AAPL Model Comments Sought

  • Posted: 23 April 2011 08:00 PM

    First, a quick shout out to DT for continuously challenging us newer amateurs (been here under a year) to continue to think, model and participate.

    I realized today that I needed a new AAPL model, so I thought about what variable I was hoping to measure. Obviously I care about stock price, but that’s too dependent on external factors so the closest I thought I could get was EPS. Measuring future EPS (for me) was trying to project future revenue (sales) and then working the EPS from there.

    I’m looking for comments on the following logic to predicting future quarterly product revenue on a per product basis, taking (for example) Mac’s.

    1) I took the quarterly sales for 10 quarters (FQ1 2009 to FQ2 2011) and calculated the total trailing 4 quarter revenue on a quarterly basis (i.e. 7 values over the 10 quarters).

    2) I took these numbers and calculated the percentage change of this rolling annual total from quarter to quarter.

    3) I calculated the average and standard deviation on this percentage change found in step 2

    4) I projected the revenue for next quarter based on maintaining this average for the trailing 4 quarters minus some percentage of the standard deviation (anywhere from 0% to 300%) to factor in risk and uncertainty.

    This certainly isn’t the most sophisticated model for projecting next quarter revenue for a given product, but am I missing something obvious that makes this approach just downright bogus?

  • Avatar

    Posted: 23 April 2011 08:29 PM #1

    I think you have a great start.  Many of Apples products also exhibit some seasonality so don’t forget YOY.  I also like to forecast YOY product launches as the drift in these events tend to affect a particular qtr’s growth rate for that segment.  On the laptop/PC keep an eye on Intel’s release cycle.  The iOS hardware is more an internal cycle so I keep an eye on the production rumors.  A company as large as Apple can’t hide everything and if you look at what competitors are producing you can make some educated guessing about the timing.  Even with all this great insight you can miss due to unpredictable events like a production ramp problem or subcomponent shortage.  The latest qtr iPad estimates serve as a good example.  As a new product it was difficult to determine the seasonality.  Additionally the product was transitioning to a new model which involves clearing out inventory and ramping production.  I think most of us assumed when Apple announced a early release in Mar, they had already ramped production, but based on the actual numbers, the ramp was slower then I modeled.  Good Luck look forward to you joining the rest of us.

  • Posted: 23 April 2011 09:11 PM #2

    pats, thanks for the great comments.

    I thought about trying to get year-over-year into the model, but thought that the rolling 4 quarter non-weighted total would suffice given each total would include 4 consecutive quarters.

    I appreciate your insight and may consider some of those concepts in the future. For now, I was shooting to be as objective as possible (i.e. not trying to predict unknowns or intangibles that might happen).

    For anyone who is interested, here are the numbers (average and std. dev.) for each product (iPad not included, not enough data yet); the method of determining these numbers are described in my first post.

    Mac mean=6.02% stdev=0.62%
    iPod mean=-0.08% stdev=1.86%
    iPhone mean=19.00% stdev=3.58%
    iTunes mean=5.37% stdev=0.99%
    Peripherals mean=6.47% stdev=2.53%
    Software mean=2.77% stdev=2.45%

    The Mac growth is nice and tight (small stdev)
    The iPhone would be much tighter stdev except FQ3 2010 was an outlier—I believe it was the switchover from the iPhone 3GS to 4.

  • Avatar

    Posted: 23 April 2011 09:40 PM #3

    Excellent start.  I usually just eyeball what you calculated.

    Additional factor: consumer confidence and employment figures.  rising or falling ?

    Additional factor: upgrade cycle ?  The Q3 quarter hit a bit of both.

  • Posted: 23 April 2011 10:10 PM #4

    Treehouse - 24 April 2011 12:40 AM

    Excellent start.  I usually just eyeball what you calculated.

    Additional factor: consumer confidence and employment figures.  rising or falling ?

    Additional factor: upgrade cycle ?  The Q3 quarter hit a bit of both.

    I have been just eyeballing it for awhile, but wanted to check if I was just seeing what I was hoping for.

    The conclusion I came to was that it depends on how pessimistic/optimistic I am on PE compression and discounting the future by assuming growth decelerates. I can get my numbers to show the entire range predicted by Katy Huberty from the high 200’s to the mid 500’s.

    I am encouraged that midpoint numbers with close to historical growth and current PE yields numbers in Andy Zaky’s range. It looks like if I toss in a dose of pessimism, the future still looks green.