Pages

Jul 30, 2009

What Hath Tech Wrought?

Here's a paragraph from today's WSJ that neatly captures some of the themes I've been exploring about technology's impact on business models, corporate life-cycles and investment horizons.

"Kodak spent $3.4 billion from 2004 through 2007 converting the bulk of its 129-year-old business from high-margin film to more competitive electronic technology. It is in the midst of cutting 3,500 to 4,500 jobs, which could reduce its work force to a 1930s-era low of 19,900 from a 1988 peak of 145,300."

It's worth noting that the enterprise value (equity value plus net debt) of EK is a little under a billion dollars today (at roughly $3.00 per share.) Including pension liabilities of $2.4 billion in the net debt calculation increases the enterprise value to roughly $3.4 billion, equal to the amount Kodak invested over four years to convert from a film-based to an electronic imaging company.

Under the most generous interpretation, the market is valuing Kodak's 129-year history at zero right now, despite it's highly recognized brand and long tradition of technological innovation.

Jul 25, 2009

AOL - R.I.P.?

(updated July 30 to include note 2.)

Saul Hansell has put together an illuminating timeline of AOL's sad decline over the past decade at his New York Times blog.

As Tim Armstrong, who recently took the helm at AOL, attempts to revive the company, I am reminded of one of Warren Buffet's most popular aphorisms: "When a management with a reputation for brilliance tackles a business with a reputation for bad economics, it is usually the reputation of the business that remains intact."

To be fair, AOL is not completely dead yet. According to press reports, the company was valued at around $5 billion based on Time-Warner's recent buyback of Google's 5% interest in AOL, but the real valuation is likely half to two-thirds that much (see notes 1 and 2). In the quarter ended June 30, 2009 AOL reported operating income of $271 million on revenue of $804 million (see page 19 of filing). At quarter's end, AOL still counted 5.8 million paying subscribers for internet access, but lost 500,000 subs in the quarter. At it's peak, AOL claimed 27 million subs and was the world's most visited website. Today, acording to Alexa, AOL.com is the 35th most popular site, only two places behind Google's Italian search site and a distant 33 places behind Yahoo. So if AOL is not quite dead, it does look a bit like the guy over John Cleese's shoulder in this classic Monty Python bit.





I have been an AOL skeptic for a long time, going back to the mid- to late-1990's when it's widely publicized service failures gave rise to nicknames like "America Off-line" and "America On Hold".

My skepticism was primarily related to the "walled garden" approach to the internet that made up AOL's content offering. While this "internet-on-training-wheels" offering clearly worked for internet novices, it always seemed fundamentally disadvantaged to me. It just seemed unlikely that a single company could pay its employees and moderators to "summarize the internet" when the internet itself was available for free. And over time, new and improved (and free) internet browsers and portals like Yahoo (also free) and search engines like Google (again, free) made the internet easily navigable, rendering AOL's walled garden of content embarassingly superfluous. At bottom, I always saw AOL as a dial-up internet service provider whose attempts to become a content company were doomed by an infinitely more powerful force that, ironically, AOL itself helped enable.

And writing this in 2009, I feel reasonably sure that my decade-old instincts were correct in the long run.

But there was a period, let's call it 1997 through 1999, during which I was spectacularly wrong about AOL (or at least about it's stock price) and I am grateful for my employer's compliance restrictions that prevented me from shorting the stock.

The AOL saga illustrates some classic themes for technology and growth stock investors.

1) In the long-run, technology matters... a lot. Ultimately, AOL's core business was dial-up internet access. Its inability or failure to develop a competitive broadband offering was an obvious and fundamental weakness that eventually and inevitably eroded its subscriber base. And as I mentioned above, the attempt to develop a content offering to offset that risk faced a powerful, if not invincible force.

