Wednesday, January 27, 2010

Fiscal 1Q '10 actual results vs. estimates

Apple beat my revenue estimate by $666M (hmm), an understatement of 4.2% on my part. Source of this is relatively evenly shared between all three main product lines, with Apple's solid crush of Mac units estimate partly compensated by my overshooting the ASP a bit. On a relative basis, most significant is the -6.5% error (-$221M) in iPod revenue which came from compounding a -2.2% error in estimating units (-470K) with a -4.4% error in estimating ASP (-$7). Finally, a 13.9% underestimation of Peripherals revenue is not as critical in absolute $M, but is still quite a nice surprise as this is the second consecutive quarter Apple crushes my peripherals estimate, so I'm now giving a boost to my model's longer-term attach rates for peripherals. Surely some sort of magic going on there.

Over to the income statement, an overshoot in GM% practically erased all of those mistakes in the revenue breakdown. Apple came in 150bp below my admittedly optimistic 42.4% GM estimate (I was tired of hearing PO's excuses for lowballing on this, and decided to completely ignore his air freight and component warnings). Thankfully it saved me from another embarrassing miss, and allowed me to waltz through most of the rest of the income statement line items within a percentage point or so of the actual figures. The only small issues I have is with OpEx which came a little too high (even a tad higher than Apple's guidance which is unusual), and OI&E for which I had ignored Apple's $30M guidance as silly given its cash position. It's inconsequential for now, but I do want to be able to model this based on interest rate trends and Apple's cash and equivalents, for the time when interest rates get back to normal and OI&E becomes a significant contributor to EPS. No luck so far. Finally, the net margin, 130bp below my hugely optimistic 22.8%. I should learn not to get carried away like that, as Apple did awesomely at 21.5% net margin. This was the first thing I noticed since I got crushed on revenue but Apple missed my EPS (and it worried me a bit at the time). I'll have to tone down the scorching profitability level very slightly.

Congrats to those who correctly guessed that Apple would make the accounting change in Q1. Yes, I got it wrong (and no, my reasoning wasn't at all based on any presumption that Joan Hoover would just tell me, a nobody, about such an important news item). I truly believe this was not the best way to go about it. For one thing, this report's figures have no clear relationship with what Apple had guided for back in October. Thus, not only the official record of Apple beating their own guidance is broken (of course they beat, but it's not "officially" sanctioned as in verifiably coming from Apple's own language in its reports), there's also no clear way to go about any top-down guidance analysis for it. You know, the thing many of us do in which we take Apple's lowball guidance, add back an average historically-derived "PO's lowball factor" to it, and aim our sights so our revenue and EPS estimates fall close to that. Not this time, Apple simply trashed their previous guidance for last quarter.

A similar disconnect is happening with investors, analysts, and financial reporters' interpretation of the numbers, in particular when trying to compare them to analysts consensus estimates. Obviously analysts mostly gave the old GAAP figures, some of them gave non-GAAP or new-GAAP, but the most quoted figure by the media was the old GAAP consensus. Turns out Apple by doing it like this has also trashed that gauge, it's no longer a comparable way to measure Apple's supposedly huge beat. So how do we know they really beat? and by how much? Well of course we know Apple, if it had reported based on old GAAP, would've easily beat even that "whisper" thing at $2.30, but people in general don't know and don't care to figure this out themselves (even I had a little conundrum in figuring that out). Apple clearly simply doesn't care for those old metrics anymore. That may be understandable given how those understated the company's performance, but why not provide a mechanism for a smooth transition? I say, respect the previous guidance that Apple itself gave to us, and also respect the analysts' consensus, even if it's way off, because no matter how wrong them analysts are, the consensus is the most quoted figure investors hear and read about, and investors rely on that mechanism for gauging any report against market expectations.

So people are naturally skeptical about all this, and some are assuming there are accounting tricks in play. Surely Apple's performance is quite jaw-dropping, looks highly suspicious to most in the face of the economic crisis, and to top it all most of the public know about the accounting scandals and simply distrust anything remotely reminiscing of an accounting sleigh-of-hand. To those that don't know about Apple's financial reporting preferences and style, I suppose this report can only make sense to them as a trick, after Enron, Worldcom, and all the recent banks shenanigans. Let's not forget Apple itself had it's own accounting scandal. No it wasn't overstating financial metrics to meet or beat expectations, but options. Still, that reinforces doubt. Well, the thing with this one is that it's just the opposite of that stereotype. Apple management tends to act, with regard to the stock, as understated as possible. The CFO and other executives tend to paint things in the least favorable way from a shareholder point of view. Who knows why, maybe they figure it's better to lay low and downplay your stock, perhaps to avoid attracting all sorts of curious eyes (competitors, the SEC, tax audits, envious freaks, and whatnot). Put it all together and you can see how the stock may have reacted as oddly as it did AH and the day after, wavering around $200 when this stock should be worth at least $288 today. It's clear to me the stock is reflecting this self-inflicted fear, uncertainty and doubt (self-FUD, although this time I think unintended).

Anyway, a smooth transition would have been achieved by only issuing new guidance based on the new accounting scheme. Obviously the same question would have come up regarding how to gauge the strength of such guidance against consensus expectations for Q2, and that's why my assumption or recommendation was to report and issue guidance this time as old GAAP, then sometime in the middle of the quarter announce the accounting change, making clear they would restate the last 3 years and forget all about that GAAP/non-GAAP stuff, and leave plenty of time for most analysts to realize they can throw away their old GAAP models and start talking their new GAAP models, and for the media to pick up on this and update the consensus data for the quarter, the fiscal year, and beyond. A bonus of doing it for Q2 is that the GAAP vs. non-GAAP discrepancy would be minimal for this quarter and for the next. So, even if some analysts don't bother or can't figure out how to make the switch, it still doesn't make the consensus worthless, because the effect is minimized. Thus, investors would have been much better able to appreciate Apple's beat. Only then, you would have a sensible way to measure whatever Apple reports against the market expectation. This, what they did, wasn't a trick of course, some of us know. But there's no clear and indisputable way (through official company filings) to tell exactly by how much it isn't one.

My performance details below. Apologies for the estimate fudging. It's all derived from what I published a few days ago, and a couple of things from my spreadsheet as it stood back then. Read the notes to see how.


. Est(*) Act Err Err%
. -------- ----- ---- ------

Units (K):
Mac 3184 3362 -178 - 5.3%
iPhone 8650 8737 - 87 - 1.0%
iPod 20500 20970 -470 - 2.2%

ASP ($):
Mac 1354 1324 + 31 + 2.3%
iPhone 619(1) 638 - 20 - 3.1%
iPod 155 162 - 7 - 4.4%

Revenue breakdown ($M):
Mac 4312 4450 -138 - 3.1%
iPhone 5351(*) 5578 -227 - 4.1%
iPod 3170 3391 -221 - 6.5%
Music 1177 1164 + 13 + 1.1%
Perph 404(2) 469 - 65 -13.9%
SW 604 631 - 27 - 4.3%

Income statement ($M):
Revenue 15017(*) 15683 -666 - 4.2%
COGS 8657(3) 9272 -615 - 6.6%
GM 6360(*) 6411 - 51 - 0.8%
OpEx 1625 1686 - 61 - 3.6%
OpInc 4734(*) 4725 + 9 + 0.2%
OI&E 55 33 + 22 +67.8%
Pre-tax 4790(*) 4758 + 32 + 0.7%
Tax 1365(*) 1380 - 15 - 1.1%
NetInc 3425(*) 3378 + 47 + 1.4%
Shrs. 915 920 - 5 - 0.5%
EPS 3.74(*) 3.67 +.07 + 1.9%

GM% 42.4% 40.9% +1.5% + 3.6%
OpInc% 31.5% 30.1% +1.4% + 4.6%
Tax% 28.5% 29.0% -0.5% - 1.7%
NetInc% 22.8% 21.5% +1.3% + 5.9%

(*) Shows differences in these estimates compared to previously published estimates due to accounting change. See corresponding explanatory notes for special cases below (all other new figures marked with (*) are derived by applying already derived new estimates as inputs in its standard formula, e.g. newTax = newPretax * oldTaxRate).

(1) iPhone ASP - Looking forward, it makes sense to include "other" iPhone-related revenue (carrier payments and iPhone accessories) by rolling it into the ASP for simplicity of calculation and the subscription accounting requirement to separately calculate the handset hardware revenue recognition having been removed. Also, I've decided not to estimate the precise effect of deferring the $25 value of the rights to upgrade the iPhone software ($10 for Apple TV), and instead assume the recognized portion of this as attached to just the current period's units, also for simplicity and because the differences are relatively meaningless when compared to the aspired accuracy in estimates. This last simplification tends to penalize the ASP on those quarters with much stronger unit sales when compared to the trailing 2-year average, somewhat offsetting the first assumption which rolls in related revenue that isn't necessarily attached to each unit sold. For this F1Q2010, an originally estimated $247M worth of "other" iPhone related revenue (about $29 per estimated unit) that had been excluded from the original $590 ASP estimate has been added back, yielding $619 estimated ASP including all iPhone related revenue sources. Notice Apple stated an iPhone ASP of about $620, so backing that out of the $638 ASP as I'm modeling it, it reveals $18 per unit (about $160M) of this "other" stuff, which is well below the $29 per unit ($247M) I was estimating.

(2) Peripherals and Other Hardware revenue estimate now includes full recognition of ATV sales. This is derived by subtracting all other estimated revenue sources, including the new iPhone and Related Products and Services revenue estimate based on the new ASP estimate explained in (1) as an input, from the estimated non-GAAP total revenue published previously ($15,017M).

(3) New COGS estimate is derived by estimating a similar proportion as in previous company reports for the old non-GAAP adjustment to COGS over non-GAAP adjustment to revenue (close to 30%), as follows: take 30% of the difference between the original and the new estimates for iPhone and Related Products and Services revenue (.3*[5351-2815]=761), and 75% of the difference between the original and new estimate for peripherals revenue (.75*[404-392]=3) as representing the iPhone and ATV adjustments to COGS, respectively, and add these to the original COGS estimate of $7,893M, resulting in 7893 + 761 + 3 = $8,657M. There are many other ways to derive a non-subscription accounting estimate for COGS based on iPhone revenue and estimated gross margins, so feel free to try them out and let me know what you come up with. It shouldn't be too far off if you stick to sensible estimates for iPhone GM%.


JavaJack said...

thanks for all of this. Very interesting. I wonder when the stock will rise to the level it should be at?

Roy said...

Thank you for sharing your outstanding analyses. I think a lot of people recognize AAPL stock should be in the $280-300 range to be consistent with the rest of the market.

Any insights on why AAPL is struggling to hold the $200 level, while other companies have been rewarded for positive surprises (AMZN, VMW to name two)?

Is it a case of just too much good news too fast? (Earnings upside, accounting restatement, iPad, all in a span of 2-3 days)? Some 50-60 million AAPL shares were dumped at declining prices over the last two days, and I don't understand the rationale for it.

Roy said...

Sorry - there was no accounting restatement, I meant to say change in revenue recognition.

Daniel Tello said...

Roy, I have no idea about a rationale for AAPL being undervalued so much. Back in late 2008 / early 2009, I know WS analysts were expecting something like $5 EPS for FY 2009, and many had targets around $100. I thought that might explain why AAPL was trading at less than half its fair value.

But this time, analysts have their estimates at the right place, or at least in the same ballpark. Their average target is about $250 which is still low, but not too bad. I say give it a few weeks and all will be fine. I can't see this continuing beyond the next report, but it might take two.

Horace Dediu said...


Excellent work as usual. Regarding valuation, it's impossible to get an idea why Apple is considered to have the same P/E as Microsoft. I don't even try.

See a graph of the P/E of Apple pre- and post- restatement here:

Daniel Tello said...

Thanks for the link Horace. Added your asymco blog to my links on the right.

Your chart looks much like the one in my previous post, except I only plotted quarterly data points. I wonder how you keep the P/E from making big jumps after each report as the trailing EPS is updated substantially in some cases. Did you do some kind of daily interpolation between the quarterly EPS points? Cool stuff, thanks.