Categories

Using Capital Expenditures to predict Apple's share price

It takes money to make money. That’s a cliché. But it’s also true. The interesting question is how much can be gained from how little.

In previous articles I explained how Apple’s expenditures of capital for equipment used in manufacturing affects their output of products. The relationship between capital in and product out should stand to reason.

The more surprising aspect of that analysis is that we get to know in advance how much Apple spends (since they tell us their budget a year before it’s spent.) and therefore it becomes possible to get a rough idea of how much they will produce. And since demand has generally been higher than supply we can get an estimate of how much Apple will sell.

The only missing piece to this logic chain is to estimate how much will shareholders benefit from the capital expenditure. I’ll try to establish the relationship through a build-out of graphs.

The first graph shows Apple’s share price at weekly resolution.

The time frame stretches back six fiscal years. The time span includes some dramatic periods including the financial crisis and the launch of the iPhone and iPad.

To illustrate the effect of the iPhone and iPad on this share price appreciation, I’ve overlaid a quarterly resolution graph showing revenues over the same time period with each product line shown separately.

 

Note that I’ve indexed the vertical scale to match approximately the highest peaks of both graphs. The two axes scales are shown separately on the left.

The relationship between growing sales and growing share price is plain to see. Note that the iPhone and iPad account for the vast bulk of  sales growth and hence they are mostly responsible for the share price appreciation. Again, this should not be surprising. The scale gives an approximation of the relationship where $50 billion in quarterly revenue roughly yields a share price in the $700 to $800 range.

As I’ve shown in the last post on forecasting production using CapEx (focus on the last chart) iOS device shipments (mostly iPhone and iPad) correlate with capital expenditures in the previous quarter. Using the historic relationship, the next two quarters could see a significant increase in iOS unit production. The sum of the low end of the range implies about 163 million units.

If the relationship is therefore CapEx -> Production and Production -> Share price appreciation, then the transitive implication is that CapEx -> Share price appreciation.

That relationship would be illustrated as follows:

The CapEx data shows the offset to the right due to the relationship with production being time shifted by one quarter. Recall that expenditures are booked ahead of output and therefore we have this apparent view into the future.

The fourth fiscal quarter’s spending has just ended and the output will be produced during October-December period.

The obvious question is whether production will increase in response to spending (as it has in the past), or, put another way, will the stacked bar areas follow the beige bars and, furthermore, will share prices follow production (as they have in the past) and thus the red line will follow the beige bar.

It takes money to make money. The relationship seems to be that $1.5 billion of CapEx per quarter yields a share price of $800. With spending reaching $3 billion per quarter will the share price reflect a similar ratio?

  • http://newcanuckworkshop.tumblr.com Junesoo Andrew Yang

    Does this help to explain in part the rapid decline of companies as well? Massive capital spend ahead of expected sales growth, which results in equally massive write downs in the case sales projections are not met. RIM and Nokia are both in situations where the cost structures are built around an expected revenue base and sales volume, but once the growth slows or declines, the organizations are burdened with disproportionate costs to sales ratios.

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

      I suspect you may be right for Nokia, which has its own factories. I think most of the other players just outsource everything having to do with production, including most of the tooling costs, which pushes a lot of that risk off onto the manufacturing partner. Apple is unusual in that it neither owns the factories, nor outsources all production costs, but operates in some kind of hybrid mode where it buys some of the machines used in the outsourced production facilities.

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

        Capital equipment is depreciated over a fixed life. The equipment Apple purchases and uses in production is probably used on three shifts per day and therefore depreciates in less than 2 years. I don’t think it’s a coincidence that a product form factor from Apple lasts about as long. Once depreciated, the equipment is “off the books” and can be disposed of. It does not create a burden on the owner in the sense of a barrier to exit the business.

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

        As a point of interest, Horace, do you think Apple retains title to the equipment after it is depreciated, or do they pass it to the manufacturer? If the former, they could keep it from being used for competing products (and copying Apple’s special production steps), but I haven’t seen anything about Apple doing “scorched earth” things with their production lines, which I’d have expected to show up in some of the Foxconn investigations.

      • http://www.facebook.com/Waterbug Steven Kan

        Depreciation is an accounting procedure and does nothing to the ownership or usability of an asset. It also has no effect on cash flow _except_ for its effect on taxes, and tax accounting is not the same as FASB accounting. Apple’s machinery will still be owned and used by Apple as long as it’s mechanically useful. Don’t confuse an asset’s depreciation life with its useful life. They’re not even remotely the same concept.