2) In the short run, don't underestimate inertia. AOL's subscription model, based on relatively modest charges to a subscriber's credit card, worked for a long time even after broadband internet service became widely available in the U.S. I suspect many subscribers carried the service for several months longer than they needed it because it required a proactive decision (and a not-insignificant investment of time) to cancel the service. True, AOL's customer service operations and policies seemed designed to make it inconvenient, if not downright impossible to quit, a practice for which the company has been widely criticized. Even today, some subscribers are unwilling to part with a long-standing email address, creating some stickiness in the subscriber base. There was also a time -- which seems almost quaint in the Twitter age -- that one's AOL Instant Messenger buddy list was considered a significant barrier to switching.

3) Sometimes, widespread customer service "failure" is really evidence of overwhelming customer demand, hence a bullish signal. When AOL switched from metered billing to flat-rate billing for internet access, it triggered a huge increase in subscribers and usage. AOL's inability to keep up with the demand surge led to its widely publicized service outages. From an investment perspective, this was actually positive, at least in the short run. Indeed, the surge in subscription revenue allowed AOL to quickly invest in the infrastructure to meet the increased demand. To be clear, widespread customer service failures that are quickly rectified may be bullish signals if they reveal massive consumer demand and a quick response by the company. This investment thesis should not be applied to businesses whose customer service failures seem chronic and institutionalized, like for instance, certain U.S. airlines and cable companies.

4) Adaptation is really hard for successful companies. AOL was not utterly blind to the fact that high-speed access presented a threat to its dial-up subscriber base. The company did experiment with broadband offerings and "Bring Your Own Broadband" discount packages for email and instant messaging. But it does seem that corporate hubris prevented AOL from striking deals that could have blunted the threat from broadband. At its peak, AOL was notoriously high-handed in negotiations with business partners, and it's easy to imagine how AOL (who thought it owned its customers) and cable companies (who thought they owned their customers) might not come to agreement on a business deal. Publicly, AOL dismissed the threat from broadband as in this comment from Bob Pittman (also reported by Saul Hansell) ten years ago.

"''Broadband doesn't sell itself,'' Mr. Pittman said in an interview. ''There is a small group that says 'Yeah, it's faster.' '' But high speeds don't make a difference to the activities that AOL members use most, he added, saying, ''Your E-mail doesn't get any better with broadband.'"

Maybe Pittman actually believed this. Or maybe this was a corporate posturing for the financial markets or for negotiations with cable and phone companies. Maybe it was all three. But when a charismatic corporate leader makes public statements like this, he runs the risk that his employees will believe it and act accordingly.

Absent a broadband strategy, AOL increased its bets on content, culminating in the ill-starred merger with Time Warner, which was a bold attempt to leverage a temporary advantage in dial-up internet access into permanent leadership as a content provider. But sometime this year, that deal will finally be unwound, leaving only bitterness and decimated Time-Warner 401(k) plans behind. More recently, AOL has let go of its paying subscribers (who were leaving in droves anyway) while shifting its emphasis to ad sales. But today AOL.com looks like a smaller and much less popular version of Yahoo. In the current ad recession, Yahoo -- the second most popular destination on the web -- has struggled to regain its former glory. Tim Armstrong faces the same challenge, but from 35th place.

Note 1. Time Warner announced that it repurchased Google's 5% stake on July 8 for $283 million, a price that includes "cash distributions owed to Google". Google made its investment in AOL in December 2005. According to this pre-spin off filing by AOL, net distributions from AOL to Time-Warner in 2006 through 2008 come to $2.85 billion (page F-5). Assuming Google was owed a pro rata interest in these distributions, the "cash distributions owed to Google" may have been as high as $150 million, suggesting that Google is accepting a mere $133 million for its 5% stake in AOL, which would imply a total equity value of $2.66 million, or 2.5x run-rate EBITDA (or OIBDA as AOL prefers it) based on the June quarterly OIBDA of $271 million (see page 19). This extremely low-sounding valuation may be plausible as subscriber revenue, which undoubtedly has a high contribution margin to OIBDA, shrunk by 27% over the prior year quarter. If subscriber revenue is halved from its June quarterly result of $356 million with no corresponding cost reductions, quarterly OIBDA would be just under $100 million, and a $2.66 billion valuation would imply a 6.6x OIBDA multiple.

Note 2. On July 30, Michael Nathanson of Bernstein Research published a report (subscription required) on Time-Warner. Nathanson estimates that "cash distributions owed to Google" made up approximately $100 million of the $283 million Time-Warner paid Google for its 5% stake in AOL. This would imply an equity valuation of $3.66 million for all of AOL.

Jul 8, 2009

Steve Jobs' Health is None of My Business

Bloomberg today reports the non-news that, "...disclosures about Steve Jobs’s health remain under scrutiny by U.S. Securities and Exchange Commission investigators over how his condition went from 'relatively simple' to 'more complex' in nine days" according to an unnamed source.

It must be vacation season if Bloomberg is trotting out a "situation remains the same" news story, which then recycles lots of commentary from legal experts not involved in the situation and doctors not treating Jobs. Predictably, nobody actually involved in the situation provided any on-the-record comment.

For the ultimate phone-it-in vacation season news story, go read this parody by Andy Borowitz. Like Mike Royko's traditional New Year's Day column or the Wall Street Journal's annual Thanksgiving editorial, Borowitz piece deserves to be republished every July 4th and Labor Day weekend.

Back to Steve Jobs' health.

Here's a simple fact that seems to get overlooked by the breathless journalists on this story: Like every other CEO I've ever met, Steve Jobs is mortal. Presumably rational investors in Apple factored this into their investment decisions years ago. Bloomberg is especially aware of this fact, having already published an obituary for Jobs back on August 28, 2008.

Here's another simple and widely reported fact. Steve Jobs announced to Apple employees that he had been diagnosed with islet cell neuroendocrine cancer of the pancreas back in 2004. Investors in Apple who previously overlooked his mortality probably factored this into their investment decisions then.

Now here's a more complex fact. Cancer, in all its forms, remains a complex disease. Nobody, including Steve Jobs and his doctors, knows what tomorrow will bring. As anyone who's had a close friend or relative battling cancer knows, the prognosis can swing from high to low and back again in mere days. Any near-term prediction about the course of someone's cancer is inherently speculative and as likely to mislead as to inform. If this were any other risk factor and any other company, it would suffice to include boilerplate language in the 10-K along the lines of "We maintain key man life insurance on certain of our senior executives but there can be no assurance that recoveries under these policies would fully compensate the Company for the loss of the executive's services."

Surely the SEC staff has better things to investigate, journalists can find more newsworthy stories to report, and we can all just say a silent prayer for Mr. Jobs and his family and leave them alone in this difficult time.

Jul 7, 2009

Buzzword Beat -
Mark Cuban on "Free"

Mark Cuban has published a new post on his weblog entitled,

When you succeed with Free, you are going to die by Free.

In the post, Cuban argues,

"Lets look at the rule that eventually KILLS all freemium based content plays:

There will always be a company that replaces you. At some point your BlackSwan competitor will appear and they will kick your ass. Their product will be better or more interesting or just better marketed than yours, and it also will be free. They will be Facebook to your Myspace, or Myspace to your Friendster or Google to your Yahoo. You get the point. Someone out there with a better idea will raise a bunch of money, give it away for free, build scale and charge less to reach the audience. Or will be differentiated enough, and important enough to the audience to maybe even charge more. Who knows. But they will kick your ass and you will be in trouble."


Any thesis in the sphere of economics that includes the words "rule", "kill" and "all" is inherently suspect, and Cuban's post reads like a breezy attempt to join the buzz-fest around the publication of Chris Anderson's new book, "Free, The Future of a Radical Price". The giveaway is Cuban's strained "BlackSwan" reference from Nassem Nicholas Talleb's book of the same name.

(Anderson's book is an expanded version of this Wired magazine article.)

