So what about the 20% of time when the market confirmation indicator does not work? (I discussed it my last posting)

1) I do not think it is best to buy on the actual market confirm day. This is because there can still be some selling after the confirm day where the market will still go down about 2 percent or so. So what I usually do, is buy over time in the week or two following.

2) Look for distribution days in the week following the confirmation. If the distribution days start to stack up, that’s very good signal that the rally is faltering. This happened a number of times in the bear market that started in 2000. Often, if I see more than 1 distribution day in the week or so following the confirm, I will either sell or not buy at all.

3) Look at the overall market direction. Has the market recently been experiencing a horribly bearish episode where it has dropped 20% or more? or was it only a correction of 15% or less? The more bearish the market has been, the more likely it will take more than one of these market confirmation signals to change its overall direction.

4) Finally Look at the earnings of the S&P 500. If the overall earnings picture of the market is poor, meaning year over year quarterly earnings are flat to down, then the rally probably won't last.

In my next posting, I will discuss some interesting publications and web sites that have helped me make good decisions.

Do you think I am full of it? I would really like to know. I have often made my best decisions when someone was kind enough to show me a better way.