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

        I agree that they’re different things, but Horace’s depreciation estimate is also close to the life cycle of a particular iPhone design, so I was wondering if they coincided deliberately in some fashion. Obviously some of the equipment has a much longer useful lifetime (the CNC milling machines come to mind), but other equipment may go out of use after a product cycle or two, at least for Apple’s use. I was wondering what happened to it.

    • GeorgeS

      Perhaps, but RIM’s writedown was on inventory (and goodwill). CapEx is a sunk cost: you can’t write it down, if I understand correctly. RIM may have had to guarantee a certain number of orders to get their suppliers to do business, as the suppliers would have to tool up for the production. With Apple owning the tooling, they can get the suppliers to give a better deal and be more flexible, as the suppliers’ costs are mostly labor, which is more flexible.

      There’s another aspect. Apple can essentially lock up a huge fraction of available manufacturing capability, putting their competitors at a disadvantage. I’ve read that Apple did just that by buying most of the Japanese output of precision CNC milling machines when they went to the “unibody” construction. They also bought essentially the entire output of a particular green laser that’s used to burn tiny holes (e.g., for the “sleep” light on the more recent MacBook line.) Apple is supposedly the largest buyer of flash memory in the world, as well.

      • Gregg Thurman

        If you go back to the original iPod, Apple bought the entire production capacity of the only manufacturer (Toshiba) of the small form factor HD used in the iPod. When competitors tried to build their own version of HD digital music players, they couldn’t, because Apple had purchased all production capacity for the next 2 years.

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

        Good on Apple.

    • Mayson

      I think they lose more on unsaleable inventory, and over-employment, than on capex.

  • http://twitter.com/kremsr Ryan Krems

    One thing to be aware of could be the fact that the machinery maybe more expensive due to tighter tolerances, enhanced techniques required for the latest designs. Therefore increased spend doesn’t always have to equal increased output (although you can see from Apple that this has usually been the case)

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

      I think there are too many variables in the tooling cost effectiveness to say any one factor has an effect like that. For example, tighter tolerances in the machinery could increase yields. Or more expensive tooling could make more parts per unit time, lowering the cost per unit of the tooling. I expect these are the kind of things Tim Cook optimizes….

  • http://twitter.com/mark_a_howard Mark Howard

    The magnitude of these purchases cannot be simply component prepayments. So they are manufacturing equipment, probably custom, purchased by Apple but installed in manufacturing partner facilities. To what end? Manufacturers would not balk at designing / installing such equipment at their own cost if it meant ongoing revenue from the largest CE company in the world.

    So perhaps it is that Apple desires exclusive use of the equipment, or a particular process, for a predetermined time. It may also grant Apple flexibility and power, in that if the manufacturing partner were not up to the task, Apple can take their stuff and install it elsewhere. They would not be beholden to a particular manufacturing company in order to get their product out the door with their unique process, which may have only a six month shelf life before competitors could catch up. Also, they pay less per component to the manufacturer as they don’t have to include the cost of equipment depreciation in their profit level.

    Why not buy the whole manufacturing company? Apple has more flexibility this way, and has control over what is important to them. It is expensive this way, clearly, but they deem it so important to be worth it. It is an interesting strategy.

    • http://twitter.com/ChristianPeel Christian Peel

      horace has addressed this in previous posts, which you can find by scouring the archives.

      If I recall correctly, one guess is that purchasing an assembler (Foxconn) or a chip maker would decrease Apple’s flexibility in extracting every ounce of flesh out of their suppliers and supply chain. I.e. if you own a company that assembles iPhones, then you would have to use your own company and it’s more difficult to have multiple companies vying to do the assembly for you.

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

      None of these expenditures are for components.

      • SmittyRedcard

        Does the esitmate of spending include new iOS devices at Retail stores? Apple has moved from a few Macs at the Genius Bar to all support staff using iPads during working shifts. The aging non-camera iPods Touch that retail staff use (third gen) are probably going out in the next few quarters as their purchase price is amortized and the software moves past what those old devices can handle. Do you think the CapEx budget covers their replacement, either wholesale or gradually?

      • claimchowder

        Apple has ~40000 retail staff. Multiply (for a very high estimate) with $1000 and you get 40mio for supplying new devices to all staff. That’s nothing compared to the 3000mio.

