Categories

A new way to value Apple

Almost all valuation models for Apple assume it’s a hardware company. The modeling algorithm for hardware is simple:

  • For each year in near future
    • For each product line
      • Compute contribution
  • Determine company value by summing contributions and multiplying by a P/E ratio

The major difficulty is in predicting the growth of each product line. This is difficult because buyers can be fickle. Companies employs all sorts of tools in order to secure repeat customers but if switching costs are low, the company can crash in value. For this reason, analysts pick various ways to predict device  sales. Some choose to index each product on production (channel checks), demand (customer surveys,)  or even on a top-down share of total (market research on growth of whole market and assumptions of share).

Although simple and convenient, this model does not probably match the way Apple is managed. The company does not build hardware products to sell and forget. It builds platforms which are best seen as sources of recurring revenues. Users are incentivized to continue buying devices. iCloud, iTunes and other Apple properties are expressly designed to that goal–and they are not cheap to run. So, the company must see the world through a different set of lenses than the default model shown above.

To think about the business like they do, we need to put the data through a similar set of lenses. I began with a modest proposal last month to value each user as a source of recurring revenue. Now it’s time to expand on this method.

The following chart shows the recurring revenues per user by product and the current and possible future size of user bases.[2]

Revenues/yr/user are bars with the left axis and the User bases are various colored circles indexed with the right axis.

The assumptions that went into the data are as follows:

  1. Device life span is as per previous article linked above
  2. User base growth for 1 year is: Mac 20%, iPhone 80%, iTunes 10%, iPod 30%, iPad 100%
  3. User base growth for 2nd and 3rd years (recurring): Mac 20%, iPhone 50%, iTunes 10%, iPod 20%, iPad 80%.
  4. The Revenue per user is assumed constant

Note that the user base growth figures are lower than the product growth levels seen historically. Obviously, the model is highly sensitive to these growth assumptions so they need to be scrutinized and tested rigorously.

Nevertheless, as a straw-man proposal,  the recurring revenues for all these products[1] is shown in the following chart:

Once the income is estimated, we can take that value and assume a profit margin (net) and then multiply the earnings by a multiple. Using a 20% net margin and 12x P/E yields the following chart.

(I added an assumed level of cash in green.)

This model would imply that the company today could be valued at $208 billion on the basis of its installed base alone. That value would be about $323 billion in one year and $620 billion in three years. Dividing by the number of shares outstanding (935m this year increasing at 2% a year) yields a share price of $222, $339 and $629 by mid 2014.

These values can be considered “lower bounds” on valuation since they assume income from previously secured customers. The speculative part of investment would be based on what the future bases will look like (so, for example, if one believes these assumptions, the $629 figure could be considered a target to be discounted to today).

Notes:

  1. Excludes Software, Peripherals, non-iOS devices, and any other service revenue.
  2. Compare this char to social media companies (MIT Technology Review)
  • http://twitter.com/realinformatik @realinformatik

    Gee Horace, thought provoking to say the least (esp for AAPL shareholders I suspect). That first chart is a thing of beauty and clarity. Do you make these charts with Numbers (IIRC)?

    • asymco

      I only use Numbers.

  • MattF

    Well, OK… How about the other side of the bookkeeping ledger? What fractions of the income from each user go to the various platform costs?

    • asymco

      The model takes a simplified view of the cost by assuming 20% net margin.

  • Lawrence

    Horace, didn't you mean to use Revenue in the labeling below the bar graphs on the 2nd chart?

    • asymco

      Yes, sorry. That's fixed now.

  • http://www.aktually.com Andrew

    Horace, this is really interesting stuff! I agree that "Users are incentivized to continue buying devices." Would you be open to doing the same analysis on, say, Google's Android platform?

    • asymco

      Of course, but we don't have any idea of what Android users are worth to Google and we can't assume anything about the cost structure of Android either.

      • http://twitter.com/WaltFrench @WaltFrench

        Methinks you're overly modest about your ability to understand Google. They ALREADY price as a service; gaining essentially zero revenues from hardware. Although they have serious competition in China and elsewhere, they enjoy very high penetration ratios in the US, so they can grow at the same rate as the number of total web users. If today's smartphones and PC usage is dominated by the upper half of income, new users will be less attractive to advertisers.

