Bob, thank for an interesting glimpse at valuation models applied to growth stocks. What I found missing in these documents is really essential if you are going to use DCF methodology to appraise a company. That is, the concept of risk. Actually, the market embodies this pretty well by rewarding market leaders like DELL and CSCO and healthy premium over their smaller brethren. In essence, the market uses a lower risk adjusted discount rate because of the competitive position that these companies have achieved.
I have great difficulty accepting Graham's heuristics because his numbers are arbitrary. As you correctly point out, the DCF analysis depends on the prevailing interest rates, and this makes for fluid valuations.
The real problem I have with these approaches is that they ought to be modeling behavior, not determining what rational behavior ought to be. Put another way, a good multiperiod valuation methodology should be generating numbers that comply fairly well with the market. If they don't, then my inclination is to question the model (not the market). So concluding that DELL, or any other issue, is "overvalued" is normative based on the assumption that the model you choose is correct. I think a more rational approach is to assume that there is a problem with the models since they do not reflect market-determined values. And as any economist will tell you, the value of anything is what is agreed to by a willing, informed buyer, and a willing, informed seller. And aren't those the people who trade stocks?
TTFN,
CTC |