      • claimchowder

        Apple has ~40000 retail staff. Multiply (for a very high estimate) with $1000 and you get 40mio for supplying new devices to all staff. That’s nothing compared to the 3000mio.

      • claimchowder

        Apple has ~40000 retail staff. Multiply (for a very high estimate) with $1000 and you get 40mio for supplying new devices to all staff. That’s nothing compared to the 3000mio.

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

        It should.

      • christophe driver

        I do understand that. But what if the share of theses investments related to components instead of final assembly (tools at sharp facility for incell screen instead of CNC machine at Foxconn for example) was simply going up with the latest iDevices? then the correlation between Capex and numbers of iDevices would no longer be the same, and therefore the correlation between Capex and share price wouldn’t either..

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

        We don’t know what the mix of assembly and component manufacturing ever was. We only have the historic relationship. I don’t see any reason to think that components have become proportionately more expensive to build or whether that mix has changed. My guess is that component equipment is the lion’s share of the equipment budget.

  • Frank

    Reminds me of Al Gore’s “Inconvenient Truth” graph going off the charts. How exciting! What does Apple have in store for us? Spending almost three times as much as their record breaking 4th quarter last year… wow.

    • suddy

      Is this all not the iPhone 5? They have a steeper launch of iP5 to 100 countries by Dec. They are anticipating increased demand for iP5 compared to previous iPhones. That’s how I interpret this data – that is this CapEx is for iP5 and is not meant for a new product launch out in the future. Thoughts?

      • Penick

        What about iPad mini and growing iPad regular sales? Takes equipment. More equipment, more product, more sales, more share price.

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

        The CapEx expenses should really be for the entire expected production run of iPhone 5s, or at least a major part of setting up the equipment for its lines. If the demand went unexpectedly higher, they might have to add more later. I suspect they put most of the production in place before launch, and maybe add a little more after launch once they confirm demand and tune the production rate.

        Horace, have you seen any evidence of Apple spending CapEx money on bottlenecked lines after the fact? I can’t think of any times when they might have done this, except maybe early in the release of the first iPhone and the first iPad, where they probably had no idea of what sales were going to be like early on.

        That said, if they are launching the iPad Mini this month, there should have been significant CapEx associated with that recently, as well as for iPhone 5.

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

        Tooling is not something you source with short lead times. Production equipment has to be ordered well in advance and it is not something that is brought on-line quickly. The sales and installation cycle time is probably 6 to 12 months.

    • Camden1

      As much as I would love to see Apple’s price continue to follow the growth of capex – it will not. One must realize that much of this recent capex is for simply replacing the production that has been taken away from Samsung. It is NOT all going towards additional manufacturing capacity. The answer to the question Horace posed in the last paragraph is clearly “NO”.

      • Kizedek

        Not necessarily. Before one can make your assumption, one must first find out what Samsung has been producing for Apple that only Samsung can produce for Apple; and what investment Apple made into Samsung’s production lines.

        1) If the assumption is that these CapEx expenditures for equipment means that Apple owns the equipment, then anything (if anything) they installed at a Samsung facility belongs to Apple.

        2) Components like screens are already being produced by multiple suppliers, so there already is an alternative supplier for many items. (eg. LG or Sharp for screens.

        3) Even for something as complicated as the A6, there may already be alternatives, such as Texas Instruments. Did Apple invest in the smaller chip fab processes for Samsung? Unlikely. And this is where competitors are going anyway. Apple may lose some production capacity initially, as they switch, but unless they were buying into the chip fab process at Samsung with their own special machinery, then additional investment elsewhere may not be necessary when switching to a competing supplier.

        With CPU/SoC the switch is a win for Apple, because it means that Samsung-the-components-supplier is not slipping Samsung-the-phone-manufacturer trade secrets that belong to Apple.

        Basically, critics can’t have it both ways. One can’t both say that Apple doesn’t have any IP or engineering know-how because the suppliers make everything AND Apple is sunk if they have to switch suppliers. If Apple is doing something pretty unique here — buying itself up further into the supplier chains — and if the CapEx expenditures are as significant as Horace claims, then Apple is in good shape to make this kind of move. In fact, this is exactly why Apple is doing things this way — to remain flexible.