        While being ever-more-connected thru mobile and all, eventually they max out in terms of the number of actual buying decisions / brand impressions they can influence: I only buy so much coffee, only so many cars; only so many ballet tickets. (SF Ballet ads show up on half the web pages I've visited of late, perhaps because as a subscriber, I used Google to check performance details.) So if I'm exposed to twice as many ads as I am now, each one will be worth perhaps half as much to the seller. Google has plenty of market share versus TV ads and perhaps they'll become a serious way to advertise low-SES products, but Google's “genius” in matching viewers and ads would seem a lot less relevant when you're promoting a product consumed by half the population.

        So, I find it hard to see rapid growth in revenue per user; in fact, revenue could grow at a rate barely faster than the online population.

        At the same time Apple is attempting to provide multiple capabilities to users, Google has long played a defensive battle against Apple in trying to prevent being locked out of mobile usage, more recently against FaceBook in trying to prevent it from being the Home Page that people never leave; and for years against Microsoft's modest share of search that they could yet grow thanks to their Office-financed war chest and potentially, from an web-enabled Office suite.

        Google's mobile strategy seems to be to suck all the oxygen from the room in terms of making it impossibly expensive for others to justify investment into a mobile ecosystem, while leaving exactly zero pricing umbrella. But Apple is ahead of them in building an expensive ecosystem to support the iProducts. Microsoft according to the CW, will also refuse to cede the mobile space. They *have* claimed Symbian; they *have* made it impossible for RIM to get back in the game (witness the headcount-reduction announced while they are supposedly in the race of their life); I don't see a credible scenario where WebOS can ever be aggressive enough on performance and features. Their strategy HAS been extremely successful. But MSFT and AAPL are fighting back on the cost front and sales numbers suggest we're settling into relatively fixed share positions.

        Google is here to stay in search, ads, a whole host of nifty web services, and mobile. It's already measured as a service. They can diddle with the RPU but their current model seems to be working about as well as it is likely to going forward.

  • EWPellegrino

    Seems like the only way to attack those numbers would be to posit either Apple growth dramatically slowing, or margins dramatically shrinking.

    Apple is currently at 10% of the US handset market by existing subscribers, the growth rate assumed would put it at 40% in 3 years, or around the entire current smartphone segment. That may be conceivable, but I think we'd be hard pressed to call it conservative. Europe is at 12% for the biggest 5 markets, requiring nearer 50% share to meet those growth requirements.

    EDIT: European number is wrong, and based on a misreading of a ComScore survey

    Obviously It is a given that Apple will be gaining an increasing portion of those new subscribers in markets such as China – but can it hope to do so to the required extent while preserving current margins and current revenue per subscriber per year.

    I guess the question is, how high do you see Apple's market share in handsets going in developed markets, and how far can it go in developing markets at current levels of profitibility.

    • asymco

      The growth rate is not indexed to US users. The proper sanity check is against the global market which today is 5.3 billion connections. If we assume it will be 6 billion in 2014, Apple's share of that (assuming about 450 million base) would be 7.5%.

      The market assumptions I use are that smartphones will be in use by >80% of developed market users and at about 50% in developing markets. I can cross-reference that against the assumption that Apple can capture 20% share of all smartphone users.

  • Pingback: Monday links: out of sync | Abnormal Returns()

  • Ed L

    Nice alternative view on the data. Footnote (1) excludes software…can you clarify then what is in the iTunes bucket? Since it appears that the per year figures are averages of platform lifetimes and revenues, that to me does not appear to be the key insight. The "per user" tracking on a recurring basis is the interesting part. And what I then wonder (though the data may be hard to obtain is), how revenues per year break down among: (1) existing users (halo/platform/network effect), (2) new customers (the retail PR story), and (3) how many customers are lost (wonder if AAPL even knows this answer). A second area that would be really interesting is to understand the halo effect (or as HW vendors like to often describe, "pull-through"). So per device owner, how much additional $ gained via apps, content, accessories, additional devices. Understanding that picture, we understand the key pressure points that will drive revenues. And the discussion then becomes a lot closer to those surrounding true subscription players, content providers, etc — cost of customer acquisition, $/user, new users, lost users, etc.

