PHOENIX, May 18 - One way to throw pursuers off your trail is to keep changing it all the time. IBM has become a master at it when it comes to financial reporting. Every year since Lou Gerstner took over the reins, the Big Blue has made some changes to its financial disclosures and/or Annual Reports. But at no point in the past, has the Armonk fudge factory put out so much fudge in lieu of facts, as was the case with its 1998 Annual Report.

In some cases, even the (revenue) figures for the same categories on page 64 and on page 86 didn't balance. In other cases, brand new fudge accounts were created and called business segments (e.g., "Enterprise Investments"). In some situations, entire reports DISAPPEARED, such as the international revenue, assets and net earnings breakdowns. Almost all business segments were renamed. And even if two were not - servers and software - their definitions were changed, and therefore, past figures had to be restated.

We've said before that IBM has turned financial engineering into a new line of business (see Annex Bulletin 97-30, 7/21/97). With its 1998 Annual Report, Armonk has turned financial reporting into abstract art, where only the artist himself knows for sure what the various brush strokes are supposed to mean.

But the top prize for 1998 goes to the Armonk's financial artist who came up with the idea of reporting the "funny money" along with real revenues. In an instant, IBM became a $91 billion-company. All it took was adding $9.4 billion of "internal revenues" to the amount earned from customers, and then calculating the pretax profits on the total of the two (i.e., including the "funny money").

And the Securities and Exchange Commission (SEC) calls that kind of reporting "financial disclosure?"

Maybe the SEC folks have been taking lessons from their colleagues at the Treasury, which has allowed the Federal Government to create $2.25 trillion out of thin air (see "What's a Trill Here, a Trill There...?", Annex Bulletin 99-11, Mar. 18, 1999). So Armonk still has ways to go with its "puny" $9.4 billion fudge factor, compared to the financial artistry of the Washington money wizards.

Yet Wall Street has lapped up all the fudge IBM slathered over its financial results, unperturbed by a possibility that the rest of it may not be as sweet. Nor were the casino players concerned about the fact that it is now nearly impossible to reconcile IBM's 1998 business segment information with prior years' figures. They've pushed the IBM stock to a new record high of $246 on May 13, up almost 50% since the time the 1998 Annual Report was released.

By merging the shrinking (-8%), but highly profitable, IBM maintenance business, with the fast growing (+22%), but marginally profitable, IT services operation, the company created a hybrid called IBM Global Services (GS). The IBM GS hybrid gives the impression of being both the fastest growing (+15%), and the most profitable business in IBM's "business portfolio," as Armonk likes to call its divisions, pandering to Wall Street terminology even with its operations.

The key words in the above sentences are "gives the impression." For, when you scrape off the Armonk fudge, what you see tare totally different businesses, with different metrics and different business results.

IBM's IT services is a booming $24 billion operation with a pretax margin of a mere 5.5% (and even that is a HUGE increase from the 2.2% equivalent margin in 1997). The maintenance, on the other hand, is a mere $5.2 billion-business segment, but with a pretax margin of 41%.

As a result, maintenance accounts for 62% of the IBM GS hybrid's published 1998 pretax profit - $3.8 billion (including Armonk's fudge; or $3.4 billion with the fudge scraped off by Annex). Without the maintenance, the IT services' $1.3 billion pretax profit makes it only a distant No. 4 within IBM's " business portfolio," behind servers, software and maintenance.

But the IBM GS hybrid gets 100% of the credit and looks like No. 1 in the eyes of the uninitiated, since the 62%-38% maintenance/IT services breakdown is never shown anywhere in the IBM 1998 Annual Report.

Battered and forlorn; shrinking and scorned, the IBM servers still contribute more than double the amount of money to the company's pretax profit than the suddenly "cool" businesses, like the IT services ($2.7 billion vs. $1.3 billion).

In fact, the S/390 alone, despite its declining $5.5 billion revenues, is still more profitable than IT services $24 billion business ($1.4 billion vs. $1.3 billion in pretax profits).

The "new news" about IBM's PC business is that even after all the Armonk fudge is applied to it, this second largest part of Gerstner's "business portfolio" lost about $1 billion in 1998 (as we said it had lost in 1992 - see Annex Bulletin 93-26, May 1, 1993).

And have been losing money on its PCs every year since... Including in the last two years (1996-1997) which had been hailed by Armonk's "pontificators" as a "PC recovery" period.

What is this new business? EI "provides a spectrum of initiatives in IT solutions, supporting the hardware, software and services segments of the company," reads the vague explanation in the 1998 Report.

"Supporting" other businesses? At a pretax LOSS of $60 million in 1998; $900 million in 1997, and $800 million in 1996? In other words, it's "Lou's Kitty;" a "stealth" R&D account; a playpen for the Armonk Emperor masquerading as a bona fide "business segment."

All of which brought back some long time memories....

In a December 1994 meeting at his office in Armonk, Jerry York, the then IBM CFO, told this writer that the company had just instituted a new "budget process," under which operating divisions were asked to contribute money to a special R&D account, "sort of a 'Gerstner kitty'," York explained.

Here's an excerpt from this writer's business diary for Dec. 13, 1994:

"Some 'kitty!' It will have 'several hundred million dollars' in it. JY said that its purpose are the kinds of investments which have new business potential, not just the enhancements to the existing lines which the divisions make.

'What makes you think that Armonk can spend that money better than the divisions?' I asked.

'I am not saying we can,' he replied. 'We'll pass the successful projects on to the divisions once they are ready for market.'

'Do you have some sort of an Investment Council to make such decisions?'

'For the time being, Jimmy Cannavino and I do it,' JY said. 'But in early January, we'll make a presentation to Gerstner, after which we may establish a more formal body to decide on such investments."

"Jimmy Cannavino," like York, a former IBM executive, was back then the company's chief technology officer. And "Lou's Kitty" has evidently mushroomed in the last four years to multibillion-dollar proportions. And has so far generated several billion dollars of pretax losses, according to IBM's own disclosures.