        Anyway, this quarter’s CapEx shows so many billions more, that even if Apple DID have to cover production facility and capability that is lost as it moves away from Samsung as you suggest, then surely there is a few billions left over to still significantly increase the production capacity of the iPhone 5 over the 4S. Where was the initial CapEx that set up Samsung as such a special supplier in the first place? If any, it must have been much less, because we haven’t seen this level of CapEx spending before. Are you suggesting that all the extra billions this quarter are in the budget merely because Apple foresaw the move from Samsung and knew it had to build a COMPLETE chip processing fab from scratch somewhere, for someone, all in one quarter?

      • Kizedek

        Not necessarily. Before one can make your assumption, one must first find out what Samsung has been producing for Apple that only Samsung can produce for Apple; and what investment Apple made into Samsung’s production lines.

        1) If the assumption is that these CapEx expenditures for equipment means that Apple owns the equipment, then anything (if anything) they installed at a Samsung facility belongs to Apple.

        2) Components like screens are already being produced by multiple suppliers, so there already is an alternative supplier for many items. (eg. LG or Sharp for screens).

        3) Even for something as complicated as the A6, there may already be alternatives, such as Texas Instruments. Did Apple invest in the smaller chip fab processes for Samsung? Unlikely. And this is where competitors are going anyway. Apple may lose some production capacity initially, as they switch, but unless they were buying into the chip fab process at Samsung with their own special machinery, then additional investment elsewhere may not be necessary when switching to a competing supplier.

        With CPU/SoC the switch is a win for Apple, because it means that Samsung-the-components-supplier is not slipping Samsung-the-phone-manufacturer trade secrets that belong to Apple.

        Basically, critics can’t have it both ways. One can’t both say that Apple doesn’t have any IP or engineering know-how because the suppliers make everything with Apple merely sticking a label on it, AND that Apple is sunk if they have to switch suppliers. If Apple is doing something pretty unique here — buying itself up further into the supplier chains — and if the CapEx expenditures are as significant as Horace claims, then Apple is in good shape to make this kind of move. In fact, this is exactly why Apple is doing things this way — to remain flexible.

        Anyway, this quarter’s CapEx shows so many billions more, that even if Apple DID have to cover production facility and capability that is lost as it moves away from Samsung as you suggest, then surely there is a few billions left over to still significantly increase the production capacity of the iPhone 5 over the 4S across all component suppliers (or ramp up iPad Nano or whatever). Where was the initial CapEx that set up Samsung as such a special supplier in the first place? If any, it must have been much less because we haven’t seen this level of CapEx spending before. Are you suggesting that all the extra billions this quarter are in the budget merely because Apple foresaw the move from Samsung and knew it had to build a COMPLETE chip processing fab from scratch somewhere, for someone, all in one quarter? We are talking about a lot of money this quarter, probably enough to build a complete factory.

      • Camden1

        Not ALL the billions but some of them.

  • Sebi

    Please tweet the CapEx of FQ4 as soon as you have the data! Thank you!

  • Nigel

    Provocative post. I can’t wait to see what happens over the next few months.

    If this ends up being an accurate predictor of share price, the next question is how long demand will continue to outstrip supply. Which is likely impossible to predict.

    • Gregg Thurman

      No, I don’t think it is. Apple manages its inventory very tightly. If Apple invests in machinery, etc., on the scale indicated by Horace’s graphs (I think they are), then Apple has already determined that future demand exceeds their future capacity. Maximum ROI from such investments occurs when the machinery purchased is used at 100% capacity. Having more capacity than is needed dilutes the ROI. Leaving a little (unit sales) on the table is not always a bad thing, especially when the competition can’t grab unfilled demand for your products with theirs.

  • graphex

    Securely fasten your seatbelts!

    I’m trying to imagine the impact resulting from money rapidly moving in the markets as share price moves from mid 600′s to 800 to 3000. ;-)

  • Tatil_S

    Horace, do you really think Apple can hit $185 billion in revenue in Oct-Dec quarter? It looks like the correlation between sales and CapEx will break for three quarters in a row. FQ3 sales were already lower than CapEx suggested, FQ4 will probably be lower, as even your forecast expects less than 30 million iPhones and FQ1 2013 hitting $185 billion seems impossible.

    Is this a harbinger of a completely new product or a different business model in how Apple will handle manufacturing components or just things getting out of hand?