    • asymco

      "Software" in Apple's income statement includes sales of iWork, iLife, OS X and the Pro tools and any other software titles Apple sells.

  • davel

    Your assumption about Apple being a hardware company clearly illustrates the reasons for the recent slide in the share price. Because no new shiny toy was introduced ( even though Apple almost said as much leading up to their conference ), the street was negative on the company and its stock price suffered.

    Now the same analysts are repeating what Apple said last quarter and upping their estimates leading into reporting day tomorrow.

    The iCloud announcement if it is executed properly brings important features to the Apple ecosystem. It brings badly needed features to the Apple devices as well as extending the usefulness of existing devices.

    The skeptic in me says this is price manipulation. I cannot believe that they are collectively this stupid.

    • asymco

      The way I came to believe in this model was by thinking what value does a competitor get from a new device sale vs. what Apple gets. Clearly, the competitor will be nervous about where to find the next customer and the one after that. The preference of iOS users to remain with the platform must be worth something. However, "the street" does not use this model. It uses the more traditional model that assumes that customers are not loyal. This model is closer to how social networks or even Microsoft are valued as recurring revenue models.

      • davel

        Thank you. Your analysis makes sense of something I have been struggling with.

        It seems to me that customers are loyal. In talks with friends and acquaintances the consistent message is I like or do not like my carrier; I like or don't like my phone. Many people I know hate ATT for one reason or other. They may stick because of various plans they are tied to. Most people I know love their iPhone or Android phone. I find that if they have HTC they tend to stick to that. If they have Motorola they will buy the next one. So there seems to be a certain amount of loyalty from what I can see.

  • Dan

    So you are predicting that AAPL's book value is $222 today and will be $339 and $629, not to be confused with stock price target?

    • asymco

      I would not say book value, and it's not a stock target either. It's the base value of the platform side of the business.

  • http://twitter.com/pcad @pcad

    Do your models assume that value from one column affects value in another? Just pulling numbers out of thin air: iTunes is worth $25 if it's a Mac User; $19 for an iPad user, but $52 if the user is a Mac User who also has an iPad.

    It seems to me Apple views the platforms as being multipliers. It could be that increasing Rev./User/Yr is almost as important as increasing # of users. In the face of stiff competition, margins on bare devices might be the most vulnerable.

    • asymco

      The value of a user is determined based on reported sales over a few years and the number of active users (presuming they've been active over that time frame). In the case of iTunes the revenues are over one year and the number of users is as reported by Apple.

    • name99

      "In the face of stiff competition, margins on bare devices might be the most vulnerable."

      I don't see this as a useful viewpoint. Apple is not in the business of selling "bare" devices; it is in the business of selling devices wrapped in more value than its competitors. Certainly to EXISTING users, an iPad at $600 is more valuable than an equivalent hardware Android Tablet at $450.

      Beyond this, as I have mentioned many times, Apple is playing a DIFFERENT GAME from its competitors. Apple perceives what its competitors apparently do not — the future is not a SINGLE device per person, it is each person using a multitude of devices with different form factors optimized to the human body. In such a world, the value is in the collection of devices and how well they work together, more than in how well each device works by itself. This is the point of iCloud — not to duplicate DropBox.

      Google kinda sorta gets this — but is so obsessed with insisting on the net and remote storage that it engineers stupid solutions which don't incorporate well the fact that bandwidth costs, and that one is frequently not connected to the net.

      MS? Ms probably has fifteen different views on the subject. What will happen is that .Net will adopt one view, Silverlight will adopt another, Win Phone will adopt a third, Bing Live Pro Plus Grid (or whatever they call their internet properties these days) will adopt a fourth, and the result will be a horrendous clusterfsck that covers every bullet point, provides lots of employment to MSCEs, and zero joy to anyone.

      In this world — and that is Horace's point — once you enter the Apple eco-system you have little incentive to leave. As long as the constellation of services surrounding the hardware is good enough, you'll stick with Apple because moving your iPhone world + your iPad world + your Mac world all at once is just too painful.
      Meanwhile, in the Windows or Google worlds, if your devices don't do a very good job of working well together, you're not any worse off switching your phone tomorrow to an iPhone while sticking with a Windows PC (or whatever).

