Lots of other people, though — including investors and the cable guys — don’t know what to make of Google’s moves.

They have two big questions:

• What is Google actually trying to do here?
• How much is Google willing to spend on whatever they’re doing?

And there are multiple theories to answer those questions, which aren’t necessarily mutually exclusive:

• It’s a sandbox for Google. It’s literally trying to figure out what goes into offering that kind of speed, and what happens when customers take advantage of it.
• Google thinks that by offering blazing-fast fiber in a bunch of cities, it will spur the likes of Comcast, Time Warner Cable and other incumbents to up their own investments in speed. And that eventually means that more people will have faster Internet, which is presumably good for Google.
• Google really does plan on wiring a big swath of the country. Because, why not? It’s Google. It can afford just about anything.
• During yesterday’s earnings call, analysts again tried to suss out what Google was up to. Both CEO Larry Page and CFO Patrick Pichette did a good job of speaking at length and not saying very much, beyond the notion that it would be good if lots of people had fast Internet.

Here’s one exchange between Bernstein analyst Carlos Kirjner and Page that is pretty representative:

Kirjner: “How do you expect fiber to have a major impact, given that it would take many billions and several years to pass something like 20 million U.S. homes — and after all of that time and money, you would be at best a mid-sized provider (Note: Comcast, the U.S.’s biggest broadband company “passes” — i.e., could provide service to — 53 million homes) in a market that accounts for less than half of your current business?”

Page: “I’m glad you’re very ambitious — the fiber industry’s not big enough for you. So I applaud that. We’d love to find businesses much bigger than our entire business to invest in, but I think that there’s only a small number of such companies that even exist. So I think on fiber, we look at places where we can provide products that we can make a very big difference in people’s lives, and we can make a lot of money and resources doing it. And I think it certainly meets that criteria.”

Does that spell things out for you? Me, neither. But Kirjner seemed to get some value out of it. In a note this morning, he writes:

• That he is now convinced that Google is not experimenting with fiber, but is serious about it.
• But! He also doesn’t think Google is going to embark on a truly massive effort: “Google is playing a very long game and does not aim to change the broadband access and pay-TV world in three or even five years … We remain skeptical that we will see a mass roll out involving billions of capex to build a fiber network passing millions of homes in the next few years, but think they will continue to expand Google Fiber’s footprint.”

Got it? Me, neither. But fascinating to watch.

Short-term, that means ads that run on phones are much cheaper than ads that run on PCs — often by as much as 80 percent.

That’s a well-documented challenge for the likes of Facebook and Zynga and even Google; the problem is just as real, if less-publicized, for other Web publishers.

If you are an Internet optimist, you have a ready answer for this problem: Advertisers will eventually catch up to their audience. If people are spending lots of time on their phones, marketers will follow the eyeballs.

That’s the gist of the argument Kleiner Perkins’ Mary Meeker made at our D10 conference earlier this year, when she unveiled her annual Internet Trends report.

Meeker updated her slides for a new presentation last night, but the key ad chart remains unchanged — she figures that between the desktop and mobile Web, there’s at least $20 billion in ad spending ready to move away from old media:

That’s a popular argument in the Internet world, for obvious reasons. But here’s the counter-argument, via a note Bernstein Research’s Todd Juenger published last month. The title spells it out quite clearly: “The Share of Time Versus Share of Ad Spend Fallacy.”

Juenger’s argument boils down to this: Yes, lots of people are spending lots of time on the Web, and now on their phones. But that doesn’t increase marketers’ desire to spend more money there — it just creates more advertising inventory, which might actually end up decreasing the total amount spent on those ads.

The money paragraph, such as it is:

We could actually make the argument that, because usage of Internet (especially mobile) continues to quickly grow, Internet ad spending (on an absolute level) should actually fall. If marketers are mainly are concerned with delivering a certain number of impressions, rather than a certain amount of spend, then everything else being equal, it would cost less to deliver the same number impressions next year than it did this year. Because in that scenario, supply is growing but demand is not. Or, perhaps marketers would look at the falling CPM’s and determine their preferred course might be to spend the same amount on Internet as last year, which would deliver X% more impressions. Getting, in that case, a lot more impressions for the same amount of dollars. The only way Internet ad spend grows is if marketers determine they want a lot more Internet impressions, causing them to spend more absolute dollars, at a lower/higher CPM (depending on which grows faster, demand or supply).

The easiest way to shrug off Juenger’s argument is to say that mobile Web ads are so nascent right now that no one has any idea what the market will look like, except that it has to get bigger, period. That’s probably right.

But we’ve seen plenty of evidence of Juenger’s argument in the desktop Web world. People keep spending more time there, but publishers have to keep slashing their rates to attract ad dollars. If the same scenario plays out on mobile, that gap might remain there for a long time.

(Photo courtesy of Shutterstock/Dominik Michalek)

Shares of the social giant’s stock jumped by more than 8 percent on Monday on news of multiple analyst upgrades. Facebook closed at nearly $26 per share, the highest price the stock has seen since the summer.

Carlos Kirjner, a Bernstein Research analyst, gave the most resounding support of Facebook’s stock over the next two years, upgrading the stock to “outperform” from his previous “market perform” rating (as my colleague Peter Kafka noted earlier).

“Facebook probably can increase the number of ad impressions per user per day,” Kirjner said, “with limited chance of seeing material deterioration in user experience.”

