Categories

Hey Big Spender

In previous posts I described the patterns of asset value growth for Property Plant and Equipment at Apple and how that change has been thus far correlated to the production of iOS devices.

The analysis was based on looking at the declared value of the PP&E assets on Apple’s balance sheet. These values include depreciation so they reflect not only the “spending” on new assets but also the value lost due to wear and tear. As such it’s not a perfect measure of investment.

The better approach is to use declarations on the cash flow statement. Apple reports a specific cash flow related to PP&E: Under Investing Activities, Payments for acquisition of property, plan and equipment.

In the latest 8-K the amount for the 12 months ended September 29 the value was $8.295 billion.

Compare this with the statement in the Annual Report from one year ago:

The Company anticipates utilizing approximately $8.0 billion for capital expenditures during 2012, including approximately $900 million for retail store facilities and approximately $7.1 billion for product tooling and manufacturing process equipment, and corporate facilities and infrastructure, including information systems hardware, software and enhancements.

Compare also with the PP&E net asset value on the Balance Sheet currently at $15.452 billion but valued at $7.777 billion a year ago (an increase of $7.7 billion).

This means that Apple intended to spend $8 billion, actually spent $8.3 billion and realized a net asset gain (after depreciation) of $7.7 billion.

So there are no major surprises. The spending was only about 4% above expectations a year earlier.

The other pattern to observe is that the spending rose dramatically into the fourth quarter. The fiscal quarterly spending was:

  1. $1.32 billion
  2. $1.46 billion
  3. $2.06 billion
  4. $3.46 billion

I illustrate this pattern (and previous quarterly spending) in the following chart.

The question remains what is this $8.3 billion being spent on?

To help answer I also added the comparable spending (from cash flow statements) of a few other technology companies. Google, Amazon and Microsoft have been spending in the same range of $500 million per quarter. Over the past eight quarters the average spending of these companies has been:

  • Amazon $487 million
  • Microsoft $583 million
  • Google $800 million

Intel is different. Its average spend level has been $2.6 billion/quarter.

But Apple has just spent about 30% more than Intel. And it spent three times in Q3 than Google, 4.8x Microsoft and 3.8x Amazon.

The level of spending and growth from a far lower level implies to me that Apple is spending primarily on process equipment (aka tooling) rather than servers. Data centers are expensive at $1 billion for a large one, but they are not being built fast enough to account for $3.5 billion in one quarter.

Note that Apple’s CapEx is above that of Intel whose fab-based semiconductor production method is known as particularly capital intensive. And Intel’s Capital Expenditures have been increasing to all-time highs.

If Apple spent at Google levels on server infrastructure it would only be spending a quarter of what it’s actually spending. If Apple spent on production at Intel levels it would only be spending three quarters of what it’s spending.

In other words, Apple is spending at the equivalent of one Intel of fabrication and one Google of server infrastructure.

In a few days we will learn what the plan for expenditure for fiscal 2013 is. Historically, the company has roughly doubled spending for each year of the iOS platform. If that continues then Apple will be spending the equivalent of several Intels and several Googles.

  • gprovida

    If Apple continues its investments in manufacturering with suppliers, this will create a widening gap in device technologies being sold and will be very hard to overcome, case in point, most competitors are struggling with aluminum chassis for laptops, desktop, and mobile devices. The Apple shift to other supplier from Samsung, although I am sure thy will buy from Samsung, they will not invest in Samsung. This creates a long term strategic advantage to Samsung supply chain competitors.

    It is interesting that reviews of iPhones and iPads, especially the 5 and Mini reflect build quality and I suspect costs that are a continuing competitive advantage, while the press seems more focused on Foxconn assembly it is the manufacturing that is the real competitive advantage.

  • http://www.isophist.com/ Emilio Orione

    Intel owns their manufacturing implants, intel is investing in their own structure. The same for goggle.
    Apple’s investing is different, while server infrastructure is proprietary, manufacturing is not. This is unprecedented, a company investing in its supply chain much more then anyone else is doing in their own infrastructure.
    With this level of spending, why apple does not build its own manufacturing facility to compete with samsung?
    They have the needed volume in their own products without having to produce for competitors. They have already integrated for processors, why outsource production if outsourcing requires this level of spending?
    I don’t understand.