  • Tatil_S

    Horace, do you really think Apple can hit $185 billion in revenue in Oct-Dec quarter? It looks like the correlation between sales and CapEx will break for three quarters in a row. FQ3 sales were already lower than CapEx suggested, FQ4 will probably be lower, as even your forecast expects less than 30 million iPhones and FQ1 2013 hitting $185 billion seems impossible.

    Is this a harbinger of a completely new product or a different business model in how Apple will handle manufacturing components or just things getting out of hand?

  • Tatil_S

    Horace, do you really think Apple can hit $185 billion in revenue in Oct-Dec quarter? It looks like the correlation between sales and CapEx will break for three quarters in a row. FQ3 sales were already lower than CapEx suggested, FQ4 will probably be lower, as even your forecast expects less than 30 million iPhones and FQ1 2013 hitting $185 billion seems impossible.

    Is this a harbinger of a completely new product or a different business model in how Apple will handle manufacturing components or just things getting out of hand?

    • oases

      It’s a harbinger of the oncoming plague of iPad Minis.

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

      I wonder if there was additional CapEx associated with tooling up the new integrated display panels, which might have shown up earlier than simple mechanical tooling for producing cases and the like, since the new display is a risky (for production) new technology.

      • Tatil_S

        That excess ~$5 billion can buy more than one LCD panel fab and I doubt you can build one in just two quarters. CapEx numbers sound astounding.

  • oases

    I can’t thank you enough for this. I wouldn’t get this anywhere else. Your blog’s full of exemplariness. Also, I’ve saved a screenshot of the last graph to my iPad – I’ll be opening it up periodically to remind myself.

  • LTMP

    I think that this is as good a model as any for trying to forecast Apple’s share price (probably better than most), but I wouldn’t expect it to provide great accuracy.

    For one thing, the market is not always rational. For another, Apple is making it’s cap/ex bets a year or more out. That leaves a lot of room for error, in either direction.

  • Oliver Bruce

    Is it possible that Apple is buying this equipment as a defensive move, protecting its IP further into the supply chain? I’ve been thinking about this after your latest Critical Path where you talked about suppliers hollowing out the innovation advantage over time by moving higher and higher up the value chain. Apple operates, as Walter Milliken has pointed out, in a sort of ‘hybrid mode’ – buying key equipment used in manufacturing and deploying it in suppliers operations.

    From a strategic perspective, it would make sense – Apple owns the valuable parts of the operation, while leaving all the ‘other stuff’ (labour, land, buildings etc.) to be carried by the supplier. It makes it far more difficult for a supplier to replicate an advantage if it doesn’t actually own the machinery with which it produces components.

    If this was the case it could also therefore decouple the capex spending from production increases.

    • Gregg Thurman

      Where I see the beauty of Apple’s strategy, is that it doesn’t require the usual early adopter/mass production cycle to get new technology prices down. Apple provides the financing, and the demand, for the new technology today, without the time delay of market adoption (price decay/production ramp). By owning the production equipment, Apple gets the benefit of the new technology at prices, and scale, that ordinarily would not be seen for two or three years, at which time it would be available to anyone, eliminating any differentiation the new technology brings to the table.

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

        Furthermore apple has complete control on the quality this machineries operate.
        They require high standards, pay for it and give the equipments to suppliers to operate, no compromise on quality is needed.

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

        An interesting point, since those same suppliers are also being price-squeezed. Buying the key equipment ensures the suppliers aren’t using shoddy gear in the production line that might affect build quality. But I suspect the main CapEx is aimed at machinery the producer simply wouldn’t buy on its own.

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

        Sometimes having too exacting of a standard can bite you in the behind. Case in point, a lot of scuffing and scraches on the black iPhone 5. This is a result of ever increasing precision / quality standards at much higher levels of production. It becomes a high-stress situation on those that have to produce the devices hence the 3 to 4 week wait

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

      It may be protecting its “design”, not so much its IP. Case in point might be the unibody design of the MacBooks. The milling equipment used for the portable Macs bay be unique to Apple due to their having purchased all tooling available. Similarly it’s likely that the new iPhone’s Aluminum chassis cannot be duplicated by competitors due to a lack of available tooling suppliers. Tooling for screens in the specific technology used and size and lamination that Apple uses may also be exclusive.

  • neutrino23

    The last chart is just mind boggling. If Apple just maintains that and perhaps grows it 20% YOY going forward it will be amazing.

    Perhaps some of this is related to new Foxconn factories in Brazil and the interior of China.

