Detecting market tops and bottoms

[Tom K]

Painful lessons

Thought I'd add my two-cents to your post.

I think I agree with your first point, but might have a slightly different interpretation. Did you mean 'never follow a single indicator'? If so, I agree. I'm an advocate of timing models - using several indicators to develop a market outlook. My NYSE intermediate model is here:

http://stockcharts.com/def/servlet/Favorites.CServlet?obj=ID350373

"The trend is your friend". True, but not at market extremes. You'll note that my model includes trend/breadth indicators AND contra-trend indicators. The idea is this: Follow the trend until a critical number of reliable reversal indicators tell you differently.

"Be neither a bull or a bear but a trader and be prepared to go either way". Amen, I totally agree.

"Follow the moving averages...." My advice is this: Develop a model that's ACTIONABLE. I get chuckle whenever I go through the Stockmarket.com public list. I see dozens of pages with trendlines, chart patterns, Elliott waves, oscillators, etc. Most of these pages are history lessons, but don't tell me a damn thing about what I should be doing tommorrow.

Do their indicators mean they're bullish? bearish? a little bearish? a little bullish? Does it mean they're 20% short or 100% long? If indicators aren't quantifiable and ACTIONABLE, don't even waste your time.