It’s especially surprising given that Kirjner, as The Wall Street Journal earlier pointed out, has been bearish on Facebook since its early days on the Nasdaq.

Indeed, many Wall Street analysts are changing their tunes as of late. The outlook on Facebook’s mobile advertising prospects — once considered the most egregious deficiency in Facebook’s overall monetization strategy — is faring much better.

Rich Greenfield, an analyst with BTIG, raised his price target two dollars on Monday, to $36. Greenfield’s main beef with Facebook mobile ads in the past was that they weren’t nearly as relevant as they should be. The sudden optimism stems from Facebook’s uptick in inserting more mobile ads, which he expects will boost short-term revenue throughout Q4 (though that may prove problematic in the long term).

Other promising reasons for the positive outlook include Facebook Gifts, the company’s push into e-commerce, as well as the up-and-coming Facebook Exchange ad-targeting program.

Facebook shouldn’t break out the eggnog-flavored champagne quite yet. The company’s shares are still down 12 points from the debut share price of $38 in May, and have fluctuated (mostly downward) quite a bit since.

But it’s probably a nice change of pace from hearing the Street’s bearish prospects over the last year.

The not-quite-as-good news, at least for the ad-tech guys: It looks like the Exchange is always going to be a side project for Facebook.

So says Bernstein Research analyst Carlos Kirjner, who figures that Facebook Exchange ads, which let advertisers pitch you based on your Web traffic history, will only end up accounting for 15 percent to 20 percent of Facebook’s old-fashioned Web ads. Those are the ones that run on the right side of the service’s page when you look at it on a PC.

That’s still a very big business, since those are the ads that generate about 80 percent of Facebook’s advertising business right now. But Facebook’s future is based on “newsfeed” ads — particularly the ones it will show mobile users — and Exchange targeting won’t work there.*

And Kirjner figures that the Facebook Exchange will be capped when it comes to its conventional ads, too. That’s for two reasons: He says there simply may not be enough demand to turn Facebook’s huge pool of regular ads into retargeted ones. And he says Facebook will move slowly with this stuff to avoid setting off alarms with U.S. and European regulators, who are already sniffing around retargeting and privacy issues.

But even if the Facebook Exchange plays a limited role for the company, it’s still an important one. Kirjner predicts that Exchange ads will help slow the decline of Facebook’s conventional ad business, which is dropping off as users head to the mobile Web.

If the Exchange wasn’t around, he says, conventional ad pricing would drop 5 percent next year, and 5 percent the year after that. But after factoring in the growth of Facebook Exchange, which adds a big pricing premium, Kirjner figures those ads will remain flat next year, and drop a bit in 2014.

That is: Without the Exchange, this chart would show a much steeper drop.

*Kirjner does think the company’s first efforts at mobile are working, by the way, and has moved his rating from “market-perform” to “outperform” — in conventional English, that’s a “buy.”

So, once again: Does that mean people really are ditching Comcast, Verizon and Dish, etc., in favor of Netflix, iTunes and Hulu?

And, once again: Maybe. But you can’t prove that based on last quarter’s numbers.

As we’ve pointed out before, there’s a seasonal cycle to the pay-TV business: The cable, telco and satellite guys usually add a bunch of subscribers in Q1, lose a bunch in Q2, lose a few more in Q3 and then gain some back in Q4.

Tally up the first nine months of 2012 and the pay-TV guys are basically flat — just like they have been for the past couple years, notes Bernstein Research’s Craig Moffett (click chart to enlarge):

For the past few years, the pay-TV guys could point to the cruddy economy as the reason for their nongrowth. But now that argument doesn’t work as well. Moffett: “Household formation, while still anemic, is showing signs of recovery. Pay TV industry subscriber metrics are not. Pay TV penetration of America’s households is therefore falling, even while the number of Pay TV subscribers is still inching higher.”

So, if you’re in the “everyone I know uses the Web instead of cable” camp, that sure sounds like the data supports your argument. Moffett is still unconvinced, though: He figures the net losses come from subscribers who simply can’t afford to pay for TV or the Internet, and are getting their fix from old-fashioned rabbit-ear antennas.

Plausible?

That all depends, of course, on whether people want to see the next three “Star Wars” movies Disney says it is lining up.

It’s a tricky bit of guesstimating, because the people who were young enough to be wowed by Lucas’ first installment when it debuted in 1977, are now oldsters who are 40 and much older.

And many of those same people (cough) despise the last three movies he put out. But if a new generation of kids like this stuff, then Disney is in good shape.

Bernstein’s Todd Juenger breaks it down in a note today. For starters, he figures Lucasfilm does $860 million in revenue a year right now (see table below).

Next you have to guess what kind of performance Disney will get from the new movies. If each of the next three three movies does “Avengers”-like numbers, they’ll generate $1.5 billion at the box office, $2.8 billion overall, and generate $1 billion each in profits, and the deal works out for Bob Iger and his successor.

Is all that doable? Yes, he says: “Not heroic, given the ‘Star Wars’ brand and Disney’s unparalleled global consumer products network.”

Here’s the math:

Whatever, comes the response from some of Apple’s most ardent fans. That’s what they said about the iPhone five years ago, and now look what’s happened.

True enough! And we should also note that there’s a scenario where the new, still-theoretical Apple TV plays by the TV industry’s rules and does very well for Apple.

But we should also point out that a TV box is not a phone, and that the parallels fall apart in some crucial places.

