To: Sam Citron who wrote (63856 ) 5/16/2002 11:26:50 PM From: Jacob Snyder Read Replies (4) | Respond to of 70976 Magic numbers: 1. semi-equip capex/semi sales of 11% is expected LT average. Deviations much above and below that number foreshadow an impending ReturnToTheMean, and then a corrective overshoot in the other direction. 2. Capacity utilization of 100% at foundries, at the cycle top; below 50% at bottoms. Caveat: we could stay at 100% utilization for many months (but not years), and semi-equip stocks would keep going up. 3. valuation matters, as always. No indicator is perfect, closely following a short list of indicators is the way I do it. AMAT P/S 1-2 at the bottom; any P/S over 6 is unsustainable, a sell signal. Notice, we didn't get to that expected bottom valuation in 2001. My explanation of that, is the Bubble hasn't fully deflated. 4. My second largest position is NTAP. If the market was Efficient, NTAP wouldn't have hit 152 in 2000, or 6 in 2001, or 27 since then. The market is a herd of manic-depressives with no long-term memory, extremely inefficient. At any given time in the last 3 years, most stocks (especially tech stocks) were selling at many multiples of what they were really worth. At other times in history (1932, 1980), stocks were selling for a small fraction of their real worth. The market can't be Efficient, because: A) investors are not rational B) investors don't make decisions independant of other investors. Instead, they show herd behavior. C) investors don't all have access to the same information, and D) they don't get access to that info at the same time. Caveat: the above will only predict the general vicinity of the tops and bottoms of the cycle, and that's the best anyone can expect to do.