Different than...what?
I think we need to look at as many characteristics of past cycles as we can discern, with a view towards discovering which are consistent, and which are not. I tend to have a bias towards fundamentals, that is, what's going on in the underlying business as opposed to what's going on with the stock market, because I view the former as cause, and the latter as effect. I don't see how the sector can start a sustained uptrend in stock prices, leading eventually to new all-time highs, until we see an end to the parade of bad news about the economy, end-user demand, semiconductor sales, and equipment bookings.
Adding to the difficulty of our task is the fact that every cycle has had some things that were different, and some that were the same. For example, on the previous two cycles, the final bottom in sector stock prices coincided well with the bottom in equipment bookings. But the depth of the decline in bookings was different, being about 50% in one case and 70% in the other. Another example is the cyclicality itself. That there are cycles in the industry has been a dependable fact of life since the beginning. But what has always been different is that the number of years for each cycle has always varied. So we have to try to use our intelligence to figure it out, distinguishing carefully between those features of past cycles which are consistent, and those which are not, and giving preference to correlations for which we can find logical reasons to believe that there is a causal connection, as opposed to mere coincidence.
I would say, no, this time is no different and the evidence is out there.
Then what did you mean, when you wrote the following:
People will claim, hubris ad nauseum, that x, y, and z will be present at the bottom of the equipment cycle. If these factors don't fall into place roughly approximating these preconceived ideas of a cycle bottom, then the bottom 'by definition' has not occurred. If such were the case, and definitions of cycles were so easy, we REALLY would all be millionaires.
You certainly seemed to suggesting, in the above, that it is not useful to take past patterns into account. Or is it only those past patterns that don't support your conclusion?
If you wait for the fundamentals to come around you'll miss half the run.
History does not support that conclusion:
geocities.com
Looking at that chart, suppose you had taken the example of the 1996 trough, and used that information to try to figure out how to pick an entry point in 1998. Two approaches that could have been used are (a) wait for bookings to fall by 50% from their peak, or (b) wait for two consecutive months of increasing bookings. One method would have gotten you in two months early, and the other would have gotten you in two months late. Both methods would have given you the benefit of 90% of the subsequent runup in stock prices.
My consistent point has been that the participants of this thread have been no more disciplined at going against the grain at any stage than either the analysts or Joe Blow.
I agree that sentiment in the thread is not the best indicator, but in other posts you have suggested that it is a contrary indicator. I'm suggesting that looking at the history of the sector and trying to discern the causal factors is much more likely to be successful than (a) sentiment taken by itself, especially with a too-small and decidedly non-random statistical sample, (b) what the herd is doing to stock prices, which is effect, not cause, or (c) analyst trends, which is pretty much useless noise in my book. |
