Up to eleven

During my talk at the first Apple Investor Conference in January I said that I would pay very close attention to the Machinery, equipment and internal use software line item from its PP&E during the next quarterly report (it appears on Page 13 in the latest report).

The reason I consider this important is because capital spending has provided reliable foreshadowing of iOS device production. This is itself because Apple invests in the equipment used in the manufacturing processes for its devices. The more spending on equipment, the more production capacity is brought to bear and the more units are produced. Since iOS devices tend to be supply constrained, the more units are produced then the more are sold.

The first quarter of the fiscal year saw a very small increase in CapEx which caused me to “backload” spending later in the year. By watching this spending, we can assess approximately when and for which products is investment allocated. If the spending happens early then we can anticipate an early update to the iPhone. If it happens late then we can anticipate a late release. So has the ramp-up begun as of the second quarter (first calendar quarter)?

The following chart begins to give us the answer. It tells the story of spending by tracking the change in asset values:

Stripping out the Machinery and Equipment portion (in yellow above) and overlaying the following quarter’s iOS shipments yield the following composite graph:

It would appear that there is significant acceleration into the second quarter. Is it something significant?

The increase in M&E is the highest for any quarter to date. Some of that is undoubtedly in service of the new iPad but some must also be the initial ramp-up for the new iPhone.

To gauge it further however, we need to look at overall expectations. The company also signals future production plans though its yearly report. As I concluded from October’s 10K, fiscal 2012 production should double from the year before (as it did from the year before that and the year before that, etc.)

At $1.3 billion, the spending is substantial and a new record, but it does not make a big dent in the overall budget for the fiscal year. Half-way through the year, the total spending on M&E and Real Estate is still only 26% of the budgeted $7.1 billion. You can see the gap in the following chart. The yellow line needs to match or exceed the blue line after the next two quarters.

In order to reach its own expectations, the spending therefore needs to double from the current rate to at least $2.5 billion/quarter over the next two quarters.

Such numbers are hard to grasp. They are unprecedented not only for Apple but for almost any comparable company. Just as a reference point, they will have to exceed the historic average CapEx rate for Google by more than 100%.

Regular readers will note that there is nothing new in this analysis. This is only an update in a continuing story. The pieces are in place for continuing the existing rate of growth.

Footnote: for those who are unfamiliar with the idiom in the title: Up to eleven – Wikipedia, the free encyclopedia

  • Is there a correlation between how much Apple spends on M&E and how many iOS devices they ship? If so, that big spike, how does that translate into volume figures?

    • RobDK

      That is exactly what Horace shows in the second figure; iOS shipments overlaid M&E with a time shift of one (or two?!) quarters.

      Looks like Apple are gearing up for the september/october launch of the iPhone 5, with a massive global launch. And maybe an iPad Mini. And maybean iPad HD, aka the iTV.

      Whilst we are seeing huge expansion for Apple here, reading Horace’s interview with Chris Brennan ( indicates that this continued expansion is necessary for Apple to continuing to drive and dominate the mobile computing paradigm.

      I have always been ‘long’ on Tim Cook’s leadership qualities. I think he is on the front of the wave in this respect. Apple is expecting to continue the exponential growth in supply chain investments, with the implication of at least a doubling of production and sales each year. Apple’s approach to product line and segmenting mean that they cover nearly all price points at a minimum of investment. Large production runs, minimal component variation and the purchasing of future production capacity mean that Apple is able to produce and sell vast quantities of iProducts at relatively low price and with feature sets that the competition cannot match, in relation to the rest of the market.

  • timnash

    Data Centers add to the CapEx for Apple. North Carolina is adding a solar array and fuel cells. Oregon will break ground soon and if Siri is to be really successful outside the US, Apple will need to roll out data centers worldwide.

    • Indeed, but we are talking about $6 billion in the 6 months starting April.

      • Much has been written on new “in-cell” display technology which allows for markedly thinner touch screens. If a redesigned iPhone incorporating this advance is in the offering, then ramp-up expenses related to this, and retooling at Foxconn to accommodate, would have to be a decent wedge in this pie — and within the right timeframe for a C4Q launch.