As a number of commentators have already noticed, Cuban's thesis says nothing about "free" or "freemium" business models that is not equally applicable to any business. Yes, business is a hyper-competitive sport, someone will eventually -- no time frame given -- come up with a faster, better, cheaper version of what you do and "kick your ass." A vague and generally agreeable prediction -- hedged by the inclusion of "eventually" so not falsifiable in the abstract or the concrete -- is not worth arguing about.

Here's my take.

"Free" business models like Google's search engine, MySpace and Facebook are predicated on low, arguably zero, marginal costs. Facebook can afford to be the digital bulletin board for 225 million worldwide users only because the marginal cost of storage and bandwidth is very small. Equally important, the capital costs of storage and bandwidth decrease predictably in general accordance with Moore's Law. So even as Facebook's user base explodes, its technology costs (per user) are likely dropping by 15-20% annually.

What Cuban should have said about "free" business models is that the low marginal costs and declining capital costs of these technology-intensive businesses go hand in hand. And if the infrastructure cost of MySpace or Facebook declines by 15% per year, a competitor can replicate that infrastructure four years later at 52% of the original capital cost. With half the capital cost, a new competitor is tempted to compete on price (more, better stuff for "free") and win away the business. Even if the new entrant fails, it will likely compete away some of the incumbent's profits.

In his article Cuban seems to be thinking of social media businesses -- Friendster, MySpace and Facebook-- although he includes Google (which deserves an asterisk if only for its multi-year history of handsome profits). What's economically interesting about these social media businesses, allowing them "to raise a bunch of money" is often explained in terms of "network effects", a popular buzz-phrase for what economists call a positive externality.

According to Wikipedia, the concept of "network effects" was introduced in the early 20th century in the context of emerging telephone systems. The positive network effect (you have a phone, making it more valuable for me to have a phone) helps drive adoption of the new technology. But in the early 20th century, adding those additional phones meant stringing expensive wires to each house or place of business. And after the first entrant incurred that sunk cost, there was little incentive for a competitor to incur the same cost to compete for the same customers with the same service. Even if some of the installation costs declined over time (cabling, electronics, e.g.) the costs of rights-of-way, telephone poles and labor likely increased, allowing the first-mover to build a long-term competitive advantage based on network effects and sunk costs. This is why most telephone systems in 20th century became regulated monopolies.

Facebook's current membership, at 225 million registered users, testifies to the potency of network effects in social media. And with the physical infrastructure of the internet already in place, a network that would have taken decades to build in the last century can arise in mere months today.

That's what's frustrating Rupert Murdoch as Facebook has surpassed MySpace in popularity. It probably keeps Facebook CEO, Mark Zuckerburg, up at night as well. Network-effects businesses on the internet generally don't enjoy the additional competitive advantage of high and rising capital costs to keep new entrants on the sidelines. And the glue that holds the network together and preserves its value may be nothing more than "community," an economic intangible that's as fragile as it is powerful. If a new entrant gains sufficient traction through differentiation or well-funded patience (think Microsoft), it too will eventually enjoy network effects. As users defect from one community to another the winner's positive externality is the loser's negative one. If your friends have stopped updating their MySpace profile in favor of Facebook, you'll probably stop looking for them on MySpace. If your friends start posting their updates on Twitter, you'll spend less time on Facebook. After enough defections the market may reach a "tipping point" (buzz-phrase alert!) as yesterday's market leader becomes tomorrow's also-ran.

Clever entrepreneurs understand the fragility and potentially transitory nature of their competitive advantage if it's based primarily on network effects. Successful ones use their early competitive advantage period to build potentially more durable advantages based on technology and intellectual property. This is an important point where Cuban and I disagree, especially with regard to Google. At December 2008, Google's 20,000 worldwide employees included more than 7,000 engineers. Not a lot of venture capital money is flowing into startups to take on that army of programmers. Google's R&D investment strikes me as a rational strategy while Cuban sees it as a costly act of desperation.

Chris Anderson seems like a smart guy, so maybe he's covered all this in his new book. I'll have something to say about that when it's available at my public library where I'll check it out... for free.

