Yes, that would be a more reasonable approach. Of course, the trouble there is determining where to set the trigger point for action. 1 s.d.? 2 s.d.? Set it too close, and you've foregone some real gains. Why don't you try some data mining to give us a good point--I'm sure that would be a good headache-reducing activity. <g>
Seriously, though, I think this is where the "art" comes in, at least for individual investors. My sales that are based on extended valuation have been based solely on feel.
If I understand you correctly, then, you're not disagreeing with akoni's basic strategy, just saying that he should use relative valuation (against some
company-specific historical average, perhaps) rather than absolute valuation as the variable. You could set the upper and lower bands at, say, two
standard deviations above and below the average, and achieve something like the originally intended outcome, no? |