    • http://www.asymco.com Horace Dediu

      Be careful with the word “investing”. When Apple purchases equipment it owns that equipment. It has title to it. That equipment is probably used exclusively for its own production. It then depreciates the equipment and expenses that depreciation. The contractor may own the facility it’s used in and may employ the workers, but that piece of equipment belongs to Apple. In that sense Apple is not “investing” in any contractor. It has no equity in the plant and no relationship or responsibility for the workers. Apple controls some of the means of production but not all. The spending buys them capacity, predictability and exclusivity. These are what were missing from a pure outsourcing relationship. It does not buy them complete control but enough control. They do give up some margin to the plant operator but that margin is offset by the risks of employing people and dealing with local laws.

      • http://www.isophist.com/ Emilio Orione

        Cool a new form of outsourcing without the downsides, but it depends on the deals with contractors. If things are as you say its like investing in equipments and outsourcing labor. Good, seems like a great bargain.
        Question is, why isn’t intel doing the same or samsung itself?

      • RobDK

        ‘Question is, why isn’t intel doing the same or samsung itself?’

        Good question!

        Samsung is probably not doing this because they do not have a 5 year product development road map. They just copy what Apple does, as much as they can. They are well known as a ‘first follower’, so they cannot do anything until others have done it first. They are also hamstrung by Android – they do not get exclusive knowledge of the OS, they have the same developer access as others in OHA. So they have no incentive to develop exclusive hardware/software integration.

      • Tatil_S

        Samsung has its own mobile CPU development team, fabs to build chips, fabs to manufacture LCD screens and God knows what else. I don’t know whether Samsun Electronics documents its CapEx in detail, but if it needs capacity, predictability and control, it already has it.

      • http://www.facebook.com/people/Shameer-Mulji/1685212657 Shameer Mulji

        You make a good point. Right now, there’s serious shortage on all of Apple’s new iOS devices & a lot that is attributable to its screen suppliers not being able to ramp up screen production fast enough. A week or two ago there was a article, stating from a Samsung exec, that it would stop making displays for Apple by some time next year. Hopefully Apple’s suppliers can not only ramp production but produce them in good quality, otherwise this is not a good move. On top I read the following post (link provided below) saying Apple’s fight with Samsung may hurt Apple more in the long-run.

        http://semiaccurate.com/2012/10/23/apple-vs-samsung-samsung-put-the-boot-in-hard/

      • RobDK

        Yes, but they do not have the crucial hardware/software integration in this development process. It really does not help Samsung with all this hardware focus if they do not have deep hooks into the OS at very early stages of the development cycle.

        A common theme with the iOS is how little RAM is needed because of Apple’s design skills in software and hardware. Whilst the competition sees it as a marketing advantage to boast about large amounts of RAM, Apple uses less to reduce energy consumption and save space/weight.

        There is also evidence regarding Apple’s integration of A/V functions onto the SoC, and how they deal with graphics and the GPU, which give major advantages in the performance/battery life ratio, and which can only be achieved with a long and deep integration of hardware and software.

      • Walt French

        These are the economics of yesteryear.

        RAM is relatively cheap — in dollars, SOC real estate and battery requirements. I you need 2X the RAM to offset a couple thousand man-years and hundred million dollars of OS investment, you’re WAY ahead by paying up for the RAM AND staffing up for the OS enhancements.

        Likewise, Android’s virtual machine was optimized for quick availability first, then better performance second (v2.2); Google is working hard to move performance-critical functions into native code so interactions are close to as smooth as in iOS. Again, it got ‘em into the game quickly — more quickly than Apple’s approach needed to / was meant to.

        The fact that Samsung is (currently) at the forefront of ARM CPU capability shows a savvy investment in tech that is real world.

        I’d give a bit more attention to Sun Tzu’s advice about knowing your competition better.

      • Tatil_S

        Having a less efficient OS requires more RAM, higher freq CPU etc. which all degrade battery life. Skimping on OS development may be the fastest route to the market and the savvier business choice if time is of the essence, but the trade-off is not limited to “slightly higher variable costs vs. higher fixed costs”.

      • Walt French

        All you say is true. And reasonable. Which *I* was trying to do, too.