Jul 1, 2009

Clueless in Chicago --
Unraveling Newspaper Economics

Major newspaper publishers met last month in Chicago in a not-exactly-secret, but definitely closed-to-the-public meeting to discuss the future of the newspaper business. More specifically, it seems, the meeting was convened to share ideas about how to more effectively monetize newspaper content on the web as traditional print subscription and advertising revenue plummet.

(As an aside, since most newspaper content on the web today is free, "monetize" must mean a price increase. Imagine if any other industry convened a closed meeting to discuss a price increase. What would the press have to say about that?)

As an avid reader (and paying subscriber) of several newspapers, I'm hoping they'll figure out how to survive. Sadly, I think the meeting must've been terribly disappointing.

A copy of "Paid Content - Newspaper Economic Action Plan" produced for the meeting by the American Press Institute (API) has been published on the web in several place including here. A quick read of the document cannot inspire confidence. The action plan recommends that newspaper publishers experiment with micropayments, subscriptions, and more ominously, coordinated industry and maybe government pressure on companies like Google, Yahoo and Microsoft to share search-related ad revenue. This last bit is called the "Fair Share Doctrine" described as "Negotiate for money, a lot more, from Google and online news aggregators for a 'fairer' share of the profits from linking and ad sales."

These ideas are neither novel nor untested. And they haven't worked so far. One gets a vague and uncomfortable sense that when the API recommends, "... [using] technology, news-industry production protocols, influence and public policy to thwart piracy" what they really mean is "maybe some government intervention can help us survive in our current form."

The 31-page API report goes wrong in its second paragraph when it asserts, "The problem is that the online business model does not yet come close to compensating for the steep slide in the print business model that it is replacing."

Guess what, it never will.

For well over a century the newspaper industry has enjoyed handsome returns from the economics of bundling combined with enviably low marginal distribution costs. These returns became even more attractive as many cities (in the U.S. at least) became one-newspaper towns. Bundled pricing, low marginal costs and monopolistic (or at least oligopolistic) market structure is a wonderful way to make a living. It is, however, not a birthright. And the government has no role helping the newspaper industry compensate for its loosening grip on its historical monopoly.

Here's Warren Buffet, whose Berkshire Hathaway has owned the Buffalo Evening News since 1977 and is a major investor in the Washington Post Company, writing in his annual letter 25 years ago:

“The economics of a dominant newspaper are excellent, among the very best in the business world. Owners, naturally, would like to believe that their wonderful profitability is achieved only because they unfailingly turn out a wonderful product. That comfortable theory wilts before an uncomfortable fact. While first-class newspapers make excellent profits, the profits of third-rate papers are as good or better - as long as either class of paper is dominant within its community.” [emphasis added]

Twenty-five years on, Mr. Buffet has changed his view of the newspaper business. During Berkshire Hathaway's latest annual meeting, he said, “For most newspapers in the United states, we would not buy them at any price...They have the possibility of going to just unending losses.”

What's Changed?
The structure of any market in equilibrium is determined by a complex and recursive interplay of technology, economics, inertia (in the form of pre-existing business relationships) and sometimes regulation. In the short term, the last three factors are paramount; in the long-term, technology dominates.

Traditional Newspaper Economics
The Virtuous Circle



Historically, the market structure of the newspaper business enjoyed a virtuous circle as depicted above. Once the sunk cost of the editorial staff is incurred, the printing press paid for, and the distribution system in place (collectively representing yesterday's technology), the incremental cost of including an additional classified ad -- or any other feature -- in the daily newspaper is negligible. Hence, newspapers had incentives to bundle many forms of content in addition to their own editorial content: TV listings, horoscopes, movie schedules, stock listings, comic strips, classified ads, etc. A reader paid for the bundled product even if he used the classified ads maybe once a year, or never read the horoscope or used the TV listings (note 1). The high fixed costs, offset by the surplus economics from bundling and low marginal distribution costs gave rise to something akin to a natural monopoly. And in most U.S. cities, the market leader has seen its competitors fade away in the post-war years (note 2).