The biggest difference: In the case of the phone business, Apple ended up revolutionizing the way handset manufacturers worked with the mobile carriers.

With TV, if Apple really wants to re-order things, it will eventually have to negotiate with two different industries — the pipe guys and the content guys — who are both quite happy with the status quo. Their interlocking interests are what make the TV industrial complex so successfully resistant to change.

And here’s more along those lines from a great Bernstein Research note, which teases out the implications for Apple, the cable guys, the content guys and the tech guys like Google, Netflix and Amazon.

Bernstein’s analysts believe the Apple as set-top box scenario is “perhaps the least disruptive model one can envision for Apple’s entry into Pay TV.” But even getting to that point will be a challenge, they argue:

While similar challenges have been overcome in the mobile telephony market, the challenges here are far larger and more complex.
First, the distributors in video share in the economics of the content that they deliver. Cannibalization matters.
Second, the Pay TV business has no analogous hardware subsidy model to the one that prevails in post-paid wireless.
Third, the U.S. cable industry has become accustomed to near complete control over the development of set-top box and cable modem technologies (e.g., all the MSO-sponsored Cable Labs work on DOCSIS and tru2way).
Fourth, cable MSOs are clearly advantaged over their telco and satellite competitors when it comes to distributing a device that supports both linear and Internet TV – for reasons of both scale and ability to integrate broadband –giving them disproportionate negotiating leverage.
All of these factors make it harder for any device manufacturer to strike a favorable deal with the largest MVPDs. A successful conclusion to the “talks” reported by the Wall Street Journal is by no means a foregone conclusion.

The growth rate remains below the level of even anemic new household formation, suggesting that penetration is falling even as the Pay TV subscriber base is still growing. And that, in turn, suggests that yes, there are homes that are cutting the cord. Whether they are doing so because of online video options (as the technology press would have it) or poverty/affordability (as we would argue) is unclear.

– Bernstein analyst Craig Moffett in a research note today, analyzing the next-to-no-growth of the pay-TV business

But how do Netflix users actually use Netflix?

For at least some of them, the answer is: They use it as an $8-a-month babysitting service.

That’s the takeaway from a focus group Bernstein Research conducted with a bunch of San Francisco moms last month. Bernstein is particularly interested in how moms use Netflix with their kids, because the team there believes that kids’ shows on Netflix are a problem, or potential problem, for kids’ cable programmers like Viacom and Disney. And the discussion they had with these moms supports their thesis:

Our panelists who subscribe to Netflix largely originally did so for themselves, but are now using the service primarily for their kids. The content selection is perceived to be significantly better for kids than for adults, and the lack of commercials and ability to control the viewing choices are seen as positives.

And the quotes and paraphrases from the focus group are excellent, and ring at least partially true. At least to this non-San Francisco-based kid-watcher:

• We found that most mothers originally subscribed to Netflix streaming for themselves, but eventually the service is used primarily by their children.
• Some mothers like that Netflix streaming allows them to watch older content, but many consider Netflix as a service better suited for children because of lack of interesting content for adults.
• Mothers like that Netflix does not have commercials and contains programming their kids enjoy (citing Dora, SpongeBob, etc.).
• Many mothers cited instances in which their children watched the entire episode library of a given show (e.g., SpongeBob) on Netflix.
• Almost universally, the mothers found that kids do not tire of watching the same programs over and over again. …
• When asked if Netflix was an adequate substitute for cable, Group 2 erupted in a chorus of nos.
• Regarding the types of shows one child watched on Netflix, one mother stated, “Mostly she watches the shows she’s seen a million times [already] on Netflix.”
• “My kids do binge viewing through Netflix. I never turn on Netflix.”
• Regarding the availability of older content: “The Columbo’s we don’t own, we watch on Netflix. Columbo, Columbo, Columbo, Columbo.”
• “I think I use it more for the kids than I do for myself.”

But it’s equally important to note that most of the Bernstein panelists don’t say Netflix is their primary source of kids’ entertainment: Most of them also rely on some combination of DVRs, pay-TV on-demand offerings and other services.

And, a bit oddly, when Bernstein conducted a similar panel of moms in New York this month, it found that most of them had no idea how the service worked. Which again means they didn’t talk to my family or most people I know.

Unlike our panel in SF, this group of mothers had surprisingly little knowledge of the Netflix value proposition. They did not use Netflix for their kids as much as the SF panel suggested since most knew almost nothing about the product.

• When describing what the product offers for kids, some uninformed mothers did not see the need for the use of such a service, as they mostly needed programs that are downloadable. They felt that OnDemand and DVR fulfilled the majority of their programming needs.
• When asked what form of entertainment service the mothers would give up first, invariably it was Netflix.
• When asking about pricing for Netflix, one mother asked, “Is it $30 a month?”

Higher broadband bills — at least if you plan on streaming a lot of online video.

And no chance that today’s cable bundles go away anytime soon.

