RANDY responds:

Although David's case for Dell being overvalued is a very apt summation of the conventional wisdom surrounding the company, this particular Fool thinks there are some assumptions being made that are not necessarily justified given the actual facts.

David discusses the possibility of margin erosion at Dell due to an industry price war or similar event, hinting that margins could roll back to the five-year average. Given that Dell's 1996 move into manufacturing servers is the principal reason for the rise in margins, it is a mistake to assume that this margin rise has been the result of stable pricing that could be undone at any turn. Rather, as Dell moves servers from only 6% of total revenues to somewhere in the neighborhood of 20%, margins could actually continue to increase.

Should some sort of economic slowdown occur, David believes that computer hardware will be the first place to cut. Given the rapidly falling price tags for these products and the increasing productivity gains that companies get from Pentium-class CPUs, I find this somewhat hard to believe. More memory will not allow computer workstations to run graphics-intensive software applications for desktop publishing, CAD/CAM, Internet publishing, or complex analytical tools -- only faster CPUs will do this. Recently, I used a x486 laptop with only America Online, Excel, and one Internet connection open and the CPU was running at somewhere around 95% to 100% of capacity the whole time. As Microsoft builds bigger applications, anyone who has Word and Excel open at the same time needs a Pentium-class chip.

David finally states that 50% growth for a decade is unheard of, assigning Dell an arbitrary annual growth rate of 30% over the next five years. This is an assumption that will probably not prove to be correct. As Compaq has $20 billion in PC-related revenues right now, it is easy to imagine Dell growing to the size of Compaq over the next five years, overtaking IBM as the number two vendor of PCs and servers and stealing a lot of market share from Compaq, IBM and smaller, privately held players. With only 150 million PCs in the U.S., if everyone who owns a car eventually owns a PC, there are still 100 million PCs to be sold --and that assumes that no one replaces the 80 million non-Pentium machines out there with Pentiums. As I believe these older machines will be replaced, I would wager that more PCs are sold in the U.S. in the next five years than currently exist, giving these companies 36 million PCs a year and 3.6 million servers (one for every ten PCs) to fight over in North America alone.

If you assume Dell can grow revenues at 2 times the overall industry rate over the next five years and grow earnings at 1.2 times this due to margin improvement, economies of scale, and share repurchases, this would put the growth more in the 40% to 50% per year range, which would give us earnings of $9.30 per share. Should Dell trade at 30 times this number, we would have a $279 share price in five years, or 27% annualized growth.

--Randy Befumo ([email protected])