Drip Portfolio Report
Friday, August 22, 1997
by Randy Befumo (TMF Templr)

ALEXANDRIA, VA (Aug. 22, 1997) -- There comes a time in every young Drip Portfolio manager's life when he must defend his first purchase. Before I continue to trace in careful detail the innards of Owens Corning Corp. as I try to determine whether or not this is an investment opportunity or a great chance to underperform the market, today I wanted to respond to two major objections to our recent selection of INTEL CORP. (Nasdaq: INTC) -- obsolescence and capital intensity.

Intel requires large semiconductor wafer fabrication plants to make those central processing units (CPUs) that power your PC. If you close your eyes for a moment, you can just see them. Whether or not you are in Santa Clara, California or County Kildare, Ireland, you can see the large sprawling, airtight facilities where men and women in bunny suits turn pure silicon into transistorized gold. As all the world knows, Intel turns on the CPU.

Remember for just a moment, however, that Intel has adopted the CPU as its sword, but in the competitive battles it has also acquired a number of other implements of war. Should its Pentium blade fail, the company has a shield composed of motherboards to defend itself. If it needs an alternate weapon, it can go for its network interface card morningstar. Should the enemy turn to flee, a few graphics chipsets in the back fired from a multimedia bow should finish them off.

As a company that views silicon as its birthright, whether the silicon is in a CPU or the motherboard of an network computer (NC) is irrelevant. Since the PC remains an incredibly profitable, fast-growing market, there is no reason to chain the fences in front of the fabs, lock out the workers, and stop making the gear. Should the market turn and some other information appliance (such as the NC) take precedence several years out, do you really think that Intel will not be able to adjust and compete?

With $8 billion in cash at its disposal -- more than Cyrix owner National Semiconductor or Advanced Micro Devices have made in combined revenues over the last four quarters -- the company seems to have much more flexibility than the facile competitive analyses that tend to surround the company would assume. Even if the NC were to become the standard, would these chips not be placed on a motherboard? Would the machine not be hooked into a network via an Ethernet card? Would there be no more silicon left in the machine that Intel could contribute? Would this NC perhaps be made of wood? Or ivory? Or some other precious substance that has the same, curious semiconductive properties that silicon enjoys?

As far as capital intensity, yes, Intel is rife with wafer "fabs" that require sophisticated million dollar machines to run smoothly. However, as you may discover in my Fool on the Hill column in the Evening News this evening, the relative consumption of capital is more a matter of opinion than fact. To put it simply, if you compare all the cash Intel takes in with the amount it spends on plants, land, and equipment to make chips, the capital spending runs at about 30% of gross cash flow. Coca-Cola, by comparison, runs at roughly 20%. Certainly less by a factor of one-half, however, to call one capital intensive and one "light" and fancy-free because of a gulf of only 10% is a bit exaggerated, in my opinion -- particularly given that Intel garners profit margins roughly 10% higher than those enjoyed by Coca-Cola, evening the score a bit.

When we consider the close to twenty multi-billion dollar fabs that Intel operates throughout the world, we realize that this $3 billion per year capital expenditure habit is one that few other companies could even contemplate. The very size of the necessary investments completely locks out the vast majority of the competition. Certainly in developing markets, one player can upset the balance and take the lead at any time. However, in established, stable markets where one player has billions of dollars in capital to deploy and the nearest competitors are either not directly in the business or but a fraction of the size, it takes sustained years of stupidity to lose a large lead in market share.

Sure, Novell managed to do so and there was an able competitor of roughly the same size at the time ready to take charge. Until Intel makes four or five stupid mistakes in a row, I think we may be alright in that department, no matter how many people believe that two guys in a garage can supply enough ivory-wood CPUs for the 150 million PCs that will be sold from now to the year 2000. Sure, and if you believe that, you'll also believe that our next investment will be in some old metal bridge in Brooklyn or in the mail company that keeps messing up the address on my magazine deliveries!

Have a Foolish weekend.

           (Please see the left navigation bar for information on this new portfolio.)