In a way, momentum investors are riding the coattails of the investors who dig into financials.

I think the "disposition effect" has some good explanatory power for why momentum works as well as it does, and continues to do so.

You buy something for $100 a share. Market price drops to around $80 for a long time. You're depressed. Good news comes out, meaning the stock is worth a lot more than it previously seemed. Let's say for the sake of argument that the new true value is $130, but of course there is no billboard that says that.

The market price starts rising on the news, and you sell not far above $100 because you anchored on what you paid and are suddenly in a profit after a long time with red ink, and "nobody ever went broke making a profit". This sale (yours and those of many others) has a short term dampening effect on the price so it doesn't rise to its true new fair value of around $130 until time and aggregate trading volume have swamped all those sales by people who were waiting to break even.

So, from the date of the good news till the date that the last of the anchoring sellers has sold, the price is on an upwards trend and momentum works.

This is definitely a real effect, provided you think there exist at least some people who are hesitant to sell at a loss but more willing to sell for a small gain, which seems very likely. We all do it, I suspect.

To the extent that this is a real effect, it makes predictions. For example, the number of such people is a function of both daily turnover and the length of time that the stock was lower than the current one. In essence, the turnover-weighted integral of the past prices is an estimate of the weighted average trader's (not long term holders') breakeven. The current market price is repelled by that average breakeven number, which shows up as momentum working.

Jim