I believe investors should adjust their portfolio compositions over time, as valuations become extended or, on the other side, exceedingly cheap. I certainly agree with taking a more defensive posture when the market starts looking a little frothy, but I have to agree with Ed Brown when it comes to remaining fully invested. The difficulty with market timing is not so much in identifying when the market is over or undervalued, by whatever measure being used, but in knowing when to pull the trigger. Markets can stay relatively over or undervalued for long stretches of time. Many studies have been conducted testing the viability of market timing strategies, using various measures of value and rules for shifting allocations. For the most part the conclusions of these studies suggest staying fully invested is the way to go. Generally, the relative gains from correctly reducing allocations prior to market declines, were outweighed by the underperformance resulting from being out of the market when it rallied. The market of the mid-late nineties provide a good example of this. The market, by my calculations, was beginning to look pretty pricey by mid-1996, and by mid-1997 was extremely overvalued in my view. Had I reduced my equity exposure at that time, it likely would have proven far more damaging than remaining fully invested during 2001-2002.
John's point is valid. If you spend some time on going through some postings you will see that I do not rely on my gut, exclusively to take a defensive posture in the accounts I manage. I rely on three indicators, two of which I write about all the time. When the yield curve inverts, it historically has meant that a recession is on the way which is bad for equities. The other indicator is when the SPX goes below its 200 DMA. I would suggest that John study the history of both of these indicators. I view them as clear messages from the market that a problem is on the way. As a good friend of mine says, the market can only do four things; up a lot, up a little, down a little, and down a lot. Any investor will do wonders for their lifetime returns by missing big chunks (not all) of down a lot. The history of both indicators I use give you a great chance to side step down a lot. But that's all it is, a chance.