  • http://realboxscore.com/ Juha-Pekka Sipponen

    A fascinating, value-adding post, Horace, with insightful comments from many people. I especially liked, for example, Walter Milliken’s and Gregg Thurman’s perspectives on how Apple (possibly) uses their strong balance sheet to extend the tight vertical control (which is their core operative model) and to offload supplier risk with unproven technologies (resulting in better time-to-scale). If so, there probably is a correlation with capex and sales, but how stable that number is (e.g. in the event of changes in the partner portfolio or technology mix) is probably the core question.

  • RobDK

    Does anyone have any idea what Apple is gearing up for? These numbers are huge. Which of the following?:

    - iPad Nano introduction in October 2012.
    - iPhone 5 introduction on the Chinese market for the Chinese new year.
    - iPad 4 introduction in spring 2013.
    - A combination of all three.
    - Maybe an iPad HD, aka iTV.

    I would also add that with Horace’s estimate of over 160m iDevices per quarter gives about 700m per annum. This puts Apple within reach of the 1 billion devices per annum which Horace previously has mentioned as necessary if Apple is to stay relevant in mobile computing.

    • http://www.facebook.com/profile.php?id=820800998 John Miller

      I believe Horace’s estimate was for 163m iDevices over two quarters, not one. Here is the relevant quote: “Using the historic relationship, the next two quarters could see a significant increase in iOS unit production. The sum of the low end of the range implies about 163 million units.”

  • Klasse

    Are the results affected in any (meaningful) way if you remove the cash portion of Apple’s share price?

  • Micromeme

    Hi Horace,
    I know its not strictly speaking capex, but can you see anything that suggests that apple sees their mapping problem as a large data quality problem rather than a software dev problem? what I mean is that software dev problems are solved by teams of 10s to 100s, where a worldwide map data cleaning and quality improvement problem is a problem of 1000′s of staff (think 43000 zip codes in the US and like numbers of subregions in europe, + more in asia). That staff would be be constantly updating and testing the data for errors and changes. Can we see that staffing or that $ expenditure for outsourcing anywhere in their numbers?

    for that matter we would get a lot of understanding if we could see that they started that kind of team a year ago (they knew all large data sets have to be cleaned) or they just started it (ie they thought they could just buy error free quality data and all they had was an algorithm problem to import it)

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

      There was a report yesterday that Apple Store employees have been tasked to spend some of their hours improving the data on their local areas. That won’t help much outside major metro areas, though.

      • Micromeme

        yes I was afraid when I saw that report that apple wasn’t really taking the data quality problem seriously since that seems so haphazard.

      • peto1

        Not at all. Asking their retail staff to participate in the data scrubbing enhances morale because they are helping to solve a problem …

      • Micromeme

        well its not sufficient to the task– though it might help at the margins. that might help with morale, but frankly if its not real clear real fast that they fit into a larger system that is sufficient then it might just as well hurt morale. same problem with talking up crowdsourcing–it has to fit into or complement a system that is sufficient to the task.

      • peto1

        Nobody said Apple is fixing their Maps problem by using some of their retail staff only. It’s just an incremental addition to the significant resources being employed …

      • Micromeme

        that is the point of my original question to Horace. Really it was: essentially how big is apples expenditure (read staff) for map quality improvement? And is it at appropriate scale to catch up and surpass google? “nobody said” is not exactly the point– what I have not heard is the affirmative story that says “apple has X ’000 map correctors in place for a world wide data cleaning and updating push”. All I have heard is they are hiring some developers in the map space and everyone is working real hard.

      • Micromeme

        that is the point of my original question to Horace. Really it was: essentially how big is apples expenditure (read staff) for map quality improvement? And is it at appropriate scale to catch up and surpass google? “nobody said” is not exactly the point– what I have not heard is the affirmative story that says “apple has X ’000 map correctors in place for a world wide data cleaning and updating push”. All I have heard is they are hiring some developers in the map space and everyone is working real hard.

      • Micromeme

        well its not sufficient to the task– though it might help at the margins. that might help with morale, but frankly if its not real clear real fast that they fit into a larger system that is sufficient then it might just as well hurt morale. same problem with talking up crowdsourcing–it has to fit into or complement a system that is sufficient to the task.

      • Frank Vaughn

        Everyone that uses maps helps enhance and correct the database. All of the usage data and corrections are fed back into the system.