That’s the takeaway from Bernstein Research’s super-savvy Craig Moffett, in a note out this morning. It’s a great read, and if you can get your hands on it, you should. For those who can’t, here’s a quick summary of Moffett’s argument:

• Broadband providers are already moving away from broadband plans that charge everyone the same price, as long as their use stays under a certain cap, and toward usage-based pricing. That move has been blessed repeatedly by federal regulators.
• That’s a problem for Netflix, as well as for anyone thinking about getting into “over the top” video, because that will effectively increase the price of those services. With usage-based pricing, if you’re streaming a lot of Netflix — or Amazon Instant Video, or some other service that hasn’t launched yet – you’re going to pay more for broadband.
• When Netflix complained that Comcast was giving its in-house streaming video service a boost over Netflix by not counting the Comcast service against its own broadband cap, Comcast responded by dropping its cap and moving to usage-based pricing.
• A federal investigation means that all the cable guys move in this direction, with speed. “Additional scrutiny from the DOJ would likely definitively end caps … and instead usher in a regime of [usage-based pricing] that would ultimately be even more threatening to online video providers than caps themselves.”
• Meanwhile, while The Wall Street Journal report raised the notion that the Feds are also looking into the cable industry’s practice of bundling channels — which means that you can’t watch the Disney Channel unless you also pay for ESPN — that’s not going to go anywhere. Because no one’s breaking any laws. “The reasons we have the unwieldy content bundles that we have today are not because Comcast or DirecTV won’t sell us individual channels, it’s because Disney and Viacom won’t sell them individual channels. We see little or no chance that the DOJ (or the FCC) will take on that issue. The Supreme Court has already made it clear that that is a losing battle.”

So, to recap: If you’re waiting for the cable/TV model to blow up, don’t count on any help from the federal government.

So why did the number of pay-TV subscribers increase in just the last three months?

They didn’t grow much — a modest 422,000 subscribers, for a very modest 0.2 percent growth rate — but they still grew.

Those numbers come from Bernstein Research’s Craig Moffett, a longtime skeptic that “cord-cutting” is a real and pervasive problem for the cable guys (at least for now). It’s not the first time he’s shown evidence of barely-there growth for cable TV — last quarter, for instance, he gathered similar numbers.

But his numbers do conflict with other reports that show evidence of cord-cutting. Earlier this month, for instance, Nielsen said that pay-TV subscribers had shrunk by 1.5 million in 2011.

The easiest way to reconcile Moffett’s numbers with other reports is to note that almost all of the analyst’s data comes from the publicly traded pay-TV providers themselves — like Comcast, Time Warner Cable and Verizon — in the reports they offer up to shareholders. Most of the other stuff you’re seeing comes from polls and surveys.

Here’s his data. You’ll need to click the image to enlarge it:

But what about all of you folks who tell me, over and over, that you’ve ditched cable for some kind of combo of Netflix, Hulu, Apple TV, or even pirate streams? Surely I’ll hear from some of you again, just as soon as I publish this.

And I believe you folks, too. I can certainly imagine many scenarios where tech-savvy people — and even not-that-tech-savvy people — are able to satisfy their video urges without paying for a TV subscription. But my operating theory, for now, remains my vegan analogy: “They’re real, and they’re out there. They’re particularly notable in certain places like New York, the Bay Area and college towns. And they over-index at certain Web gathering places, like this one. But McDonald’s sales are still chugging along.”

But people are still watching a whole lot of TV — perhaps as much as ever. So the more nuanced answer is that some people are swapping out the Web for TV, some of the time. And other times Web video consumption may end up leading to more TV-watching.

Here’s some evidence supporting that idea: New data from Bernstein Research that tracks TV viewing habits for Netflix streaming subscribers. It shows a drop in kids’ TV viewing — but an increase for networks like AMC and FX. Presumably that’s because Netflix users are discovering old episodes of shows like “Mad Men” and “Sons of Anarchy” on Netflix, which is prompting them to watch new episodes of those shows on cable.

This chart shows a dropoff for programming on kids’ networks like Disney and Nickelodeon:

And this chart shows spikes for AMC whenever the network shows new episodes of shows that have old episodes on Netflix:

Bernstein put its research together using data from TiVo users. So on the one hand the numbers are pinpoint accurate, since they’re tracking actual TV usage. And, on the other hand, it may not be representative of the entire country.

Still, it makes intuitive sense, and certainly syncs up with the way the Web gets used in our house: Our kids are unaware you can do anything with an iPad other than watch “Backyardigans” and “Dinosaur Train.”

Meanwhile, I just plowed through the first four seasons of “Breaking Bad” on Netflix (and iTunes) last month — and am going out of my mind waiting for the fifth season to start on AMC this summer. Let’s cook!

But that hasn’t happened yet. And it’s possible that even as Web video grows, TV will continue to do just fine.

That’s the thesis of Bernstein analyst Todd Juenger, who made his case to investors earlier this week. Two slides from his presentation sum it up well.

First, he notes that even though eyeballs have moved away from broadcast TV, ad dollars have not (click to enlarge):

Even more important: Though the Web ad business is growing, TV continues to grow, too. And while other old media industries have shrunk, their losses haven’t turned into equivalent gains for the Web (click to enlarge).

But what about consumer spending? After all, Netflix is streaming more than 2 billion of hours of video every three months. That has to cut into TV, right?

Not really, says Juenger, noting that overall TV viewing is still up. Instead, he says, Netflix, iTunes, Amazon et al are eviscerating the DVD business. Important distinction.

And yes, all of this could eventually change, particularly if the digital guys figure out how to break up big cable’s lock on programming. But as we keep pointing out, that’s a very strong lock. It’s not going away anytime soon.

Latest example: It appears that the pay-TV guys actually increased their subscriber totals last quarter. Not by much — something like 240,000 new accounts, which translates into a growth rate of 0.2 percent — but an increase is an increase.