      • Much has been written on new “in-cell” display technology which allows for markedly thinner touch screens. If a redesigned iPhone incorporating this advance is in the offering, then ramp-up expenses related to this, and retooling at Foxconn to accommodate, would have to be a decent wedge in this pie — and within the right timeframe for a C4Q launch.

      • Jim Zellmer

        I propose another idiom, one that I have recently become familiar with as a parent during our present graduation season.  

        “Money Cannon”. 

        The graduation use refers to event festivities. Horace has illuminated the Apple use case. 

      • vincent_rice

        I like this analogy very much Jim!

        It’s very similar to the state when a project you are working on gets close to fruition. You suddenly find that you need to spend shockingly quickly and freely to get the thing out of the door.

      • timnash

        If Siri is to make it out of beta and be truly available worldwide (that is with a good response time and for a wider range of languages and dialects) Apple will arguably need Data Centers in Greater China, Japan, Australia and 2 in Europe. Even if these (and Oregon) cost $500m compared to North Carolina’s reported $1bn, this will add considerably to CapEx. What we don’t know yet is how fast Apple wants to roll them out.

  • mjuarez

    Horace, great analysis as always. Just points out that, even though 2011 was a record breaking year, 2012 could be roughly 2x that. Unprecedented indeed.

    Just a quick comment: On the overlaid graph, that last bar is not comparable to the three ones before it, but to the one in Q1 2011 (4 quarters back). Maybe it’s obvious to everybody else, but it wasn’t immediately obvious to me. Maybe point it out with an arrow or something?

    Another thing might be, add a second graph, where you extrapolate visually the next two quarters, based on same percentage growth as last year’s. The numbers are clearly helpful, but seeing what the graph will look like if AAPL just matches last year, would definitely have more impact.

    Also, can you please consider using vertical and/or diagonal X-axis labels? The way they are right now, it’s incredibly confusing to look at, especially when the bars are so close together.

  • aharon


    If Iphone 5 needed to double M&E as budgeted it’s hard to
    believe so little money was used so close to launch date (even if it’s October).

    Could we infer that beside Iphone 5 there must be a pretty big surprise, likely a new product, in order to justify all the capacity that was budgeted and not yet implemented?

    On the other side, Isn’t unlikely that Apple made public a budget figures that included capacity for new/secret product?

  • André Jansen Medeiros Villar

    And the new campus ? Apple have already accounted some expenses in other quarters, but the building activity are yet to start, and presumably to be accounted.

    I’d suppose the complex buildings + research centers + power plant + landscaping will cost upwards of a billion dollars and that construction expense probably haven’t been accounted yet.

    They must be waiting on cupertino’s final approval to spend some big money.

    • jdsweet

      The giant numbers Horace presented above are *just* for the M&E component of CapEx, not total CapEx. The strong correlation (and thus predictive power) comes from looking at just the M&E.

    • I don’t think ground will be broken on the new campus before October. They are still getting approvals for their plans. And once initiated, the burn rate of construction will be spread over 2 to 3 years and not add up to one quarter’s M&E. Bear in mind that a new car manufacturing plant costs about $2 billion ( Apple will be spending more than this (a lot more) per quarter. The numbers we’re dealing with here are astronomical.

      • safi kariv

        Which lead to What:
        A deal with China Mobil?
        India? Russia and Brasil?
        Reduction of prices for older models in order to capture mid-quality markets?

      • None of these require capital expenses. Capital expenditures are physical goods that are purchased for production. It does not include components which are part of products sold. It does not include license agreements or new employees or distribution agreements. It does include buildings, land, tooling, office equipment, internal use software and improvements to stores. This is why this measure is significant. Technology companies don’t usually spend on these items. Industrial companies involved in manufacturing are called “capital intensive”. They buy factories or ships or oil drilling platforms. Apple is spending more than any of these “capital intensive” businesses.

      • What you say is true. But I noticed something interesting in the latest campus plans. Along with the main round building and its accompaniments, there is a cluster of on the edge of the site, just East of Tantau Ave. This is called “Research Facilities” while the function of the round building is “office, research and development building”

        Now let’s recall some issues
        (a) Infinite Loop is not supposed to close. It will remain (at least for now) HQ perhaps meaning something like all the execs and some bits of HR, accounting and so on move there when the engineers move out.

        (b) The round building is 2.8M square feet; the “Research Facilities” extra buildings are 300,000 sq feet.

        Given all this, one really has to wonder what is being done in “Research Facilities” that can’t be done in the ample space of the round building. This suggests to me that the “Research Facilities” buildings will perhaps be involved in materials science in some way — they will have special facilities for toxics, perhaps a clean room, those sorts of specialized facilities that wouldn’t make sense (and wouldn’t be allowed) in a more general office building. Which, if it is true, suggest Apple moving a lot more deeply into “deep” manufacturing — eg no longer just buy screens subject to the schedule of Samsung and Sharp, but figure out how to make better screens.

        This seems a heck of a leap for Apple, but I cannot otherwise see the point of these extra buildings. I think it would be extremely interesting to track what permit applications Apple submits (for toxics and suchlike) over the next few years.

  • Walt French

    Always intriguing.

    But there are four very different usages of the CapEx that might be surmised from these numbers, all very different in implication:
    – partnering w/ Foxconn for new iPhone/iPad production. Business as usual, just much more. Assumes no major disruptions on way to 1 billion iPhones/year.
    – first couple of global Siri data centers. (Maiden est. $1 bil). Siri as the new “home page/portal.” Could be hugely disruptive asymmetry.
    – iTV ramp in partnership w/ a new partner. Everybody besides me understands this.
    – expanded ratio of CapEx to sales.

    Regarding the latter: capital is cheap world-wide. (US, EU & JA central banks would LOVE to increase lending; FB shows how little earnings & growth equity investors ask.) I don’t know why Apple would need to be the provider of capital to manufacturing partners unless it’s essentially a takeover of unprofitable firms — Sharp? — who might face higher borrowing costs.

    So your “up to eleven” analogy, with its implications of volume no matter the distortion, seems to mischaracterize what is actually an extremely conservative policy.

    Too conservative, even, I’d argue. My analogy would be that Apple is in the same place the West was as the Cold War was winding down. The “Evil Empire,” by over-emphasizing profit-free industrial development and starving consumers for high-quality products to stimulate demand, was wobbling and the Berlin Wall soon to fall.

    Apple needs to continue to emphasize the consumer delight aspect of their products. Google has ALREADY moved its search algorithms toward more people-friendly tacks to blunt the asymmetric appeal of Siri; I sure hope to see it leave beta in a big way.

    • Walt French

      I guess I don’t really know what the “up” and “down” arrows mean, but I take them to be a way to say the writer’s being obtuse or obstreperous about some theory that’s disproven elsewhere.

      If whoever voted down would let me know what the basis was, I’d appreciate knowing which of my ideas looked more than just “unlikely.”

  • LTMP

    I know it almost seems irrelevant these days, but the entire Mac lineup could be getting a design refresh. The laptops will almost certainly see new form factors, shifting towards the Air design.

    • An interesting thought. Retina MacBooks and potentially iMacs are thought to be on the menu. I just can’t see it being a meaningful portion of $6b given the size of that segment.

  • safi kariv

    Last quarter Apple net at 48%. Does future efficiancy allow apple to expand unit production at the same rate while ExCap at a reduce rate?

  • In thinking further about what elements may account for such an outsized expense level, I thought it useful to make a list:

    1) Data Centers (courtesy of timnash)
    2) Retooling for redesign of high-volume product (i.e. iPhone)
    3) Unannounced new product (courtesy of aharon)
    4) China

    I recognize #4 is short on specifics, but in my mind, it’s the only region with enough critical mass in and of itself to single-handedly sway a balance sheet. It’s a region that Apple is still in its early days as far as achieving scale, and is one that requires the addressing of more particular needs and requirements in order to do business and be successful there.

    Would welcome other ideas to add to the tally.

    • Capital expenses are specific assets which can be depreciated (or not in the case of land). New products per se don’t require capital expenditure and distribution expansion does not either. Data center equipment and tooling do qualify.

      • Yes, which makes expenditures of this size all the more mind-boggling. I cannot fathom data center and retooling amounting into the billions alone without some new element not previously observed.

        I also can’t keep myself from thinking about China given its importance in solidifying growth and achieving the requisite scale you’ve cited in past posts. It’s the one obvious variable that could provide an avenue for return on such a large investment. I’m trying to imagine what kind of expenses Apple might incur, at scale, towards growth in this region.

      • Tim F.

        The next several years of the iOS revolution are the critical period determining if Apple scales (or does not) from low 9-digit annual unit shipments (hundreds of millions) to low 10-digit annual unit shipments (more than a billion).

        There hasn’t been a billion unit computing market in history yet. Never mind one participant in the market achieving that scale on their own. Never mind that one company doing it most profitably or completely vertically-integrated.

      • Just making 3 commas of something, let alone selling it, is mind-numbing. I get tired just thinking about it.


    This is a perfect example of how Hon Hai, HONG FU JIN PRECISION INDUSTRY and (ShenZhen) CO., LTD are all collaborating to utilize liquidmetal alloys in Apple components.

  • oases

    Sales in the quarters after the launch quarter tend to go a little slack. iPhone’s and iPads may not be supply constrained if order delays only last for the first eight weeks after launch or so. Or if the prospective buyers can be shown to be willing to wait rather than go off and buy Samsung. Are you sure that the annual sales numbers aren’t close to the demand (at the given price and distribution level)? I don’t feel like you’ve _proved_ otherwise.

    • Management discusses “supply/demand” balance and usually they state that the products are not in balance in a given quarter. They began to reach balance on the iPhone in the second quarter post-refresh but this is a fairly recent phenomenon and it’s not the case for the iPad. So my statements about supply constraints are based on company statements. The thing to note is that the refresh cycle is such that there is always one “transition” quarter when the market anticipates a refresh and buying slows down (though there is usually still y/y growth). This demand reduction is mostly self-imposed because the company works on a predictable yearly cycle.

      • Walt French

        Guess we’re in a transition quarter (or two) then: while I had my oil changed yesterday, the shop manager saw me checking my phone & asked whether I was going to line up for the new one this summer: you know, the one with LTE, bigger screen and faster CPU.

      • oases

        Still confused. If they can sell x number in launch quarter, how could they be supply constrained in any subsequent, lower selling quarter? Re iPad: delivery times went down to 3-5 days a few weeks ago didn’t they? I can’t say that management is telling fairy stories, but it just doesn’t compute for me.

      • jawbroken

        The launch quarter likely has a couple months of production to build up inventory so it makes perfect sense that they could sell more in a launch quarter than in subsequent quarter and be supply constrained. This also applies to all their component suppliers. Not sure how you’re looking at it to come to your conclusion.

      • Delivery times are not the only indicator of balance. There are hundreds of operators which place orders with Apple and that order book is prioritized differently than consumer direct, which is what you see online.

      • Delivery times are not the only indicator of balance. There are hundreds of operators which place orders with Apple and that order book is prioritized differently than consumer direct, which is what you see online.

      • Delivery times are not the only indicator of balance. There are hundreds of operators which place orders with Apple and that order book is prioritized differently than consumer direct, which is what you see online.

      • Delivery times are not the only indicator of balance. There are hundreds of operators which place orders with Apple and that order book is prioritized differently than consumer direct, which is what you see online.

      • Delivery times are not the only indicator of balance. There are hundreds of operators which place orders with Apple and that order book is prioritized differently than consumer direct, which is what you see online.

      • oases

        Okie dokie. But how many more units they could have sold is a mystery isn’t it? The difference between another 10% and another 50% is huge.