    • Micromeme

      actually now that I think about it, properly speaking a map database is a capital asset and money spent to improve the asset quality is probably capex.

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

      Maps data is partly crowd sourced. Google’s has been for years. Broadly speaking maps data is commoditized and has been for some time. You can track the effect by looking at the market value of companies that create the data sets. Navteq was bought for $8 billion in 2007 but may not be worth $1 b today.

  • JohnDoey

    The thing that is coming is when the phone makers go out of business. iPod got bigger and bigger and then the big MP3 player makers died and iPod exploded. Everyone who doesn’t have a PC class OS is just a phone.

    • http://search.websonar.com:8080/ Duane Bemister

      Correct and Apple has already disrupted the iPhone with the iPad.

    • Ptmmac

      I think you are under estimating the competition in this market. The cell phone business is heading into the trillions of dollars range. Every large manufacturer of electronic components is trying to get into this business. No one else had a chance against Apple in the market for music players. Samsung alone made over 7 billion $ last quarter. This is the largest battle between corporate giants the world has ever seen. Buckle up, this is going to be complicated. Apple will not win in a walk. They will lose some battles. The question is how will they respond to adversity?

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

        Don’t forget, near the end of this month Microsoft will be entering the game in a big with Windows Phone 8, Windows 8 & RT, plus the fact they will create their own device, Surface.

        Don’t count them out yet. The Windows ecosystem is huge and it isn’t going anywhere anytime soon. Plus the fact that own the de facto productivity suite in Office. The Windows and Office franchises plus the fact that MS is a much bigger player in the enterprise, gives them a lot of leverage.

        The competition in the mainstream computing space over the next 2 to 4 years is going to be huge.

      • http://search.websonar.com:8080/ Duane Bemister

        If not out they are certainly on the ropes. There is a huge exodus from pc’s to macs. Apple continuously states half of their computer sales come from switchers. The same thing will soon happen with mobile. Apple knows this so I assume the extraordinary capex expenditure must be focused on trying to solve the supply – demand problem.

  • Sacto_Joe

    The only thing I’d caution is that EPS may be about to start compressing. Notice I said “may”. If so, that could deflate the potential stock price somewhat. However, it’s clear that Apple is throwing everything they’ve got at the production issue. Therefore, I basically agree with Nate’s conclusions.

    • chandra

      Who is Nate?

  • http://twitter.com/MacsFuture Lex McFarley

    Great analysis Horace. Where in the 10Q or 10K are you pulling the Capex info from. I see references to equipment but I can’t figure out how you got your number of spending reaching $3 billion per quarter.

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

      Measure the change in asset value to get spending. The result understates spending as assets also depreciate.

  • Giovanni

    Great article, as usual. Congratulations!

    Horace (or anyone), any educated guesses on the TYPE of CapEx you are describing?
    That is, will most of the CapEx be destined, as in the past, to buy manufacturing equipment (to build more hardware iPhones etc, which is where Apple makes most of its revenues)? In that case the correlation “CapEx -> Revenues” is very clear, as indicated in the article.

    However, if most of the CapEx will be destined to build data centers (for web services), that correlation is less clear. Apple has indicated its recently launched iCloud initiative, which require data centers, will be expanded in the next 10 years.
    Also, Apple’s competition with services-oriented companies, such as Google and Amazon, will also increasingly require more Apple capacity for services.
    Apple’s CapEx for data centers might be for services related to Apple TV, maps, or whatever Apple has up its sleeve for iCloud, who knows?
    I am sure these new initiatives would likely succeed, but they are not yet the historically proven cash cows (such as iPhone) shown in the chart.
    So far, Apple services (e.g., iTunes) were aimed to increase hardware sales, but are NOT a major source of revenue by themselves.

    Thus, if a significant part of CapEx is for data centers,
    the correlation “CapEx -> Revenues” (thus the predictive value of CapEx to stock price)
    seems much less clear cut than indicated in the chart.

    • Tatil_S

      $3.8 billion for the last quarter of FY12 is more than Google has ever spent on its own CapEx in any quarter and I presume most of that is for Google’s data centers. High water mark for Google is about $2.5 billion and it happened in one quarter, the rest of them are all less than $1 billion. It is hard to imagine Apple could be spending so much more with only three data centers.