Chart via Bernstein Research (click to enlarge):

So what do you want to make of that? Not much, probably — just like the subscriber declines we’ve seen in previous quarters haven’t meant that much. If we saw a sustained move in one direction or another, that would be meaningful, but for now it’s more or less a flat line.

Google has asked federal regulators for permission to sell pay TV in Kansas City, where it has been working on a broadband/fiber buildout, and the WSJ thinks it could launch in a couple months.

It would be awesome if Google could use this as an opportunity to break up the cable bundle, and let people buy individual channels instead of big expensive blocks of programming they mostly ignore.

But that’s very unlikely to happen, because the cable programmers love the bundle, and don’t have any incentive to break it up (see: Viacom’s Philippe Dauman and ESPN’s John Skipper last month at Dive Into Media). If Google wants cable TV programming, it’s going to have to play by cable TV’s rules.

So what can Google offer that will make someone switch from Time Warner Cable, which dominates the pay TV business in Kansas City?

Very high-speed Internet access, for starters. And perhaps Larry Page will figure it’s worth his while to offer the service at an extremely low margin, because the whole project is a very expensive test, anyway.

Beyond that, here are some guesses/predictions from Bernstein Research analysts Carlos Kirjner and Craig Moffett, who have a pretty good handle on this stuff. They’re the ones who predicted on Tuesday that Google would file for cable licenses “very soon.” This prognostication comes from that same note:

DVR in the Cloud: “We would expect Google to store (and make available to consumers) the content across all or most TV channels it will provide, making available to users not just the live stream but also past content going back several days or weeks, if not longer. With the addition of good search and discovery and user interface capabilities, this would make DVRs obsolete.”

TV Where You Want It: “We would expect Google to offer access to something like ‘Cloud TV,’ described above, including the live TV stream, across multiple devices, such as computers, tablets and handsets.”

A Programming Guide That Doesn’t Suck: “It is not very hard to imagine a much better user interface than the one currently provided by the MSOs and satellite TV providers,” presumably along the lines of what they’ve been showing off with the revamped Google TV.

Again, the big caveat here is that none of this happens unless the cable programmers play along. And while none of the stuff described above seems truly mind-blowing, some of it — like live mobile streaming — will stil require programmers to give Google capabilities they haven’t given to heavyweights like Comcast and Time Warner Cable.

And even though the cable programmers often tangle with the cable providers, they’re at least comfortable with them in general. Google, though, still scares the bejesus out of lots of traditional media companies, so I’m not sure how many of them will play along.

One pretty good bet: Google’s foray into cable TV won’t include anything from Viacom, since the cable giant is still suing Google in the YouTube copyright case. So no Snooki for Kansas City.

[Shutterstock/Dwight Smith]

Here’s a helpful reminder, via a Bernstein research note out today. By 2016, analyst Carlos Kirjner predicts, the majority of Web search queries will come from mobile devices.

One interesting caveat to that projection: It doesn’t include searches from tablets. I gather this is because Kirjner is trying to distinguish between searches conducted via wireless networks and those made using broadband connections, via Wi-Fi, although I’m not quite sure why that matters.

But if you added tablet searches, you’d reach the tipping point that much earlier. And it will get there much sooner in the U.S., anyway, because smartphone penetration is much higher here than other parts of the world.

Given that Apple’s iOS devices use Google search now — last fall, Google said that two-thirds of its search traffic comes from iOS — all mobile growth is good for Google.

But at some point, Kirjner suggests, Apple may decide to jettison Google for a competitor like, say, Microsoft. And that could cause real problems for Larry Page and company. But those problems would be much, much worse if Google hadn’t created an iPhone competitor in the first place.

Now that new HP CEO Meg Whitman has decided to keep the PC business rather than spin it out as her predecessor Léo Apotheker had proposed, HP appears to be heading back, however slowly, toward the kind of stability for which it had long been known.

But can it still perform? Bernstein Research analyst Toni Sacconaghi took the opportunity to examine the company’s business units and finds that, yes, the prospects for growth aren’t unreasonable. The current portfolio of HP’s business, he estimates, is likely to grow its sales organically by a combined 2.5 percent on an annual basis going forward. Add to that additional growth from the planned shifts toward higher-margin IT service, networking and software businesses, and the picture gets brighter, Sacconaghi writes in a note to clients issued yesterday. HP, he argues, should be able to grow its per-share earnings by 9 to 10 percent annually over the next three to five years.

Part of his assumption is that HP uses about $4 billion of its $8-$10 billion in annual free cash flow for share buybacks, which will help goose earnings on a per-share basis by reducing the number of outstanding shares; plus another $3 billion annually for acquisitions, buying companies at reasonable multiples of about five times revenue.

So how does each sector of HP’s business look?

PCs: Sacconaghi reckons that the global market for PCs will grow about 6 percent a year on a unit basis, down from historical averages of about 10 percent, and that it will grow on a revenue basis by about 2 percent a year through 2015. As the biggest supplier of PCs in the world, HP will, at the very least, hold its market share. While PCs are HP’s largest business, accounting for $40 billion in its last fiscal year and about one-third of its overall sales, the unit generates only 13 percent of operating profits. And, yes, while tablets are a threat, Sacconaghi sees no reason that PCs and tablets can’t grow together. The increasing need for an Internet connection everywhere and the increasing amount that consumers are willing to spend on technology, coupled with price declines over time, mean that consumers will be willing to spend more on notebooks and tablets.

