Last Thursday, at about the same time we were publishing my

column on Judge Jackson's back-of-the-hand response to

Microsoft's

(MSFT) - Get Report

proposal on remedies in the

Justice Department

antitrust action,

Goldman Sachs

analyst Rick Sherlund was telling the troops at Goldman that a split-up Microsoft would be worth less than today's whole.

That was precisely opposite to my continuing thesis, first published here in late 1998 and repeated in that column last Thursday, that a Microsoft divided along almost any plausible lines would be worth more, in the market, than the present whole. Sherlund is a first-class analyst, in many ways the ax on Microsoft. He hasn't built his reputation on being wrong.

So who was right?

I was. Because his analysis, though intelligent and rational -- perhaps

fatally

rational -- doesn't go far enough. It ignores ... us.

Briefly, Sherlund says a split Microsoft would be worth less for two reasons: First, by splitting the company into approximate halves, the court would sever two complementary and more or less out of phase revenue streams. Looking at Microsoft's history, Sherlund says, when its

Windows

revenue was down -- that is, during the long intervals between new releases -- the revenue from its applications package,

Microsoft Office

, was up.

Thus the two revenue streams, like a twisted strand of DNA, supported and offset one another, tending to level overall revenues. Because investors like relatively even and predictable revenue, Rick says, the more erratic revenue flows from each of two split-off Microsoft divisions would be relatively less attractive to investors, reducing the amount they'd be willing to pay for the two issues by about 10%.

Second, because two companies would necessarily incur some duplication of expense items -- administrative costs, support costs, field offices and field reps, etc. -- the value of each would suffer an additional 5% or so, in terms of share price, at the hands of unhappy investors.

Presto: A split Microsoft is worth 15% less than the unified whole.

Both are sound arguments. I think the history of Microsoft's income is less perfect, in terms of cyclical revenue flows from the two product categories, than Rick asserts. But certainly Redmond has seen an offsetting effect from Windows and Office revenue cycles -- a force for stability that would no longer be present on the balance sheets of each of the two proposed new companies. And sure, general and administrative costs would be somewhat higher running these businesses under two roofs than under one.

The argument founders, however, on issues of investor behavior and enthusiasm. And on post-split market realities.

Just as there is no rational, mathematical explanation why a $100 stock which splits 2:1 should find its shares selling the next day for $54 -- in effect, an 8% bonus for no additional underlying value -- I think a severed Microsoft's parts would benefit from a similarly irrational but very real increase in total share price.

Funds and institutions would be eager buyers of both issues. Individual investors would be eager buyers of both issues. Upward price pressure would inevitably increase the total price to something like 1 + 1 = 2.1. Or 2.5, or 3.

Moreover, as post-split share prices (inevitably) diverged, the delta between pre- and post-split total shareholder value would grow as well. My guess is that as the Microsoft Office side of the house pushed aggressively towards a services model -- an application service provider (ASP)-style business, in which some buyers lease access to Office applications through online connections, rather than buying the products in shrink-wrapped boxes -- it would begin to grow both revenue and its share price more rapidly than could the Windows group.

There are also, probably, more new-product opportunities on the apps side than on the OS side; the "Office Company" would no doubt exploit those opportunities, reaching beyond the already-substantial Office franchise to new products, new markets and new alliances.

Overall, and despite the relative stability of original equipment manufacturers revenue the Windows group would enjoy, I think the Office group, newly unleashed and able to enter many new businesses, would become the more valuable part of the former whole.

On the Windows side, evolving the Windows platform into a services business will be more difficult, riskier, a generally smaller and slower-growing opportunity. Pursuing the old "Windows everywhere!" goal -- under, to be sure, a new and more politically correct banner! -- would add new revenue streams, but as a partnership-based business, would be the victim of the general conservatism of the management of the "Windows Company" partners

Remember that companies such as

Compaq

(CPQ)

and

Dell

(DELL) - Get Report

have not led the market because of bold new innovations, but rather the opposite: because they have been careful to moderate progress, not getting ahead of their customers' expectations.

Rod Canion, cofounder of Compaq and its first CEO, used to say at every turn that the company only "innovates within the standard." That is, it moves forward only as fast as the broad underlying standard embraced by its customers evolves. Michael Dell has also said many times that Dell listens to its customers and then gives them what they want, rather than leading them into new and perhaps threatening areas.

Those have been appropriate strategies for top-tier boxmakers, companies with market leadership that was theirs to lose. But they certainly discourage big leaps forward at the operating-system provider to those companies.

The "Windows Company" will be a rich business opportunity, to be sure, if Judge Jackson gets his bifurcated way. But probably not so rich as the "Office Company" franchise.

Either way, the combined value of the two is likely to exceed the value of the two in their present form, because unhitching them allows both management groups to maximize their opportunities. And more importantly, in terms of share prices, it would provide incentives for us to value the two companies illogically higher.

After all, the market's hardly a rational place.

Jim Seymour is president of Seymour Group, an information-strategies consulting firm working with corporate clients in the U.S., Europe and Asia, and a longtime columnist for PC Magazine. Under no circumstances does the information in this column represent a recommendation to buy or sell stocks. At time of publication, neither Seymour nor Seymour Group held positions in any securities mentioned in this column, although holdings can change at any time. Seymour does not write about companies that are current or recent consulting clients of Seymour Group. While Seymour cannot provide investment advice or recommendations, he invites your feedback at

jseymour@thestreet.com.