Intel : The power of a monopoly

August 11, 2007

A company with a monopoly is the best stock to buy. Of course, anti-trust laws have rightfully abolished most monopolies from existence, but in a few niches, natural monopolies still exist. Many of the richest people in the world became rich because they owned monopolistic companies (Bill Gates, Carlos Slim). The CPU market is a natural monopoly. CPUs are one of the most complicated chips to make, costing millions to research and design, and then millions more to build the fab plants capable of making the complicated designs. The manufacturer who has the largest market share is able to spread these enormous costs over a larger number of chips, attaining lower costs per chip, better brand recognition, and more funds for developing the next generation of chips. The second-place competitor has no chance against these staggering odds, unless the management of the leading firm is almost criminally negligent in running the company. Market share is of paramount importance in this industry.

Intel has previously stumbled. They rested on their laurels, diversified into other lower-margin chip businesses, while incorrectly believing that consumers want ever more powerful chips regardless of power consumption. AMD, their scrappy competitor, correctly divined that power consumption and heat dissipation are becoming key issues, and designed chips with those assumptions in mind, gaining market share in the server segment. Otellini realized the danger when he became CEO of Intel in 2005, and plunged Intel into a price war with AMD, while at the same time drastically changing the research direction of Intel’s new chips, and selling off non-core businesses to concentrate on CPUs. This price war has resulted in declining earnings and sharp drops in stock prices for both companies.

The basic structural fact is that AMD has greater unit costs than Intel, and therefore cannot possibly compete on price. Their strategy of taking the technological lead was viable when Intel wasn’t paying attention, but now that Intel has realized the danger, every single technological advance AMD comes up with will very likely be either already accomplished by Intel, or will be very quickly copied. Intel has already re-taken the technological lead from AMD; its latest chips are already runs faster and cooler than AMD’s, while AMD is trying to play catch-up with its Barcelona chips. Furthermore, Intel is more advanced in the timeline to shifting to the 45 nm production process, which should herald a new generation of faster chips. The war will be over when AMD can no longer sustain the price war and either dramatically cuts research expenditures, or raises the prices of its high-end chips and thus gives back the market share gains to Intel. I believe that Intel will succeed in forcing AMD back into its previous role as provider of cheap lower-end CPUs, which will allow Intel to regain its historically high profit margins (it might be counter-productive to actually kill off AMD completely, since that is likely to immediately raise anti-trust issues in many countries).

I believe that Intel at present prices is a great value, and a testament to the short-term time horizon of most investors. The price war has forced Intel’s gross margins from the height of 59% in 2005 down to 47% in the latest quarter. Revenue has slid from a high of $38.8 bil in 2005 to a low of $35.3 bil in 2006. Net income for the first quarter of 2007 is already 20% higher than last year. Continued cost cutting as well as the rapid growth of CPU sales in foreign markets should continue to lift earnings, and an end to the price war at an indeterminate time in the future will cause a huge jump in earnings. I expect that Intel should achieve earnings of $6 bil in 2007, and possibly $9 bil or more in 2008. At a PE of 22 (corresponding to a 10% growth for 10 years at a discount rate of 9% by DCF analysis), INTC is fairly valued at 2007 predicted earnings, and will probably rise to around $35 in 2008.

I view Intel as a stock with excellent long-term value. While I typically look for a much larger margin of safety, I also think that in this case, paying up for a quality company is worth it. Accordingly, INTC is at this time my largest holding.

{ 2 comments… read them below or add one }

balle August 13, 2007 at 4:22 pm

Even though I agree with the broad assertion that INTC is a good buy at these level, I completely dis-agree with the basis for the reasoning. Actually, the reasoning presented is shallow and made on false statement

“The basic structural fact is that AMD has greater unit costs than Intel, and therefore cannot possibly compete on price.”

This statement is wrong. Actually, AMD has always done better on COGS than INTC. They have always had a smaller die size which basically gives you more wafers per die and hence a lower cost die. Because of the lower COGS, they have historically sold prices CPU and are still in business.

“The price war has forced Intel’s gross margins from the height of 59% in 2005 down to 47% in the latest quarter.”

Lower margins is new reality for INTC and Wall Street will have to adjust to that. The higher INTC margins came from being a monopoply where OEMs had little to no choice in terms of CPU vendor. That has changed. They WILL SELL AMD even if AMD parts are not performing as well, just keep INTC in line.

valuegeek August 14, 2007 at 4:25 pm

I agree that AMD has a lower COGS than Intel. However, when I mentioned “unit costs”, I was including the cost of research, as well as the cost of actually building the fab plant. These costs wildly outstrip the actual wafer costs, which is basically highly purified sand. Even if you strip out the cost of research, in the semiconductor industry, the company that can maintain the highest utilization ratio of its fab plants will end up with the lowest cost chip. These indirect costs are not neatly reflected in the income statements, but make no mistake, they define the cost structure of the semiconductor industry. This is why outsourcing chip manufacture is a possible strategy for AMD. If you cannot maintain close to 100% utilization of your fab plant, you are going to lose money hand over fist until you sell your fab plant.

I also agree that most retailers will continue to sell AMD chips, just to keep Intel in line. However, as long as the most advanced CPUs, the ones that has the highest profit margins, are completely dominated by Intel, than I think higher profit margins for Intel is not unreasonable.

Leave a Comment

Previous post:

Next post: