Chickens, Eggs, and Analyst Opinion

An open letter to Steven Milunovich

- by Paul Murphy -

When HP announced, just recently, that it was combining the headquarters operations of its PC and printing divisions, but would continue to report their results separately I assumed they doing it to provide legal cover for the transfer of costs from the PC side to the printing side.

The core argument at the time of HP's takeover of Compaq was, as you may recall, between those who wanted to continue HP's long history as an engineering innovator and those who wanted to align HP more closely with Intel and Microsoft. Thus the nearly simualtaneous announcement of the "HP Invent" slogan and the laying-off of most of the company's innovators signaled the victory of the pro-Microsoft side and the defeat of Walter Hewlett's last ditch proposal that the printing business be spun off to give Fiorina et al a clear field in which to demonstrate their PC marketing expertise.

Since then HP's senior management, which won the merger battles, lost the fight in the marketplace as Microsoft lost interest in the Itanium and HP's bigger customers were left with nowhere to go but IBM first and Sun later.

The printing business, meanwhile, continues to prove Mr. Hewlett right by minting money faster than the PC people can lose it -leading to the present re-organization and, I'm sure, some interesting internal tensions.

The surprising thing about this, however, isn't the effect it's having in setting HP up for dismemberment -that was fully predictable given DEC's experience with the same strategy- but the apparently total lack of reaction in the professional analyst community. This kind of legal and political CYA in the HP executive suite signals impending catastrophe, yet HP's stock barely wavered with the announcement and I could find no obvious public record of any analyst or commentator who publically saw the re-organization as a response to disclosure requirements.

So maybe it wasn't? or maybe the analysts follow each other's opinions and one of the leaders has another agenda? Analyst opinion isn't open source and doesn't come with explanatory footnotes or explication of alternative views, so there's no immediate way for an outsider to know which came first: the opinion or the research.

We can, however, look back at other analyst pronouncements and see whether history sheds light on analyst current behavior.

The most common expression of analyst opinion is in the form of an earnings forecast for some company. There have been hundreds of thousands of such forecasts and thousands of academic studies of their contents. In general these studies either conclude that earnings forecasts are reasonably good or divide forecasts into predictions and targets to conclude that targets have predictive value but predictions do not.

What's going on here is that most earnings forecasts aren't forecasts at all, they are announcements under the analyst's name of consensus opinion arrived at in consultation with company management. As a result they represent targets to which company management commits, not forecasts, and are correspondingly likely to be about right as management works to achieve them.

There are cases, however, where an analyst will issue an earnings unblessed by management. Such predictions are generally wrong, but inherit credibility from the target form of forecast because the two kinds of forecast can be hard to distinguish and most of the press tends not to. As a result headlines like: Sun meets earnings forecast, misses on revenue incorrectly blame the company, instead of the analyst, for the forecasting error.

Earnings forecasts, whether real or not, have the power to move the market and are therefore open to abuse. For example an influential analyst can usually guarantee a disproportionate share price dip simply by issuing an overly optimistic forecast a week or so before the actual earnings report -and there's simply no immediate way for an outsider to know whether the analyst was mistaken or covering someone's short position.

More subtly, ill conceived negative analyst comments can become self-fullfilling prophecieS in part because they will predictably be quoted in ways which further damage the target company's competitive prospects.

In September of 2003, for example, Sun brought a lot of their heavyweights, including chairman and CEO Scott McNealy, to Ottawa to pitch java engineered security to some senior players in Canada's federal bureaucracy. The resulting meetings were reported, somewhat differently (although both operations are owned by Transcontinental Media Inc), as "SUN talks smart cards with feds" in the Ottawa Business Journal and as "Sun courts Canadian government with smart card plan" on

The Ottawa Business Journal piece (bylined Ottawa Business Journal Staff) concludes with this paragragh:

Also on Thursday, Steven Milunovich, a prominent analyst at Merril Lynch & Co., sent an open letter to SUN's board, as well as McNealy, saying the CEO needs "a makeover" and the company needs to cut up to 7,000 more workers and drop out of some businesses. Otherwise, Milunovich predicted Sun's ultimate value will only be its customer list for another tech company that buys it

The itbusinss article (bylined Scott Foster) concludes with this paragraph:

Also last week, Steven Milunovich, a prominent analyst at Merril Lynch & Co., sent an open letter to Sun's board and McNealy himself which said the CEO needs "a makeover" and the company needs to cut up to 7,000 more workers and drop out of some businesses. Otherwise, Milunovich predicted Sun's ultimate value will only be its customer list for another tech company that buys it.

The quotation from Milunovich relates to Sun, not the story being reported and has no more business here than an alternative paragraph puffing Sun's R&D capabilities would have had -but both stories would have been widely read among Ottawa's extremely risk averse civil service decision makers and influencers.

Among analysts Milunovich, in particular, is an interesting case in this context because his apparent success within Merrill Lynch bears no obvious positive relationship to the quality of his comments about Sun - a company whose products I use and greatly admire.

Here's an excerpt from's Report Card: Steven Milunovich (Alison Zombs 09/12/00) in which he gets three of a possible three stars under the criterion: "makes money for me:"

As Milunovich explains, "We've argued that the profits in the computer industry are going above and below the personal computer: above, into enterprise computing and, below, into information appliances." The first positive trend he sees is renewed demand for large servers -- particularly UNIX servers. While demand for "big iron" declined significantly in the first half of the 90s, the requirements of current applications necessitate a return to powerful processors. Milunovich sees Sun Microsystems (SUNW:Nasdaq - news) as the primary beneficiary.

At the time Sun shares sold for about $120, three months later and after a two for one split, they sold for $34; now they're around $4.20.

By October of 2003 he had direct advice for Sun as reported by Paul Krill for infoworld

In his letter, dated Oct. 2, Merrill Lynch's Steven Milunovich recommends Sun de-emphasize its Sparc hardware architecture and focus on the Intel x86 platform, spin off its crown jewel Java programming language, and pass on battling Microsoft on the desktop with the newly announced Java Desktop System, formerly known as "Mad Hatter."

Right now, of course, SPARC/Solaris is pretty much set to be IBM's only competitor as x86 gets replaced by either, or both, cell and throughput computing while Milunovich's much hyped "industry standard architecture" (the Itanium) has been all but abandoned.

In June of 2004 Martin LaMonica writing for zdnet, reports that:

Milunovich said that overall, it appears that HP made a smart move in acquiring Compaq Computer, which improved the company's market share in servers, PCs and storage. To better compete, Milunovich said, HP should continue making acquisitions of other companies. He noted that HP would be the most likely acquirer of Sun Microsystems.

About a month ago he issued a proclamation that Sun should buy either Red Hat or Novell; presumably as part of its commitment to open source.

I don't think his pronoucements make any sense but you don't get to be a Merrill Lynch vice president as well as their Global Technology Strategist by being an idiot, so I have to assume that this man's opinions have consistently earned that three star rating by making money for someone. But who? Not, I think, Sun shareholders.

And his advice on HP's re-organization? CBSmarketwatch quoted him as saying: "that there is a strategic rationale to H-P's move, and that it's "a slight positive given the possibility of improved management of PCs, both tactically and strategically."

But not a word about Walter Hewlett, the SEC, or GAAP compliance.

Paul Murphy wrote and published The Unix Guide to Defenestration. Murphy is a 20-year veteran of the IT consulting industry.