Printers: As with PCs, HP is the market leader in printers, and as that market grows its revenue by about 3 percent a year, HP will probably hold on to its share of the market. Tablets may have a long-term impact on printing behavior, but Sacconaghi argues that consumer behavior tends to change slowly. A decade ago, he says, investors worried that email and the paperless office would kill printing. “Despite these ostensible headwinds, HP and Lexmark collectively grew their supplies revenues by 7 percent per year between 2000 and 2010, driving low to mid single digit growth for the printer industry.” HP’s printer division accounted for $25.8 billion in revenues in the last fiscal year, or about 20 percent overall, and because of its 17 percent operating margin, delivered 29 percent of the company’s profits. It’s a classic razor blade business, as HP loses between 20 and 25 percent on the printing hardware, only to make it up on supplies that command a 60 percent margin. The presence of tablets could actually boost printing, and thus increase the sale of those supplies down the road. Overall, the printer business should continue to grow at about 3 percent a year, Sacconaghi says.

Enterprise, servers, storage and networking: Sacconaghi expects HP to grow revenues in the ESSN unit by about 2 percent a year, slightly below the industry growth rate of 3 percent. He bases this on the expectation that HP can grow its networking business, hold its market share in the Intel-based servers, and lose share in its enterprise storage business to players like EMC and NetApp, and also lose share in the Unix storage business because of Oracle’s decision not to support the Itanium chip.

Services: This unit should grow by about 2 percent annually, slower than the market, which is growing at 4 percent. “We believe that outsourcing is a maturing business, and that EDS, which HP purchased in 2008, is and was an underperforming asset,” Sacconaghi writes. “As a result, we forecast HP’s outsourcing revenues to remain flat.” Expect it to grow in line with the enterprise hardware business, and along with the consulting business.

Software: The market research firm IDC expects the end markets for HP’s software to grow by 7 percent a year. If you assume that HP makes more software acquisitions, which Sacconaghi does, then you can reasonably predict its software revenue will grow by 10 percent a year.

So what’s the risk to all this? Leadership. “Ultimately, our projections for HP are predicated on the company choosing a disciplined financial approach to running the company that includes modest revenue growth, small and leverageable acquisitions, a strong operations focus, and high returns of capital to shareholders,” Sacconaghi writes. The last two CEOs who tried to “transform” HP — Carly Fiorina and Léo Apotheker — failed.

HP’s isn’t “broken,” but in fact offers a favorable risk for patient investors, Sacconaghi says. He rates HP shares as an “outperform,” with a price target of $37.

IBM, he says, should meet expectations for the quarter by delivering per-share earnings of $3.31, slightly better than the consensus estimate of $3.21. On top of that, he expects IBM to raise its guidance for earnings on the year to $13.35 per share or higher. Sacconaghi says IBM may earn as much as $13.60 a share this year, depending on how much it ultimately saves from workforce reductions and a lower tax rate. He says that IBM has beat its consensus in each of its last 15 quarters and raised annual earnings guidance in nine of its last 11 quarters.

All good, right? Sure, but how long can IBM keep this sort of thing going? Certainly not forever, especially in a tough global economy. Revenue growth this year will be difficult to compare to last year, Sacconaghi writes, especially in light of a stronger U.S. dollar, a slowing business cycle in hardware upgrades and a slowdown in services growth over the last 18 months. As such, his estimates for revenue growth are below those of the Street consensus: Where the Street expects IBM to report sales of nearly $112 billion in fiscal 2012, Sacconaghi expects $109.3 billion.

“IBM has benefited from a favorable currency environment, which has boosted the company’s headline revenue growth number, which is likely to reverse and pressure IBM’s reported revenues in the first half of 2012 at current spot rates,” he writes. “While we don’t expect this to lead to earnings misses versus the consensus given that IBM hedges, we believe that revenue estimates need to be revised downwards from current levels.”

In a scathing assessment from an analyst who’s rarely so negative on HP, Sacconaghi argues that investors are universally opposed to the Autonomy deal based primarily on the high price that HP is paying, which works out to 11 times sales, versus the option of using HP’s precious cash reserves elsewhere.

Unhappy HP investors have little recourse, Sacconaghi writes, and Autonomy shareholders seem especially eager to get the deal done. As the Financial Times reported yesterday, the fact that 42 percent of shareholders signed on to the deal so early in the tender process gives an indication of their opinion that the price HP is paying is a favorable one that’s not likely to be met by another bidder. It’s also an unusually high response rate this early in the process.

And there’s little if any chance that HP shareholders could stop the deal now even if a majority of them wanted to, he says. By paying cash for Autonomy, HP doesn’t need shareholder approval, and according to the terms of the deal, HP has to follow through absent a “material adverse change,” such as new negative information about Autonomy (like a major legal problem or something on that scale).

There’s certainly plenty to dislike about the deal. As Bloomberg News reported, HP sold $4.6 billion worth of bonds in order to help pay for the deal, though with yields on corporate bonds near record lows the timing to take on debt could be worse.

