19.2% As with Nestle, P&G the switches to EBIDA for 2016. So Hershey’s improvement is inflated twice, first by starting in the wrong year with a lower base, and second by ddding amortization and depreciation for 2016, while excluding them for 2013. Those factors combined boost Hershey’s gains by a factor of five, from the actual 115 basis points to a huge 590 basis points.

Trian’s position that Mondelez raised EBIT a lot more than either Hershey or Nestle appears correct. Comparing EBIT in one year to EBITDA in another is like comparing grapes to avocados. As I’ve previously stated, P&G’s weak defense of its own record is the best argument for putting Peltz on its board. These new, bogus numbers only strengthen the case for Trian.


You May Like