This is a simple observation, but one that I believe is critical. In its most recent earnings report, Intel (INTC) announced that it would be spending $13B in CapEx, up ~$2B from 2012 levels. Now, I know some people are screaming and shouting about Intel's ~50% factory utilization rates and think that Intel is incredibly stupid for spending more on CapEx in light of the low utilization, but the more I read the tech press and comments from individuals here on Seeking Alpha, the more I realize that most people are completely and utterly confused about what Intel was talking about when it gave the CapEx number. Allow me, then, to explain.
That Extra $2B Is For 450mm
The capex number of $13B actually includes an extra $2B for initial 450mm wafer plant development. This is for much further out into the future (think 2017+ timeframe), so this spending isn't going to be depreciated this year. To quote from the most recent call,
Ross, there's two parts to the CapEx this year. I mean, two large parts to the CapEx this year, Ross. One is, you are starting to see in our first significant investments in bricks and mortar and some equipment for the 450-millimeter transition which happens later part of this decade. We have to make to those investments earlier. Now that we've solidified our relationship and contractual relationship with ASML, it gives us line of sight to that conversion and therefore we're now in a better position to predict the exact timing and deploy capital for that, so I would treat that as a more of an extraordinary event that's not related to the day-to-day business in terms of volume in 2014, 2015, 2016.
So for the actual production CapEx for 2013, for the unit volumes the firm is expecting, Intel is only spending $11B, which is about flat from 2012. This level of capex seems perfectly reasonable since demand across the firm's products is expected to be overall up as the data center business grows, headway in tablets is made, and as the PC stabilizes (and perhaps begins to recover with the Haswell launch). Intel isn't increasing the amount that it's spending on capex for 2013's products, so please ignore the noise that a lot of the tech press is making around this.
Intel isn't overspending for 2013, but it is putting in the brick & mortar, coupled with some equipment spend, for the 2017+ timeframe at which point Intel, along with the rest of the industry, moves to 450mm wafers.
The problem with Intel is that people want short term results from a business that is making very carefully chosen long term strategic bets. If you want to swing/day trade, stay away from Intel - this isn't for you. The technology here is advanced, the business is well run, but it takes time for Intel's investments to pay off. This means, naturally, that it will take time for an investment in Intel stock to pay off. The 4.3% dividend is a nice bonus, but anybody buying the stock should know that there is a long term story that needs to play out here. Intel has executed very, very well, but it won't be until 2H 2013 that Intel's big bets either pay off hugely or flop.
I'm betting a good chunk of change that the bets pay off, but I derive a good deal of comfort from knowing the company, its history, and its technology inside and out. Intel isn't for everybody, and that's why despite my own bullishness I have not been explicitly recommending that my readers buy the stock. That being said though, the capex isn't a problem and looks to be just about right...that extra $2B is for much later, and therefore won't be depreciated until much later.