With all that has been written about accounting scandals recently, there has been surprisingly little in the way of concrete proposals to bridge the gap between strict generally accepted accounting principles (GAAP) and the creative art form commonly referred to as pro forma accounting. It’s a topic that we confront frequently, as many of the companies we own are turnaround plays, a process that by its very nature is rife with special accounting charges.

To make this a little clearer, we urge the adoption of guidelines for BBS accounting: Before Bad Stuff. First, it should be said that there’s nothing inherently wrong with companies providing pro forma or adjusted earnings.

True, one might argue that many so called “extraordinary” items seem like fairly run-of-the-mill business expenses, and when a “non-recurring” charge occurs for the fifth time, you have to wonder. But it is a useful exercise for corporations to break out specific charges, and show what effect these have on the bottom line. These hypothetical scenarios do offer a tantalizing look at what the future might bring. It’s up to investors to decide how probable they are.

Where this game of “pretend” and reality becomes particularly confusing is with the business of analysts’ forecast earnings. What kind of earnings, real or BBS? It turns out that analysts are nearly always playing the pro forma game, predicting earnings unimpeded by pesky bad stuff.

Those who provide this information in the media also have to take part of the blame. Log on to your trusty financial Web site such as Yahoo, Quicken, Money Central or Globeinvestor.com and the number provided for the current quarter’s earnings is the correct GAAP one. The calculation shown for the price/earnings ratio, a key benchmark, also uses this number and is thus also accurate.

However, a strange thing happens when you go into the earnings estimates section of these sites. Along with the consensus for future quarters, there is a section on the earnings history where the forecasts are ostensibly compared to the actual numbers. It is the comparison of these numbers that generates the notion of positive or negative earnings “surprises.” What is really surprising is that the column labelled “actual” is nothing of the sort and does not match the number reported for the quarter elsewhere.

Confused? Let’s take an example. Office supply retailer OfficeMax Inc. (OMX- NYSE) just reported its fourth-quarter results. The loss came in at $229.5 million (US) or $1.95 a share. That’s GAAP.

Take off the bad stuff: store closings, asset impairments, and a big income tax reserve and the loss would be a much more modest $14.3 million or 12 cents a share.

Now look up the forecasts. In the earnings history the estimate for the fourth quarter is shown as a loss of 12 cents a share and the actual as 12 cents, for a surprise of zero.

It is impressive that these analysts had the prescience to exclude just the right number of charges to exactly meet the company’s reported net loss. A nice trick, but “actual”? Hardly.

What we need on these Web sites is a toggle. A happy face icon could show us the theoretical results without those nasty charges, while a click on the unhappy face of an accountant could show us the actual number based on GAAP.

As for OfficeMax, the stock jumped to $6 from $4.42 before the results were announced on March 4. That jolly move shows that investors feel the firm is basically on the right track and the fundamentals of negligible debt, an improved cost structure, and healthy cash flow will mean that real, true, unadulterated profits are in the not-so-distant future.

This is miles from the stock price of $1.62 where it was positioned when we first wrote about the firm in this column in December 2000, but still a long ride from our target price of $13.44. It looks like Mr. Happy Face is commanding attention.