Now imagine you're a newspaper subscriber (maybe you still are). If you could disaggregate the horoscopes from the weather from the sports from the local news from the international news from the business news from the TV listings from the almost non-existent stock price listings, how much would you pay for the parts of the paper you actually intend to read? Probably less than the $10-$15 per week it currently costs at the newsstand. Probably less than the $6-8 per week it costs to subscribe.

Probably a lot less.

This is the problem faced by the newspapers. Bundling is a pricing strategy that delivers surplus economics to the supplier by enticing customers to buy more than they would if the bundled products were sold separately. By weight, the majority of your local newspaper (and its website) is information sourced from third parties (ads, stock listings, classifieds, lightly edited excerpts of corporate news releases, etc.) readily available elsewhere on the internet (note 3). By allowing readers to disaggregate the newspaper's traditional bundle of content, the internet may be exposing the market value, or to use the API's term "true value" of the original editorial content produced by the publisher itself.

As publishers experiment with revamped online pricing models they may find that the true value of their original content will give horrifying meaning to the term micro-payment. No newspaper has a monopoly on "the news." It certainly has no monopoly on the third-party information it republishes. The newspaper industry suffers from a notion that it should enjoy monopoly economics on content ("Hey, that's copyrighted!") when in reality its historical monopoly was control of a distribution channel and much of the profit was based on aggregating and organizing other people's content. In the internet age, that distribution monopoly no longer exists and others, like Google, do a pretty good job of aggregating third-party content.

Copyright should be respected. But if a reader can get his daily dose of international news as readily from the Washington Post, the New York Times or a foreign newspaper, copyright on a particular rendition of the news will not give rise to monopoly economics.

Like the music industry before it, the API's view of the newspaper industry confuses the surplus economics arising from bundling and distribution monopolies for the natural economics of their copyrighted content. Copyright does indeed confer a monopoly right to a particular form of expression, but in no way guarantees that consumers will pay handsomely for it, if at all. The music industry has spent the past ten years battling piracy when the larger economic problem has been the unbundling of the album format. It turns out that customers prefer to pay $1.29 for one song they really want rather than $14.99 for the twelve songs the label bundled on a CD album. Losing the additional $13.70 per transaction really hits the music label's revenue line. A twelve-year old kid downloading thousands of songs he can't otherwise afford does not.

If newspapers no longer command a monopoly on distribution and can expect no surplus economics from bundling third-party content -- including ads -- they may find that the ratio of the "true value" of their editorial content to their historical revenue approximates the 20% of the average paper that is made up of original content.

When new technologies completely undermine an industry's market structure, that industry needs to be rebuilt from the ground up. The newspaper industry will fumble along (much like the music industry) if it starts from the premise that its historical economics represent some kind of natural order.


--------------------

Note 1. It should be noted that subscribers to print newspapers generally pay less than the actual cost of writing, editing, printing and delivering the newspaper... often a lot less. What drove newspaper profitability in the past was advertising sales, but that requires the aggregation of a large audience, which requires aggregation of diverse content to appeal to a large, diverse audience to attract the advertisers.  Another virtuous circle... or vicious cycle if it starts running in the wrong direction.

Note 2. Noam, Eli M., "Media Ownership and Concentration in America"

Note 3. This morning's complimentary San Francisco Examiner landed on my front porch despite my wife's repeated attempts to discourage them from delivering it; I guess they need the circulation numbers to support their advertising rate base. The paper, including ad inserts, totals 54 pages. A quick inspection shows the content is allocated as follows:

Third-party content
Ads: 34.5 pages
Classifieds 4.0 pages
Movie listings 2.0 pages
Weather 1.0 pages
Games 1.0 pages
subtotal 42.5 pages

This leaves 11.5 pages of news content, but of course the "World", "Nation" and "California" sections (one page apiece) appear to be entirely made up of syndicated pieces by the Associated Press or others. So the actual original content produced by the Examiner comes to about eight pages, or about 15% of the total newspaper.