PE=growth rate is a good rule of thumb, but the growth rate has to be based on a much longer time period than two years.
This is because the idea of stock valuation is that the stock is worth the future sum of it's earnings, where the n'th years earnings are discounted by (1+t)**n, where t is the T-Bill yield.
Looking at Intel over a 5 point time period (1999-2003), the earnings would be: $1.05, $1.51, $0.49, $0.68, $1.00.
I think if you did a linear regression on this it would have a negative slope.
Even going back to 1996 (8 years from 1996 to 2003), the EPS back then were $1.70. (adjusted 4:1) The growth in earnings is still negative, so PE should be very low.
So let's go for 11 years! In 1993, earnings were $5.20, but adjusted for splits, they were 0.65.
So, from 1993 to 2003, earnings growth is 50%. (A linear regression would be better than taking the endpoints, but I've spent enough time on this post already.) That works out to less than 5% growth per year.
Now do you understand why a PE of 50 for Intel is unreasonable? Especially considering that the last decade was a lot friendlier to the PC industry than the next decade is likely to be.
Petz |