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

      I would note here that Apple’s data center plans appear to be mostly known at the moment (North Carolina, Oregon, and apparently Hong Kong), and it’s not clear that Q3 or Q4 spending could be going very much to any of the known locations, except maybe land purchase and site preparation, which wouldn’t be the major part of it. The solar buildout in NC is probably in there, but it’s 20MW, and solar installations now cost around $1-$2/W, so we’re talking about tens of millions of dollars, not billions here. Also, the NC center was reported to cost approximately $1B. So it’s hard to see this reaching $3B, even if both the Oregon and Hong Kong centers were being built out in these quarters (which the Hong Kong one certainly isn’t).

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

        Actually you forgot one more city on that list – Reno.

  • RichardinMelbourne

    Recently reported Solar Farms powering iCloud (15.09.12 Apple 2.0) could
    put a large dent in some of the recent Capex investment. Capex may not
    be all in Asian supply chain.

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

      Capex is not all supply chain. Some of it is for server infrastructure. The proportion is unknowable, however the scale of spending is beyond what can be absorbed by data center outfitting.

  • Westech

    I am somewhat late to this party.

    Most of the discussion on this blog is about high capex as it relates to new products. I have been astounded that the current process seems to rely on hand labor. I had believed that it was more automated than it now appears to be the case. Automated mass production gives many advantages, one of which is lower cost. One must always run the numbers, which I can,t do because I don’t have them, but with thscale of production that Apple has it likely would result in very significant cost reductions which would give Apple greater leverage over how they design their business plan going forward.

    For this scale of operation the capital cost would be very large. This could be used to increase their margins, or to lower costs and prices to the point that their competitors can’t meet. There are certain caveats. In addition to meeting the return on investment criteria, the equipment must flexible enough to be easily modified to be used on new products, and the cost of doing that must also be taken into account.

    Their are potentially other benefits, including better quality, less wastage, and less dependent on the vagaries of labor. It might also lower costs to make it possible to bring some of the manufacturing back to the USA.

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

      From what I’ve seen in the reports about the Foxconn line, it sounds like most of the assembly is automated up through the subassembly stage, with manual steps only for final assembly, test, and inspection. With all the flex cables in the design, it may be easier to use hand assembly for the final steps. And test and inspection aren’t things I’d trust entirely to automation, except for basic stuff. Machine vision is getting better, but people are still much better at it.

      I would expect the labor/automation setup tradeoff is part of Tim Cook’s planning, and I wouldn’t be surprised to find that human labor is a shrinking part of the assembly process over the last few years.

    • JaneDoe12

      I have been astounded that the current process seems to rely on hand labor.

      I don’t know if this will answer your question, but Horace explained (in a Critical Path podcast) that Apple uses numerical control equipment (CNC) to make the aluminum covering (unibody) for the MacBook Prod, iPad and iPhones. He said that CNC is not designed for manufacturing — it’s normally used for prototypes and small batches. This equipment can produce glass thin enough and able to bind to the touch panel beneath it, and still keep everything in very close tolerances the way that Apple demands it. But to produce enough unibodies, thousands of these machines would be needed! I think that a lot of hand labor would be required somewhere during the process.

      Another interesting point that Horace made was that there are only a few companies in the world that make this equipment. If Apple buys all of it, and the companies allocate all future production to them for years going forward then no competitor will be able to make a unibody. And maybe that’s why we haven’t seen any.

      All of this is from the Critical Path episode #10 – The Means of Production, aired on October 11, 2011. It was an excellent discussion of both the unibody and capex. (The unibody part starts at 22 min.) When Horace explained the capex calculations, you realize that he has a predictive model — he figures out how much is allocated for equipment in factories, and then he connects that to production and volumes.

      The Critical Path #10 – The Means of Production.

  • Mayson

    I wonder how much (if any) of that 3bn is going to TSMC for a new 20nm fab dedicated to A7 CPU? If little of it is, as seems likely, then next year will see some really big capex.

  • Pingback: Analysis: Apple stock headed for $1,000 per share

  • Pingback: Analysis: Apple stock headed for $1,000 per share | Every Single HOUR!

  • Paul

    Is the Cap-Ex scale noted anywhere? I highly doubt Apple is spending close to 100 billion dollars per quarter…

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

      The CapEx scale is on the graph. See the $3,000m in the upper left corner.

      • Paul

        Oh, I missed that. It is still kinda hard to see though. Thanks for the update anyway!

  • tfd2

    Would love to see this graph updated.