Accounting practices cloud P/E picture

Benj Gallander and Ben Stadelmann
Wednesday, October 14, 2009

Has the market rebounded too far, too fast from its ignoble lows of last March? One of the rallying cries of those who believe there is more good stuff to come is that equities are inexpensive by the standards of historical valuation.

For example, Globe and Mail columnist David Parkinson reported on a study by National Bank that found that the price–earnings ratio for the S&P 500 was currently in similary territory to that which it occupied at the end of past recessions. In the article, the ratio was indicated to be sitting at 14.4, well over the ratio of 7.6 marked in 1980, but not far above the average of 12.7, taking into account all of the recessions since 1973.

A current P/E of 14.4 doesn’t jibe with the implosion of corporate profitability over the past year. That contradiction is explained by this “current” number being the average that analysts have forecast for the coming 12 months. The actual P/E, as S&P defines it, is for the trailing 12 months, and that comes in at 140.8.

That juxtaposition raises the question: Did analysts foresee a P/E of anything like 140 a year ago? The answer, as you might guess, is no. If fact, the number proffered 12 months ago was around 10. The yawning chasm between prediction and reality is largely attributable to titanic losses chalked up by US corporations, especially in the financial sector.

The bizarre part is, from the point of view of analysts, most of these losses are irrelevant, because they are “extraordinary items.” That’s because analysts use “operating earnings,” which exclude all manner of write-offs and one-time charges.

What we call real earnings, as defined by GAAP, S&P prefers to label “reported” earnings. In an Orwellian twist, reported earnings get short shrift — and barely get reported at all. It’s operating earnings, with all the ugly stuff removed, that grab the headlines.

This semantic twist encourages corporations to engage in big-bath accounting. Given the severity of the recession, where losing money is the norm anyway, it’s the ideal time to take the axe to goodwill and other intangible assets. This trick isn’t new; companies have used the technique to grease the forward path for decades. What has changed are the scale and perniciousness of the practice, so much so that it is making a mockery of the whole concept of the price/earnings ratio.

This is made clear by an examination of the historical data for the S&P 500, which is helpfully available in
Excel-spreadsheet form from the Standard and Poor’s website. Taking the data for the past 20 years and dividing it into five-year chunks shows that, for the period 1990–1994, the average difference between the operating P/E and reported P/E was 2.4. For 1995–1999 it was a similar 2.3. But from 2000–2004, this gap jumped to 6.8. And from 2005–2009, it is a broad 20.8.

Would the real P/E please stand up?

This is not to say that the calculation of operating earnings is worthless. When conducting a financial analysis, it is a useful metric, just like EBITDA, or “cash cost” of production when looking at a gold mine. But these tools should never be confused with real, bottom line profitability.

Looking ahead to 2010, analysts currently project an average operating P/E of 16. Economists at S&P reckon that reported P/E for next year will be 24.3. That’s a large disparity, especially considering that the big-bath write-offs should be mostly over by then.

One thing is for sure: in order for the present rally in stock prices to be justified, corporate profits in America need to perform a V-shaped rebound next year. And when we say profit, it means the real, unvarnished truth.