      • Wapjog

        Google cloud services (such as mail and calendar) have "offline" access since a long time. You probably meant Music offering in the cloud. On that one, right, for the time being, Google offers only cloud storage; but compare apples with apples: This service is competing with nothing from Apple. You are weirdly comparing it with a yet nonexistent competitor from Apple. How do you know Google is not working on or buying some company that has already negotiated some music rights, and intends on offering a similar product as the one described by the iCloud announcement?

        On .Net, Silverlight, et al, you are completely confusing the different layers of platform, runtime, SDK, and software just to write a tagline, I do not think it makes sense at all.

      • name99

        No, I did NOT mean music in the cloud. Music in the cloud is a (not very interesting, IMHO) variant on DropBox.The INTERESTING part of iCloud is that it is a set of APIs. What are these APIs? They are ways for an app running on device A to change state ins ucha way that the changed state is picked up by other devices B and C.So, to take the simplest example to explain, as I am reading through an eBook (in a 3rd party app, not iBooks), every time I change the page, the app indicates to iCloud my new page. I open the eBook on a second device — and it automagically opens to the page I was at on the first device.A more sophisticated version of this would be something like Instapaper, which stores a numbers of articles that I am reading on multiple different devices, and for each article retains where I am in the article. The Instapaper example shows some of the issues here including- what is stored on different devices? For “small” content, like articles, it makes sense to store the full library of content on each device. For “large” content it does not, and a well-written app will allow a user to maintain different libraries on different devices.- what is synchronized? Instapaper understands that a person may well want to be reading a different article on his iPad and on his iPhone. It ALSO understands that if reader switches from the iPad article to the iPhone article, it makes sense to open the article at the position it was at on the iPhone.In other words in this case it makes sense to synchronize page numbers across devices, but not what is open on each device.Instapaper provides this by writing a whole lot of code and maintaining its own server. Apple wants this capability to be open to all apps.Apple understands that it can't decide what is best for each app to synchronize, so what it has provided is APIs and the backend servers to allow apps to specify in a number of different ways (whichever makes sense for their app and coding style) what is to be synched and how. Apple takes care of the horrible technical issues — all the networking (including how to minimize bandwidth and power), conflict resolution etc.There are many aspects to UI of this that are not yet clear. For example — Chrome to Phone and similar Safari to iOS apps. One way to get a web page you're reading to your phone is to explicitly push it in the browser. Is this the only, or best way? What if you are away from your mac and realize you want to continue reading that page?What if Safari on iPad had something like a souped up version of the Bonjour page — you go to this page in Safari on iPad, and it shows you the devices (macs, iPhones, etc) connected to your iCloud account. You choose your iMac, and it shows you in some form (mini pages? URLs?) the web pages open on your iMac. You choose the one you want, and the page is opened on your iPad.This is the kind of thing we are moving towards.Like I said — the point is to make using all your devices together easy and a pleasure, rather than a chore where you are constantly complaining about how something is on one device but you want it to be on another. If Apple does this properly, most users won't even really notice — they'll just find that, day by day, their devices, in some difficult to describe way, just seem to keep working better; more and more often, they're able to just pick up work and state from one device and continue in another device.

      • http://twitter.com/fiftysixty @fiftysixty

        Very good points, and as a developer I agree fully. There are shortcomings to the iCloud approach (it's tied to a single account, so no inter-account syncing, for example) but the upsides of what it brings will greatly overshine the downsides. And, as you pointed out, the really big deal of iCloud is not what Apple apps do with it, it's how it can be integrated into 3rd party apps.

        If often thought to myself that the design thinking in Apple really permeates everything, not just the design of hardware and software user interfaces: the APIs are really beautiful in their own way. In my opinion, Apple has succeeded in providing APIs that offer just the right mix of flexibility, power and ease of use. And that goes for iCloud as well: if you're using standard methods for storing data on the device, there is virtually no learning curve to migrating to iCloud. The bulk of the work is in designing the user experience and interface, and it seems to be the way Apple wants it to be.

