To: Bob Davis who wrote (75509 ) 10/29/1998 12:11:00 PM From: Chuzzlewit Read Replies (1) | Respond to of 176387
Bob, thanks for your response. Since my criticisms of your methodology seem to span the gamut of valuation issues, I suggest we take them one at a time. So I will start with what I consider one vanilla issue, and a second more difficult issue. First, the stream of projected earnings is a vanilla issue (IMO) because you have assumed the consensus analysts' projections are shared be the stockholding community in general. Clearly, this is false, and I will marshal three pieces of evidence in my favor: 1. Over the last several years, earnings "surprises" have exceeded analysts expectations by a very healthy margin. I believe that the lower expectations are the result of companies trying to manage investors' expectations. 2. It is not unusual for a company to meet expectations and the price of the stock to drop precipitously on the announcement. 3. The emergence of whisper numbers also points to the belief on the part of investors that growth may be significantly different, and generally higher than analysts' published estimates. I contend that all of these pieces of evidence are indicative of the fact that investors do not accept analysts' estimates blindly, and that those investor estimates are significantly higher than analysts' estimates for Dell. This issue alone is sufficient to render a DCF approach to valuation problematic at best. And now for the not so trivial issue: risk and market leadership. I believe that maintaining a commanding position in an industry provides a company with a significant advantage over its competitors, so the differential risk premiums inevitably lead to what some observers contend is overpricing. I don't know how one could quantify this effect, but I am certain it is real. To take an obvious example, I am sure that the market's risk estimate of Micron is considerably higher than IBM. What we are really talking about is the risk not only of disappointing investors (which might be measured as the standard deviation of the EBITDA estimates) but also the bankruptcy risk. These considerations can lead to wildly different valuations depending on the assumptions that individual investors make. ************** Because of what I believe are the inevitable shortcomings of both traditional (balance sheet based) and DCF (cash flow based) methodologies, I have developed a heuristic which makes sense to me. I use what I have called CNPEG (Chuzzlewit's Normalized PEG). Using this approach, I take the forward PEG of the stock in question and divide it by the PEG of either the industry group or the PEG of the S&P500. This sidesteps the issue of valuation, and instead focuses on relative valuation. I don't know how useful this method will ultimately be. It might need to be adjusted by using relative betas, and I am toying with a number of refinements. It might also need to be adjusted by removing companies from the index (S&P or industry) which are either near break-even or are losing money, because those kinds of businesses can significantly distort the P/E of the group. I look forward to your response, and perhaps we can broaden the discussion. TTFN, CTC