To: Mike Buckley who wrote (7106 ) 11/6/1999 7:09:00 PM From: Chuzzlewit Read Replies (1) | Respond to of 9068
Mike, Thanks for your comments. I think we are much closer in our thinking than you may suspect.Regarding your comments about value, I also agree with you but only to a point. For those who are as entrenched in using a PEG ratio as I am, one reason it is a valid measure worth adding to an arsenal of other valid measures is that it inherently has strong ties to a discounted cash flow of future earnings. An expert on the subject required a lengthy article to explain that so I won't try to go into it here. Actually, I am quite expert in the use of DCF (he said modestly <VBG>). I have modified the standard PEG approach. The major difficulty with PEG is that it is insensitive to discount rates and risk premia. Note that an increase in interest rates will may or may not affect cash flow or growth, but will have a profound influence on the value of future cash flows. There is no a priori reason to assume that a stock is properly valued when its PEG is at 1.00 -- that will depend entirely on two important factors: the inherent riskiness of the stock and the prevailing risk-free interest rates. So, in order to adjust for these parameters I calculate a YPEG and then divide through by the YPEG for the S&P500 (which you can obtain from Yahoo). This normalizes the YPEG for the market as a whole, but it inherently assumes a stock with average market risk (beta of 1.00 by definition), so I multiply the resulting number by the beta of the stock. I call this measure CNPEG2 (for Chuzzlewit's Normalized PEG). Now I use the heuristic of a CNPEG < 1.00 as a buy point. In spite of what I believe to be an improvement over the standard PEG metric, CNPEG2 still suffers from two problems. First, it relies on relatively short growth horizons (5 years), and second, the PEG for the market as a whole tends to be over-stated because of the presence of several companies in the S&P which have little or no earnings, and are valued on their inferred intrinsic value. TTFN, CTC