        Apple’s longstanding emphasis on software solutions that minimize hardware needs has both pluses and minuses against the more brute-force approach (high-speed, battery-hungry CPUs, extra memory, code interpretation) that the competition uses. Smartphones and mobile tablets have been, and will continue to be, an incredibly dynamic industry, which is EXACTLY the environment where flexibility and rapid pivoting are most valuable; Moore’s Law is mere icing on the cake.

        Any reasonable competitive analysis would list iOS as a tremendous asset, as is Apple’s recent ramp-up in semiconductor technology. But other firms have assets, too, and it seems incredible to wish competitors’ talents away with statements such as, “They just copy what Apple does…they cannot do anything until others have done it first.” Especially when that competitor is successfully selling a huge number of first-class devices, it would be foolish of Apple to dismiss a competitor with false bravado.

      • RobDK

        OK Walt, I admit I was a bit flippant in choice of language, maybe I have read too much at businessinsider.com… And I agree with your comments on Android development and optimisation path.

        However, I do believe my original comment stands; it is difficult for Samsung to invest now in hardware/software integration, and begin buying up capacity further down the line, when they do not have deep insights into, and early access to, Android OS development.

        As long as mobile development is hampered by battery technology, and there is a design imperative for lighter and volumetrically smaller devices, then the optimization of hardware/software integration is going to give a competitive advantage. Apple has this advantage over the Android ecosystem.

      • Walt French

        Samsung has at least three areas with which to integrate software & hardware:
        1. Their home-grown Baidu,
        2. Their membership in the OHA, where they work with Google on low-level integration, and
        3. Rumors of an Android fork.

        Google is staying pretty hard-core about Android being widely available for OHA, which I take as their choice to keep Samsung from thinking they’ll do #3. Nonetheless, the fact that Amazon (a company with ZERO consumer OS software experience) has built a customized version with hardware, means that Samsung obviously has more than enough resources to undertake the same.

        And I would only be mildly surprised, if recent trends of profitability in Android handsets continue, to see #3 happen. Like Google and Apple, any big business in this area competes with its partners, and Samsung needs to have assurance that a key supplier (Google) doesn’t step on its airhose.

      • http://www.facebook.com/people/Shameer-Mulji/1685212657 Shameer Mulji

        “Any reasonable competitive analysis would list iOS as a tremendous asset…..”

        Which makes it all the more worrisome with Forstall out of the picture at Apple. He was the one that championed using OSX as the foundation for iOS and provided the leadership to bring it to where it is today. For Apple’s sake, I sure hope Federighi is up to the task.

      • http://www.asymco.com Horace Dediu

        Don’t believe anything you read in mainstream media about any executive from any company, ever.

      • Walt French

        Samsung just announced the first mass production of a chip using the “A15” ARM architecture. The phone that uses it features an industry-leading speed score for browsing, notably faster than the iPhone5’s A6 CPU gets.

        Both of those factoids suggest that we are approaching a time when phones are “good enough,” with all the industry implications that would have.

        And that your recitation of the “fast follower” story is not the most important way to understand the industry.

      • Folke Sonin

        Because of the nature of semiconductor manufacturing. Intel produces wafers that are cut into chips. Each new generation of technology roughly doubles the amount of chips that can come from a single wafer. Also, most of the manufacturing steps are completely automated and done in a hard vacume. So semiconductor manufacturing is very capital intensive but requires very few highly skilled machine operators. It is a very different business than final assembly of a consumer electronics item.

      • http://www.facebook.com/people/Carlos-Carbajal/620312515 Carlos Carbajal

        wrong. Semiconductor fabs employee several thousand workers per fab . . . it’s very labor intensive and not as automated as you would think. Most of the labor is covering 24-hour shifts as the fab is a 24/365 operation in most cases but running the machines is only part of the picture . . . fabs have their own IT departments and there is huge facilities and safety infrstrcutre that is required.

      • Tatil_S

        A fab is said to cost $7 billion if built from scratch and requires a couple billion dollars of investment every few years to keep up to date. 2000 employees, each earning $100k including benefits would cost $200 million annually. There may be a lot of need for labor in an absolute sense, but it is still a capital intensive industry compared to final assembly of gadgets.

      • http://twitter.com/handleym99 Maynard Handley

        The issue is not only the number of employees, it is the level of skill required.