  • Pingback: Tuesday links: the high cost of complexity | Abnormal Returns()

  • jbelkin

    Nice to have ‘captive’ plants, you lock out competitors without real capital outlay plus actually running an overseas plant …

  • Joe_Winfield_IL

    My (highly scientific) analysis of this M&E is that it is being used for a three-step plan:

    1) Dig a moat around 250 million user installed base
    2) Annually dredge the moat to keep it nice and deep
    3) Build freeways into the moat-protected user base and expensive toll bridges at the exits

    • Scott Sterling

      Well put, and exactly as it should be for Apple.

  • Pingback: The how and why of a 4-inch iPhone |()

  • Its worth taking Depreciation schedules into account when looking at Apple’s Capital Expenditures budgeting.

    While Apple’s $6 billion in expenditure is indeed massive, its also worth noting that the sort of “capital good” that Apple is likely to be spending money on is probably going to have a much shorter economic life (and depreciation schedule) than oil-wells and automobile plants – which typically are depreciated over at least a decade or two. Tooling and Jigs to punch out iPad 4 and iPhone 5 cases will become worthless (and therefore “fully depreciated”) as soon as the iPad 5 and iPhone 6 model is announced – usually within twelve months.

    • Tatil_S

      Why? Apple will continue making iPhone 4 and 4S for a year or two after iPhone 5 is introduced, so Apple will need many of these machines even after then. Even if the volumes are not very high for the older versions, the surplus machines can be sold. Do you have any specific knowledge that these machines will be become worthless after only 12 months?

    • Depreciation schedules vary. It depends greatly on the asset class. One thing to keep in mind is that as device tooling is used 24×7 its useful life may be used up in one third the time of tooling used on a single shift. Therefore it’s possible that Apple can depreciate the iOS device tooling in one to two years, comfortably coinciding with its product cycles.

  • Pingback: Apple investiert in Maschinen und Ausrüstungen: iPhone 5 in Sicht? | Apfelpage()

  • Pingback: Apple Said To Be Investing In Equipment (And Machinery) At An Exponential Rate. The Question Is: What For? ‹ RazorianFly()

  • It is possible that the capital spending boost represents business-as-usual for Apple rather than an early warning of a major product introduction or update.

    With nearly 100% annual growth, Apple may simply taking greater responsibility for financing capital outlays and Apple’s vendors (such as Foxconn, etc.) doing less … in exchange for lower prices for components.

    That seems likely if financial markets have tightened significantly, which would make it more difficult for vendors to make adequate capital investments to keep up with Apple’s growth. Apple has ample cash overseas to conduct an operation such as this — and it would tie vendors more closely to Apple to boot.
    Now the numbers. Hypothetically, if Apple revenues expand by 100% y/y, that represents more than $100b in incremental sales … or (let’s say) $40b to vendors. Assume a single vendor supplies all of $40b worth of components, the vendor’s P/Book ratio = 1.5 and P/Sales ratio = 2. These suggest the company’s capital costs (book value) will be about $52 billion.
    This may not be so amazing if it happened one time, but if Apple’s sales grow by $100b EVERY YEAR, vendors need to invest about $50b EVERY YEAR to keep up with the demands Apple is placing upon them. It shouldn’t be surprising if these smaller companies — many based in China, with its undeveloped financial markets — have difficulty keeping up with this pace. Nor should be be surprised if Apple uses its excess cash to ensure a steady supply of parts (to avoid production shortfalls) and to gain greater influence over vendors that Samsung, HTC, et al depend upon to compete with Apple.
    This is all hypothetical: I’m not saying that Apple is financing 100% of the capital projects of all of its vendors, or that all of Apple’s vendors can be consolidated into a single unit. The exercise simply illustrates the magnitudes involved in Apple’s growth and suggests that conditions in financial markets may be causing some capital outlays to appear on Apple’s balance sheet rather than the balance sheets of its vendors.

    • The trouble I have is with the use of “financing”. Financing implies Apple lending money as capital to be used in increasing capacity. In the statements I read, Capital Expenditures imply Apple *owns* the equipment. It is not acting as a bank but actually buys and has control over the equipment itself. The supplier only owns or leases the premises and supplies the labor. Financing component purchases is a different matter and would be accounted under accounts receivable.

      • Well, I’m not claiming certainty with any of this, but only speculating.

        I didn’t mean to suggest that Apple was lending money to other companies — though perhaps it does some of that. I meant that perhaps Apple is purchasing capital equipment or other assets and placing it at the disposal of other companies. I’m not certain whether this might include assets such as patents.

        Apple absolutely MUST ensure that component suppliers invest (say) $50 billion each year, or Apple’s growth will grind to a halt. If component suppliers can’t gain access to that much financial capital, what is Apple to do? Apple could lend money to component suppliers or provide financing by purchasing their stock, but I’m suggesting that it Apple might also be purchasing assets the suppliers need and put the assets at their disposal.

        It is not proof, but there appears to be evidence of this in the cozy relationships Apple has with some of its suppliers. Is there any doubt whether Apple has ordered Foxconn to raise wages at its iProduct-producing factories? Or that Apple has arranged for Foxconn to hook up with Sharp? Isn’t it wonderful how key suppliers seem to put Apple ahead of their other customers, all the while Apple’s production costs are at or below those of competitors? I think Tim Cook is great, but he’s not able to make everything come together by magic …

  • One implication of Apple always being supply constrained is that this means that Android has actually helped Apple by keeping the iPhone from becoming a train wreck of missed expectations. Imagine if there was no Android competition, and then Apple tried to make another 30 million phones a year? Could the quality be kept up? Would Apple have spent enough on Cap Ex to cover the larger demand in phones?
    Even if they could at launch could they have kept up with the surge in smart phone demand over the last 3 years?

    • Android is doing a number on feature phones, much as BlackBerry did before the iPhone showed up. As users moved to these platform devices (after abandoning non-platform devices) the users became educated to what could be done with a device. Some will demand more and switching costs will determine whether they stay with a platform or move on.

  • Pingback: Report: Apple breaks records with production investments, but for what?()

  • fiftysixty

    Horace, this is very interesting, and indeed the scale of things is staggering. Can you foresee a situation where the bottleneck for Apple is their vendors’ ability to increase production simply because the scale becomes so big? Am I correct to assume that your projections for 1 billion base call for quadrupling the current production capacity, or will even more be needed? And does that directly translate to CapEx, so that two-three years from now we’ll see Apple spending the equivalent of one automobile plant each month? Where are the limits in manufacturing new manufacturing capacity, availability of qualified work force, raw materials, shipping capacity? Should we look at Foxconn Brazil as a hedge against these kinds of limits? A lot of questions, but this whole issue is the elephant in the room that very few seem to want to talk about.

  • GuruFlower

    What I don’t understand is how Apple could spend that much money on “buildings, land, tooling, office equipment, internal use software and improvements to stores” and keep it secret? 7.1 B represents 81% of their current Property, Plant and Equipment line. It’s 51% if IBM’s PP&E, for goodness sake.

    And how can you spend that much in 6 months? Whatever these plants are for they won’t be ready for the iPhone 5 release, that’s for sure. This speaks to something down the road.

    • I’m not sure what you mean by keeping it secret.

      I suggest reading the posts from October last year that put the story together. The relationship between spending and iOS device output is pretty clear. The growth in spending predates production by one or two quarters (two quarters in the early years of iOS and one quarter more recently.)
      I put the analysis together last October. You can read it here:
      There’s also the video of my talk on the subject:

      • GuruFlower

        Thank you for sharing that. The links completely clarified the question for me.
        I’m new to your site and busy trying to catch up on your thinking.
        I was actually searching yesterday to see if you had written anything about what a discounted dividend analysis might say to APPL’s current and future valuation now that we’re going to be getting one. Perhaps we’ll see that topic at the appropriate time.

  • Pingback: Re-imagination: the $36 Trillion addressable market | Rainmakers()

  • Pingback: duch Apple investiert in Maschinen und Ausrüstungen: iPhone 5 in Sicht? | affklik()