But even with corporate bond rates at relatively low levels, HP is paying more to finance its debt than its peers. Compare, for example the five-year fixed-rate bonds that HP issued yesterday with a coupon rate of 3 percent. In July, IBM issued five-year fixed-rate bonds at a coupon rate of 1.95 percent. I’m not exactly an expert on corporate bonds, but the way I understand it is like this: If you think of bonds like a credit card, HP is paying more than 1.5 times the interest on its debt than IBM is. That block of five-year bonds represents $1.3 billion worth of the $4.6 billion offering, making it the biggest portion of the debt offering disclosed in an HP regulatory filing yesterday.

Separately, two securities law firms said today they have launched investigations of HP and its board of directors. The Briscoe Law Firm, headed by former Securities and Exchange Commission attorney William Briscoe, and Powers Taylor LLP announced they’re investigating what they call “potentially misleading statements” by HP between Nov. 22 and and August 18. During that period, the firms allege that HP executives and directors made false statements or failed to disclose material information about the TouchPad tablet and the webOS operating system running on it. HP, they say, “lacked a reasonable basis for their positive statements,” and thus caused HP shares to trade at artificially high prices, peaking at $48.99 on Feb. 16. HP shares were nowhere near that high today, closing at $22.93, up 23 cents.

Fifteen years after their initial Comcast investment, Microsoft’s vision of a Windows-based gateway to the television still hasn’t materialized. Now it is Google’s turn to storm the fortress. And, like Microsoft before them, they have decided to do it from the inside.

Bernstein Research’s Craig Moffett, in a note (reg. required) savaging the notion that buying Motorola will allow Google to disrupt the TV business. Moffett does see a role for Google in helping cable operators measure and target TV advertising, though. For a less pithy take, read AllThingsD.

Moffett expects to see the cable guys like Comcast and Time Warner Cable lose a total of more than 300,000 subscribers, while the satellite TV companies may eke out about 70,000 new customers.

Those results could be balanced out by growth from the TV services offered by AT&T and Verizon, who have added 412,000 new subscribers. And overall pay TV numbers could end up positive for the quarter, but Moffett says that’s no sure bet at all: “Our conviction in a positive aggregate number is all but zero.”

Moffett is quite skeptical that pay-TV customers are actually cutting the cord and satisfying their TV needs with Netflix, Hulu, etc. For quite some time he’s been arguing that pay TV growth is a function of new household formations — and since there’s none of the latter, there can’t be any of the former.

Again, note that Reed Hastings and the rest of the Netflixers have been diligently announcing that they don’t believe there’s cord-cutting either — and if there is, they’re certainly not contributing it. But there are a lot of people who will see the next few weeks’ numbers as evidence that cord-cutting is, indeed, for real.

Apple is succeeding in the category because it reinvented it. Now anyone that hopes to compete in it must do so by peddling products similar to it. And because of that, Apple will dominate the tablet category in much the same way it dominated the portable music player category.

So there is a tablet market, but it’s been subsumed by the iPad market, just as the MP3 player market was engulfed by the market for the iPod.

Consider this observation from a new Bernstein Research survey: “We find that consumers are not interested in form factors that deviate from the benchmark set by Apple. Few consumers, less than 15 percent prefer the 7″ screen size versus the 10″ screen of the iPad. Over 50 percent of respondents are firmly in favor of the 10″ screen, which leads us to conclude that the 7″ tablet models recently launched, like the BlackBerry PlayBook, are destined for failure. Consumer’s preference for the 10″ form factor explains the lukewarm response to Samsung’s 7″ Galaxy tablet and the rapid introduction of larger screen models in that series.”

In other words, success in the tablet market may well be dependent on how similar a manufacturer’s offering is to the iPad. Which is quite a challenge given the formidable combination of hardware, software and app ecosystem that the iPad represents — not to mention the sheer power of its brand.

Said Bernstein, “Fifty percent of respondents preferred Apple over all other brands. There is a remarkable degree of unanimity in consumer’s preferences for the iPad over competing products. … In the US, we find that Apple has more than double the brand appeal of BlackBerry, HTC, Motorola, Nokia and Samsung combined. These manufacturers have a very high level of brand equity and visibility in adjacent categories. It is striking that they hold so little appeal for consumers in tablets.”

No surprise then that Bernstein sees the tablet market playing out in two ways, each with Apple in the catbird seat.

UPDATE: Some additional perspective on this from Instapaper creator Marco Arment (via Daring Fireball)

Analysts at Bernstein Research are offering a different take. They argue that passage of the deal may well lead to less competition and higher prices. However, they say that’s not such a bad thing.

The firm notes that American wireless firms have lower profit margins than their European counterparts and that consumers here consume twice to three times as much data, but pay only half or a third as much per megabyte for that data.

“Given the economies of scale of wireless, the constraints of CEOs and historic decisions with regard to spectrum, only further consolidation of the U.S. market is likely to deliver better infrastructure and lower prices,” Bernstein said in a report. “In short, more wireless competition may (though there is no guarantee of this) deliver lower prices near term, but it will almost certainly deliver lower quality infrastructure.”

The surest path to higher quality, the firm reasons, is consolidation.

“If U.S. regulators want to redress the lamentably lousy wireless service Americans receive, one of two things must happen: U.S. operators must be able to reduce their cost to serve (probably through consolidation) or Americans must pay more per unit consumed,” Bernstein’s analysts say in the report. “If the ‘Just Say No crowd’ gets their way, they may well end up with cheaper telephony near term (although this is far from certain), but they will certainly receive a poorer wireless service.”