        I know a high tech manufacturer that does the “hard” manufacturing (creating and processing the wafers) in the US, and the “easy” manufacturing (packaging) in Asia. This is a deliberate split, based on things like how the “hard” manufacturing uses large numbers of engineers, and these people are available in and want to live in the US, while the assembling requires mostly semi-skilled labor.

        Compare with companies that work in Silicon Valley where, likewise, salaries are far higher than say starting a SW company in Cleveland, but Silicon Valley is where the talent pool is, and if you require that talent, that’s where you have to be.

      • http://twitter.com/handleym99 Maynard Handley

        Intel does not use massive amounts of semi-skilled labor, so they have different incentives from Apple.

      • mieswall

        Horace,

        what’s happening with depreciation costs? Eventually a correlation with capex and depreciation could give some clues on what kind of capex expenditure apple is doing. Perhaps if investments are mostly in machinery, depreciation should be a higher proportion of previous capex’s.

      • Tatil_S

        Why do you think machinery depreciates fast? For example, CNC machines can last for a very long time. I’ve seen twenty year old basic units being still traded. The machine parts that wear out on routine basis might be considered costs and not even show up on the CapEx. I doubt you can back calculate the types of expenditures based solely on depreciation.

      • Shane

        A)This discussion is based on depreciation from a Finance perspective. Depending on the cost accounting methods that Apple is using they will depreciate all machinery they own likely within 3-10 years. This has nothing to do with the actual functionality of the machinery. It is simply common practice accounting principles – often 20%/year is the figure used,
        B) Even if the machinery is technically still viable and working that doesn’t mean that it can be used in new production as the designs/tech requirements of what Apple wants is moving quickly thereby rendering the machines useless for future production. And they’re not likely to sell these machines to other companies later on as they don’t want that tech being used again on lesser products by lesser companies.

      • Tatil_S

        5 years later, I doubt there will be much others can learn from Apple’s second hand machines that they cannot learn in the open market.

        I was also thinking of 10 years or so for the machinery and in my mind that is not fast depreciation, but Staiger’s comment knocks the wind out of that line of thinking.

      • http://twitter.com/JoshStaiger Josh Staiger

        Apple states in their 10-K:

        “Depreciation is computed by use of the straight-line method over the estimated useful lives of the assets, which for buildings is the lesser of 30 years or the remaining life of the underlying building; between two to five years for machinery and equipment, including product tooling and manufacturing process equipment …”

      • Tatil_S

        Thanks, I did not know Apple specifies “2 to 5 years” for depreciation of such equipment.

      • cellojoe

        What do you think the latest 10K portends? Only 10 billion for the year.

      • http://www.asymco.com Horace Dediu

        2012 saw far greater spending than anticipated. If you assume the excess (over $2 billion) was carried forward for unanticipated reasons, and assume the usual over-spend seen in the past then the y/y growth for 2013 is about 65%. Not the 80% to 90% of past years but it’s still a significant increase from a large base.

  • http://alexandersmith.co/ Benjamin Alexander

    Apple is spending at a greater rate than the Apollo program: http://benjaminalexander.tumblr.com

    • http://twitter.com/JessiDarko Jessica Darko

      … and in contrast, Apple is having a positive effect on humanity. Unlike the Apollo program, which was cold war dick swinging (“we can put a nuke in your tiny country in 24 minutes, and here’s the proof”) that only produced PR and Tang, Apple is advancing the state of the art forward. The apollo (and NASA) budgets go into one-off un-economical programs…. and all of that money has to come out of the economy.

      Every dollar NASA spends is a dollar some business didn’t invest in new products or some consumer didn’t spend to better their lives (creating demand and growing the economy.)

      If you dispute this, read Bastiat’s “Parable of the Broken Window” or Henry Hazlitt’s “economics in one lesson”.

      The people who support NASA and other boondoggle government programs account for the alledged benefits and are quick to make them up– but they pretend like the costs are zero.

      If we hadn’t had NASA, we would have had the pace of growth of high technology accellerated by 20-30 years. (EG: the iPhone would have happened in the 1980s.)

      • http://geekfun.com/ Erik S.

        Complete and total bullshit. Yes, war is incredibly wasteful and destructive, but there almost certainly would be NO silicon valley, or the equivalent, without decades of government strategic industrial policy investment in the guise of military and space spending. What is certain is that the US high tech industry was underwritten by massive amounts of government spending over decades before commercial and consumer spending could generate enough revenue to fund the pace of change we’ve grown accustomed to. On the other hand, you have your fantasy of what might have been, if only things weren’t as they are.