        So, what iCloud will do is it will bring inter-device syncing within the reach of every 3rd party developer. And in a year from now, we will have apps that use iCloud in really creative ways, and we will have nearly all apps use it for basic syncing. And once the users get used to it, anything else will feel really cumbersome. I suspect Google could create the same kind of service for 3rd parties, but will they? Or are they satisfied with what they already have in place?

  • Michael Anderson

    Actual net margins will depend heavily on the mix of products sold. I believe margins are highest on the iPhones (over 50%), in-between on Macs (30–35%), and lowest on iPad/iPod Touch (~25% or less). Of course, sales success will also depend inversely on margins. As long as Apple keeps iPhone margins at 50%, I'd expect Android to continue maintaining a worldwide unit sales lead in phones. Conversely, if iPad margins stay at current levels, I'd expect Android tablets to continue to struggle.

    That said, a 20% net margin seems suitably conservative.

  • Steven Hofman

    Wonderful model and actually pretty conservative.

    The only assumption were I beg to differ is your projected growth of iPod. I would assume that growth rates will be much lower (maybe even negative) as people stop using stand-alone MP3-players and switch to iPhones and Android phones instead of carrying around an additional device.

    • Michael Anderson

      I can't speak for Horace, but I'm pretty certain the iPod column only includes iPod Touches. Otherwise it would be much higher. It's not clear to me that iPod Touch sales should decline. Some of the sales will be cannibalized by phones, but it's also a very recent market that is nowhere near saturation.

    • asymco

      iPod data represents iOS only, i.e. iPod touch.

  • http://twitter.com/alvalassopoulos @alvalassopoulos

    Very interesting approach. Just a quick note though. I believe that iTunes user base was build by iPod users who to a great extend at least so far overlap with iPhone users. Nonetheless, as iphone users grow in numbers so should iTunes ones. Thus I believe we should expect a great increase in iTunes user base as well – though the small revenue per user makes the difference insignificant.

    • asymco

      That may be but I note that with iOS 5 comes PC (iTunes-)free mobile computing to Apple devices.

      • http://twitter.com/fiftysixty @fiftysixty

        My understanding is that the iCloud works tied to an iTunes account, meaning that multiple devices will be linked to a single iTunes account, with purchases, music in the cloud, app data etc. shared between the devices. If that is the case, then a good percentage of new users will go straight to new iTunes users as well. Of course, in families the devices will likely be tied to a single account so that family members can share the purchases and sync photos etc. so the conversion rate won't be 100%, but my understanding is that there definitely is a conversion rate. PC-free means iTunes-desktop-software-free, not free from the iTunes store. At least, as far as I've understood but I'm not totally sure of this, and I might be wrong.

  • disc1979

    Horace, did you put a P/E multiple on EBITA?

  • Jeff

    Horace, what are your thoughts on the crazy earnings beat? 7.79 per share… 125% earnings growth…

  • John

    It is not that Apple somehow lures people in with shiny gadgets. There is a lot of deep thought, design and engineering that make their products incredibly useful for customers. Using iWork and AppleScript on OS X has given me a lot of leverage to produce things in my career I wouldn’t have attempted otherwise.

  • http://www.fxexchangerate.com/ fxgeorges

    Apple might be more evil than microsoft if they take over. I want nothing to do with the ifad craze going on.

  • Pingback: From smartphones to televisions, how Apple will avoid the fate of RCA()

  • Pingback: Is Innovation Valuable? | asymco()

  • Balaji Anand

    Horace when u were explaining about the general method to evaluate the enterprise value for hardware companies you have mentioned we should sum up the contributions and multiply with P/E. But I hope we should be taking net profit rather than contribution to multiply with P/E.

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

      Of course. The method illustrated assumes a net margin. (Revenues * net margin = earnings)

  • Pingback: LATEST TECHNOLOGY » Apple’s worthless innovation process » LATEST TECHNOLOGY()

  • http://www.seoserviceaz.com Arizona SEO

    iTunes account, with purchases, music in the cloud, app data etc. shared between the devices. If that is the case, then a good percentage of new users will go straight to new iTunes users as well.