It’s an interesting argument, though probably not the one we will hear from AT&T as it lobbies for government approval of the deal. Both the Federal Communications Commission and Department of Justice have to sign off on the transaction, while several state attorneys general have said they are looking into the matter. Sprint has also vowed to fight the deal.

For one thing, January’s restructuring of the board of directors has left a bad taste in the mouth of a pair of shareholder advisory firms, who have publicly called upon HP investors to vote against the re-election of as many as three sitting directors and against some say-on-pay proposals.

Sources familiar with the situation confirmed to me that after today’s event, HP’s investor relations team plans to mount what’s being described as a “road show” to counter the recommendations to vote against management made by Institutional Shareholders Services and Glass Lewis among retail investors. Exactly who is involved and whom they plan to visit couldn’t be determined. HP had no comment about it.

At least part of the road show’s mission will be to drive home the highlights of the strategy that Apotheker lays out in his keynote today. But there is some nagging concern that a sufficient number of shareholders, put off by repeated instances of board room drama over the last decade — Carly Fiorina’s ouster in 2002, the pre-texting scandal in 2006, and last year’s departure former CEO Mark Hurd — may vote against the three directors standing for another term: Lawrence Babbio, Sari Baldauf, and Ken Thompson.

That enough investors would vote against management to make a difference may seem unlikely at first until you consider the number of shareholder lawsuits stemming from the Hurd affair that are currently pending both in federal courts and in the Delaware Chancery Court. The fear of an embarrassing defeat for HP and its new board at the annual meeting on March 23 isn’t an unreasonable one.

Then there are the larger questions. As Bernstein analyst Toni Sacconaghi pointed out in a recent note to clients, HP’s stock has underperformed the S&P 500 since Hurd’s departure, and the lag has been driven mostly by uncertainty among investors about its strategy and about Apotheker himself. Its disappointing results in the first quarter didn’t help matters.

There are numerous questions around HP’s hardware strategy, particularly around the PC business. While it has promised to put the WebOS platform it acquired last year when it absorbed Palm into tablets and into every PC it ships, HP is still seen as far behind Apple on the tablet front. We all know why that is.

But the questions go deeper than that. Apotheker, a former CEO of the business software giant SAP, last week sent a strong signal in an interview with Bloomberg News that he plans to acquire some software companies.

What kind of acquisitions? He’s ruled out SAP, for one thing. And in a meeting with Bernstein’s Sacconaghi last month, he said any acquisitions would not be so large as to “keep investors awake at night,” meaning, Sacconaghi suggests, that they would probably not exceed $5 billion. Aside from buying software companies, he’s also signaled that the days of cuts–the hallmark of Mark Hurd’s tenure–are over. Cutting costs is out, investing is in.

But then there’s the larger issue about whether or not Apotheker can steer Hewlett-Packard, the storied Silicon Valley icon, on a course that restores its former glory. The question marks around him on this score are considerable because the task is just so huge. HP is a sprawling $126 billion juggernaut meaning change comes slowly, often in barely perceptible steps that leave impatient investors wondering what’s taking so long.

At the outset, the Apotheker’s strategy appears to be summed up pretty simply: Bring the market-leading position in hardware to bear and combine its offerings with a newly ascendant software business, which together will feed into an IT services business that aims to compete with IBM. It’s not going to be easy and the most important important thing that Apotheker has to do is inspire both analysts and shareholders alike that he’s the man to get the job done. As yet both are understandably skeptical mainly because Apotheker is an unknown quantity.

When Apotheker’s predecessor Mark Hurd took over at HP in 2005, there was very little doubt about what kind of CEO he would be: A relentless, unsentimental cost-cutter, and this much was clear before he was even officially on the job.

And while Apotheker has pointed tentatively in the direction he’d like to take HP, there are still more questions about him than there is certainty. His most important job will be to convince all concerned that he’s the man who can steer HP back on a course to greatness.

PREVIOUSLY:

• Another Advisory Firm Singles Out HP Director Babbio
• Shareholder Group Contends HP’s New Board Is Too Chummy
• “Peripatetic Polyglot” Léo Apotheker Wants to Save HP’s Soul by Buying Software Companies
• HP Earnings Today: Will Léo Apotheker Speak?
• Michael Dell Thinks HP Paid “Way Too Much” for 3Par
• Judge: HP Can Re-Investigate Hurd Departure
• Is This the HP Board That Will Allow Us to Stop Thinking About HP’s Board?
• Meg Whitman, Patricia Russo Among Five Joining HP Board
• HP Plans Another Probe Into Hurd Departure
• Léo Makes It Official: SAP’s Bill Wohl Joins Hewlett-Packard
• Want Enterprise Growth, HP? Think Services
• Mark Hurd Really Wants to Keep the Jodie Fisher Letter Private
• Mark Hurd Doesn’t Want You to Read the Letter That Cost Him His Job
• HP Networking Head: “People Are Tired of Paying for Cisco”
• HP Names Ex-SAP Chief Apotheker as CEO

Nokia spends about five times as much on mobile R&D as Apple. In fact, Nokia has nearly as many engineers working on its smartphone software platforms as Apple does for its entire product line. Says Dediu, “Symbian alone may cost twice as much to develop than the iPhone (including the hardware).”

A shocking metric, if correct. And a pretty dismal return on investment–unless there’s another version of Symbian in the pipeline that will best iOS.

PREVIOUSLY:

• Not Seeing Much Return on That Massive R&D Spend, Are You, Nokia?
• Nokia: Big and Slow