        If you dispute this, you need to realize that your understanding of the totality of human behavior is simplistic dogma. Unfortunately, your chances of having such an epiphany are slim to none at this point.

      • http://twitter.com/JessiDarko Jessica Darko

        What you call “bullshit” and “simplistic dogma” is actually the results of science. Specifically the science of economics. Alas, since politicians like to spend money to buy votes, and the space program was a great way to spend other people’s money to buy votes, they pretended like it was “helping people” and the economically illiterate, such as yourself, fell for it. Thus, ironically, you post bullshit nonsense here based on your simplistic dogma, in rejection of, and ignorance of, the science of economics.

        Worse, since you have no knowledge or understanding to work from, no amount of argument or logic or facts can reach you, making debate impossible. You’re just giving knee jerk ideology.

      • Kent Crispin

        Calling economics a science dooms your argument from the start.

  • RobDK

    Apple is paying an as yet unnamed firm, probably TSMC, to build/tool/operate one or several ARM SoC fabs. As such this does require investments on par with Intel’s. These fabs are designed to ramp production up to +500 million Apple designed Ax SoC’s per year for the many iOS devices. This is going to hurt Samsung.

    On top of that comes the large investments in screen, battery and casing production. All these factors will give Apple a 5 year technology lead which can be continually updated and re-invested in to keep Apple continually ahead.

    With current macro-economic concerns in the global economy, and a slowing down evident in Europe and China, the only question is whether Apple will slightly reduce the pace of investment, or whether they will use this opportunity to use their surplus cash holdings to gain strategic advantages, and be ready when the upturn in the global economy starts.

    • Walt French

      An interesting question is how much of the $2.6 billion allocation is to cutting-edge silicon fabs and how much goes to maybe-not-cutting-edge, huge-volume assembly.

      The former seems awfully important, as I’m not aware of any body with idle capacity for a couple hundred million CPUs a year, and the hold-up value that Samsung has as a consequence of it. The latter seems to address the frequently-reported issues with constraints on new models.

    • http://twitter.com/WalterMilliken Walter Milliken

      Certainly the story that Apple tried to buy exclusive access at TSMC is suggestive. And the kind of money they’re now spending, it’s a bit hard to see what they’d be buying. Semiconductor fabs are one of the few places you can easily spent that kind of capital. Or nuclear plants, but I doubt that’s part of Apple’s “green energy” approach….

      • http://twitter.com/WalterMilliken Walter Milliken

        Another thought — I wonder if Bob Mansfield’s new “Technology” division is some kind of skunkworks leading to an Apple silicon fab capability? Right now there aren’t very many cutting edge fab facilities Apple can make use of, and possibly none that they can trust for their volumes. (The ones I can think of: Intel, IBM, Samsung, TSMC; there are probably others.) Maybe Apple is thinking it’s time to revisit the “fabless” semiconductor model that’s the current “everyone knows this is how you do it now”.

        Or it could be they’re just extending their outsourced manufacturing with own equipment. But would that need Mansfield? And if it doesn’t, what *is* he doing? One other thought I had here is that maybe they’re going after integrating the baseband on the main CPU chip — something that Qualcomm is doing, and NVIDIA, Broadcomm, and Intel seem to be going after. Messing with the radio stuff would be a big reach, but something Apple could probably manage with a lot of work. And it would give them another potential way of differentiating, as well as shrinking components and possibly saving more power. Right now they’re stuck with whatever they can buy from Broadcom, Qualcomm, etc. Which everyone else also uses.

        I also wonder about their just-reported $1B 2012 R&D expenditure — R&D shouldn’t scale much with sales volume, it should scale with number of products and their design complexity, mostly. Apple has only added one new product line — the iPad mini. And maybe the burst of new redesigns raised R&D a fair amount. But it still seems a bit high to me.

      • N8nnc

        Apple has only announced one new product. Who knows how many they are working on?

        I too was struck by the merging of wireless and semiconductor teams. I take it as a sign that Apple is “all in” on wireless and will have the volumes to lead by de facto standards.

        Forstall’s exit jibes with iOS being their recent weak point, at least relative to the superlative performance of other areas.

      • http://twitter.com/handleym99 Maynard Handley

        Look at Intel’s Rosepoint software radio system…

        This is REALLY interesting, powerful stuff. It would make perfect sense for Apple to at least look into whether they could do the same thing rather than having to pay Intel (and accept Intel schedules, and the overhead of an Intel system that does more than they want, and is a separate chip rather than a cell in a SOC).

    • Fred Stein

      Thanks so much. This is huge. I stumbled onto this just being curious. Do any mainstream analysts discuss this.

  • r.d

    according to iMac article in wired.
    Friction Stir Welding is an aerospace tech
    invented by Wayne Thomas at TWI Ltd, a U.K.-based company.
    so Apple has to license the patents.

    Plasma-enhanced chemical vapor deposition also is a aerospace tech.
    Carbon Fiber is as well.
    Glass-LCD Lamination is all new.
    Robots to do the Chamfers.

    Coloring of Aluminum.
    Robots to do the assembly is the final frontier.

    • http://twitter.com/__MarkW__ Mark Wilcox

      >> Robots to do the assembly is the final frontier.
      Actually fully automated assembly lines for phones have been around for several years. It turned out it was actually cheaper to pay people in places like Mexico and China to do the assembly than run and maintain the robots. Not sure how much the economics has changed since I last heard but I expect it’s still true.

      • r.d

        NeXT had a fully automated factory in California in the 90s.
        Problem with robots is that it couldn’t do small detail
        work of individual soldering or putting on tiny resistors and such.

        As Foxconn is learning, they can no longer afford to have 200,000
        employees without problems. most probably only last a year
        or two before burning out.

      • Walt French

        Pretty sure that times have changed quite a bit. I’d be surprised if a human being could mount, and hand-solder, even one iPhone circuit board a week.

        Take a look at some of the teardowns. I know my eyes aren’t what they used to be, but many of these parts are on the order of 0.01″.

      • http://twitter.com/WalterMilliken Walter Milliken

        I believe all circuit board-level assembly is completely robotic now — projects I worked a couple years on that fabbed only a half-dozen prototype boards still had to be made by robot pick-and-place machines.

      • http://twitter.com/JessiDarko Jessica Darko

        As of at least 4 years ago, Robots could accurately, and rapidly assemble extremely small parts on PCBs, much faster and more accurately than humans (in fact, so accurately that the whole concept of “rework” had to be reworked because when there was an error, humans couldn’t fix it economically.)

    • Tatil_S

      Patents only run 20 years. Their first patent is from 1991, so it may be too late for them to collect unless they added more patentable refinements to the process later.

  • http://twitter.com/mccarthytony Tony McCarthy

    I’m “following along at home” and have a slight disagreement with the quarterly spending on PPE. From the cash flow statements, I have in billions Q1 $0.376, Q2 $1.462, Q3 $2.996, Q4 $3.461, Total $8.295. So the discrepancies are only in Q1 and Q3. I assume its highly likely that I erred somehow….(?)

    • http://twitter.com/mccarthytony Tony McCarthy

      Found my error! And still impressed by both Apple and Asymco wrt this analysis!

  • rj

    It would be interesting to know the relative efficiency of server infrastructure spending by Apple, Amazon, and Google. Google and Amazon are highly skilled in this area – I think they’re generally considered to be the two leading companies in the world. Apple has presumably acquired talent and is building its expertise, but is rumoured to be using quite a bit of off-the-shelf hardware (and perhaps even Microsoft’s cloud computing services) in its initial iCloud implementation. At this point, it doesn’t appear to be in the league of the other two. But does it matter? Does Google’s expertise give it a 10% spending advantage over Apple, or more like a 2x advantage?

    • Tomas

      I think Apple is using Amazon’s cloud services for some of its iCloud features while it’s getting the NC and Oregon data centers up and running. I think they also uses Akamai, for iTunes delivery, in which they used to own a lot of stock.

  • Ernest

    With all the new entrants, phone/tablet manufacturing capacity is squeezed world-wide. Given that Apple pays less and expects more, some parts suppliers may pick alternative contracts over Apple. That could explain the sudden Apple supply delays and difficulties. It could also explain the massive growth in infrastructure investment. When Apple invests in equipment at a foreign manufacturer, it turns things around: it guarantees capacity for itself and leaves only the excess, if any, for competitors.

