Monday, August 4, 2008

Rationally Pricing Shares of Apple

Folks eager to place valuations on Apple have done all kinds of gyrations based on the predicted earnings and the PEG ratios that are reasonable for assumed earnings growth rates. Being a skeptical consumer of such argument, I have a more basic question:
How do you figure Apple's earnings?
What Are This Quarter's Earnings?

On June 11, shortly into the last quarter of Apple's fiscal year ending in 2008, Apple released a new product: the iPhone 3G. This gadget sold fairly well, selling out initially in many locations and demanding so much performance from carriers to activate or modify buyers' accounts that activation was problematic in some countries.  A headline at Seeking Alpha gushed that "Apple's iPhone [was the] Biggest Consumer Electronics Launch Ever[.]" The article made comparisons to historic launches of consumer electronics -- Betamax, Windows 95, XBox, and the original iPhone -- and concluded that Apple had crushed every other consumer electronics launch ever, bagging some $433 million in sales in the product's first weekend.

There are some tear-down estimates that the 3G iPhone costs Apple less to make than the prior iPhone, with guesses ranging from $100 to $173. Assuming the thing's cost to Apple is $173, and Apple's gross receipts per phone are $433 (surely inaccurate, given the nature of its author's calculations) a rough take on per-unit profit might be something like $260 apiece.

So, Apple's profit for a weekend in which it sells a million units should be something like $260 million, right? Ahh, no. Not even neglecting taxes.

This is because Apple is lying about its earnings. The rule beneath which Apple fibs on its iPhone earnings is the rule governing "subscription accounting". If Apple were to sell a magazine subscription for $48 for a two-year subscription, you'd understand that it would not be entitled to book the whole $48 as revenue when it was paid up-front by a subscriber, because Apple would have 23 more issues to send out. Everyone can tell the magazine publisher needs to send all the issues to earn all the money in a 2-year subscription for a monthly magazine, and it's easy to tell how much is left unearned. So it's no surprise that magazine publishers are allowed to claim that the $48 isn't all revenue -- and isn't all subject to tax during -- the first month in which it is received. The magazine publisher books the cash as received, but books a negative asset representing the stuff still owed to customers, and by adjusting that later number slowly realizes the $48 over the life of the subscription.

Sounds fair, right?

Apple, when it takes money for an iPhone, parts with one whole entire iPhone and its pre-installed software. If Apple never sent the customers another thing, I wonder what court would find Apple owed anything more. It's not at all like the magazine subscription, in which one can immediately tell that twenty-three issues have yet to be delivered: once the customer has functioning hardware and software and is activated with a cellular service carrier, the stuff that's left to be delivered is de minimis if it exists at all. What asset does Apple really owe customers after they've activated the phone and seen that it works as advertised? Why should Apple not be seen as realizing the whole sale the day it's made?

When Apple first introduced subscription accounting for the iPhone, it was pretty clear that Apple was using its ongoing service revenue sharing and the fact that without service the product wasn't fully delivered to claim that Apple should be entitled to stretch the phone revenues' recognition across the life of the service plan Apple's customers were required to adopt. Apple, after all, pre-arranged the service plan for its customers and was a participant in delivering the services through its exclusive carrier partner. Even though Apple had pretty much delivered everything, it certainly hadn't received all the revenue, and it had some (at least arguable) performance left in supporting the carrier in handling visual voicemail through a package that ran on the carrier's own network.

Now, however, there are multiple official and approved carriers in the world -- some of them, non-exclusive in some countries and subject to competition -- and Apple isn't receiving a share of service revenues. Apple receives for its phones a payment by carriers who subsidize the cost of the phone to their own customers as an incentive for the customers to adopt high-end cell plans. Apple isn't delivering something incomplete any more than is RIMM or NOK. If RIMM or NOK aren't entitled to use subscription accounting for handset sales, why should Apple?

Why does this matter?

When Apple says it's entitled to use subscription accounting to spread a phone's revenues over 24 months, it's saying that in the 3-month quarter in which a phone is sold it will book revenue for somewhere between 1/24 and 1/8 of the actual sale made in the quarter, and will amortize the rest across the rest of the 24 months. Rapid sales growth, therefore, won't show up in Apple's quarterly reports. Apple will be able to hide the true extent of its profitability from onlookers, who will be forced to guess about Apple's per-phone price to resellers and the extent of its phone subsidy from carriers.

This might help Apple stave off competitive pressure from parties who would like hard data on Apple's performance as a fulcrum against which to lever future negotiations, but it's hell on investors who are trying to work out just how profitable Apple's sales growth is.

Did Apple take a beating clearing old iPhone inventory in the June quarter? Who knows? It's diluted in all the prior income from old sales made when the iPhone was a hot new item.

Will iPhone 3G sales make as much money as iPods? Who knows? iPod profits, though booked in the quarter of the sale, are mixed into a music business performance result that includes music sales from an online store that, despite surpassing other music retailers and becoming the biggest music retailer in the US, is described as a near-break-even business by Apple execs -- while the iPhones' profits are amortized across two years, which in the high-fashion world of high-end cell phones may be longer than the service life of the phone model that was just sold.

In short, the effort to ensure competitors can't work out the details of Apple's deals means that investors and analysts can't do it, either. Spreading revenue over time and mixing it with non-iPhone revenue streams (e.g., Apple TV) means that Apple has made it nearly impossible to work out what its profit is from the business to which many onlookers cast their eye for profit growth.

The Problem.

With profit (and profit growth) an enigma, pricing Apple shares has become risky. Risk isn't good for those holding shares. You know that the minute the unit sales number drops, folks will run for the doors, so you can't believe for a minute that Apple's subscription sales formulae have actually created a belief in the validity of deferred revenues at Apple. You also, therefore, can't accept the claim that the accounting principle will help smooth seasonal revenues: everyone knows Apple's business is highly seasonal and responds to gifting occasions and academic calendars, and will expect performance accordingly.

Apple's only real effect in using subscription accounting (enhanced by Apple's simultaneous practice of mixing different types of revenue streams into its reported product categories) is to make it virtually impossible to tell how much money Apple is making on iPhones.

But, who cares?

Folks who try to work out the price at which Apple shares "should" trade like to look at multiples like the PE (price-to-earnings) ratio, and the PEG (price-to-earnings, divided by per-share earnings growth) ratio, to see whether the stock is overpriced or is a bargain -- or to project future prices using plausible reference ratios. Once the true earnings for a quarter becomes a black box, and the growth in the earnings from the prior-year's quarter is rendered meaningless by the use of subscription accounting in a small number of high-growth market segments, both PE and PEG will cease to reflect the price of the company in relation to the company's actual performance.

To get a "real" PE or PEG, one would need to use cash flow numbers and other pieces of the quarterly statements to try to "back into" the real numbers for profit and growth. This, of course, is a risky endeavor: the whole point of the Apple accounting and reporting strategy is to obfuscate internals from detailed inspection by outsiders. While this may make it hard to work out how profitable the Apple Store is, or the iTunes business, it also makes it completely impossible to say much about Apple's PE or PEG going forward, assuming the iPhone is a material part of Apple's business.

If the iPhone isn't a material part of Apple's business then this chicanery is for naught. Assuming the iPhone is significant in terms of sales and profits, though, a future whose profits and cash flows result from an admixture of accounting principles and deferral practices will mean a future in which both PE and PEG ratios will be available either based solely on Apple-engineered accounting fictions (that will understate earnings when there is level or positive earnings growth, and overstate them in periods like last quarter, in which products were refreshed and much of the quarter was spent without stock on the shelves; and consequently will also misstate earnings growth because the earnings are falsely smoothed across eight quarters). To get "real" numbers will require -- let's be honest here -- guesswork that works against the very principle of acquiring solid evidence of stocks' value.

Apple's profits from Macs may be soaring, and the iPhone may be flying off the shelves, but Apple's accounting practices aren't going to help anyone perceive the shares as a buy even if they are. Given the impact of oddball accounting and hyperconservative forward earnings guidance, there's really no way to be sure.

The Upshot

Despite excitement in recent years over Apple's music business and its entry into the phone market, Apple's profits at present come largely from Macs. This is good news: the music business may enjoy more music sales, but the iPod unit growth seems to have come to an end. Yes, last quarter iPod sales showed an increase compared to the year-ago quarter, but it was also a decrease from the prior quarter. It was also impacted by super-low-priced iPod Shuffles, which are a "unit" though the profit in them is scant. Apple needs the business to prevent competitors from establishing a beachhead, and Apple does profit from the things, but it's no iPod Touch. So growing units with Shuffles might be "growth" under some definition, but it's not the kind of profit growth that makes investors weep with joy.

The fact that Mac sales are growing means the platform is doing well -- and Apple, as a major developer for the platform, stands to make good software revenues. The platform's growth is good for Apple also because it's darned near the very same platform that the iPhone runs (and the iPod Touch), leading to some potential network effects. Games made for the iPhone may yield development for the Mac, for example. Developers may try developing multilayer games in which iPhones or iPod Touches are used as remotes to control actions in a multiplayer game run on a Mac, or in the cloud, or tracked by the players' clients in an ad-hoc fashion. Wild things might be in the pipeline.


What we do know is that Apple's iPhone growth is unknown (it's just entering dozens of jurisdictions) and the segment of its business with known recent high growth is about a third of Apple's business. Big sales in low-profit-growth areas -- like the music business has become -- effectively acts as a brake on metrics like PEG.

Apple is likely to turn in some more good Mac numbers over the back-to-school and holiday quarters -- benefiting Apple's per-share metrics immediately -- but virtually anything Apple does in iPhones will impact share price solely because of buzz and future sales anticipation, not near-quarter growth metrics. This isn't because Apple can't make money on phones, but because Apple has chosen to report numbers that won't help observers learn what Apple makes when it sells phones. We'll be left with analyst projections and the backs of our own napkins.

If you've read my piece on analysts and on non-analysts, you will understand why I suggest one stick with the backs of one's own napkins.

No comments: