I second corpraiders sentiment. I believe that if you back test an equal weighting to all of the sector spiders ETFs especially prior year 2008/2009 and rebalance them annually you will find they outperform the total market funds. It's not so much that there is buying/selling going on as much as there is constantly over representation of the sectors that are in favor, and when they go out of favor they cause a greater drop in the market. I no longer have an etfreplay account but in sure someone else might be able to check this out
Fundamental or equal weighted indexes will generally outperform cap weighted indices before costs for a simple reason. They Accept more risk in the form of a higher exposures to the the value and size factors (stock exposure will be on average cheaper and smaller in such indices.)
Funds tracking such indices will also always be more expensive than their cap weighted counterparts for a simple reason: they require more transactions to rebalance the fund.
Imagine the S&P 500. If GM represents 2% of the S&P 500 on January 1, then a capitalization weighted fund will own exactly 2% of GM.
On the other hand an equally weighted index will hold 0.2% of GM (1/500 of the index)
Now if GM grows to become 3% of the S&P 500 this growth will be reflected in the stock price of GM automatically (and in the stock prices of other companies in the index that have shrunk relative to GM.) The Cspitalization weighted index need do nothing, it is already automatically rebalanced by price movement.
The equal weighted index on the other hand must sell some GM stock and buy some of the other stocks that have become relatively less valuable in order to maintain equal weighting. This costs money in the form of transaction costs, capital gains, bid/ask spreads, and impact costs.
In other words capitalization waiting is always cheaper, and fundamental weighting or equal weighting are simply quantitative ways to pursue value or size exposures, much like adding small value/funds to your portfolio.
No free ride. And also note that cap weighting has zero exposure to the momentum factor (or the value or size factors) as suggested in the original post.