    • Walt French

      I tried hard. But after setting aside the irrelevant pejoratives and the non-sequiturs, I couldn’t find what your meaning was. Looks like you’re saying that Apple pays less but actually pays more to get capacity, and this is obviously bad.

    • http://twitter.com/JessiDarko Jessica Darko

      You’re relying on a fallacious assumption. Apple expects more, and pays a lot more. For instance, Apple underwrites major equipment purchases for the manufacturers, and is willing to pay billions upfront giving them needed capital.

      Apple is a good customer to have. Sorry that reality doesn’t jive with the “androids winning” fantasy.

  • Pingback: CHART OF THE DAY: Apple’s Exploding Capital Expenditures (AAPL) | Digital Wealth

  • oases

    I like the Jeffrey Bernard quote: ‘You can see a train when it’s coming cant you?’

  • http://twitter.com/jefsmk Jeff Smeker

    Could the majority of this recent spending be the cost of Apple replacing Samsung as a supplier? Sure smells like Apple is phasing them out, but the lost production must go somewhere else, no?

    • http://twitter.com/JessiDarko Jessica Darko

      Well, Apple did just introduce a major new product— the iPad mini is really new, requiring new machines to manufacture. (Though I’m sure Apple is using previous generation tools for parts of it, some of it is inevitably high capital cost.)

  • yetanothersteve

    Has CapEx continued to be a predictor of future sales? That tells me it’s primarily manufacturing. It makes sense that even though Apple outsources its manufacturing it would purchase the equipment. Apple would have a much lower cost of capital, and much greater visibility as to its future plans. (Plus Apple HAS the capital so it can insure capacity.)

    To a large extent one would expect these costs to have a fairly linear relationship to units or at least revenues. Presumably Apple already works at volumes where utilization rates are near maximum achievable.

  • http://twitter.com/JessiDarko Jessica Darko

    The Annual Report is out. First let me give Apple kudos for not spending money on the fancy glossy annual report full of stock photos and “vision statements”. Just the straight 10k like Warren Buffett intended.

    2013 looks to be $9.15B. I’ll leave interpretation up to others. Not a shocking growth, it seems to me.

    I suspect they are NOT building a semiconductor fab. I’d expect a $14B number if that was what they were doing (because they would still be spending $8B or so on tooling etc, for products, unless they started spending on the fab in 2012’s $8B)

    Here’s the relevant section quoted:

    “The Company’s capital expenditures were $10.3 billion during 2012, consisting of $865 million for retail store facilities and $9.5 billion for other capital expenditures, including product tooling and manufacturing process equipment, and other corporate facilities and infrastructure. The Company’s actual cash payments for capital expenditures during 2012 were $8.3 billion.

    The Company anticipates utilizing approximately $10 billion for capital expenditures during 2013, including approximately $850 million for retail store facilities and approximately $9.15 billion for other capital expenditures, including product tooling and manufacturing process equipment, and corporate facilities and infrastructure, including information systems hardware, software and enhancements.”

    • MOD

      Yes, this matches the financial statements. Note 3, page 59:

      Beginning Gross PP&E (end of 2011) balance at cost $11,738, (end of 2012) balance at cost $21,887. Increase of $10,119 during 2012.

      Accumulated depreciation beginning balance $(3,991), ending balance $(6,435). Increase of $(2,444). This is confirmed by the disclosed depreciation of $2.6 bil on page 52.

      If I am interpreting this correctly, they spent 25% over the $8 bil budgeted in 2011.

  • Pingback: CHART OF THE DAY: Apple’s Exploding Capital Expenditures (AAPL) | BizNax

  • http://twitter.com/ddl_red84 Doug Lopata

    The 10-K is out, and the AAPL predicted YoY CapEx spend looks flat relative to 2012. Still a big number at $8-9B spend… Interested to hear how Horace and the thread interpret this?

  • http://twitter.com/fivetonsflax fivetonsflax

    Horace, I wish you would dig in more to the data center stuff. Can Apple deliver services with as much success as they deliver hardware? Why or why not?

  • Pingback: The real threat that Samsung poses to Apple | asymco

  • Pingback: 苹果第四季度支出暴增,资金流向何方? - 阅趣-阅读的乐趣

  • Pingback: The new age of Capital Intensity | asymco