Further to the above Walgreens is replacing GE next week in the Dow. WBA being down on the year 11% so far you might think would weigh on the index except for the fact GE is down 26% so far.
The SP500 has a small momentum factor (due to adding winners and dropping losers) component as well as the dominant market component. You can test this by regressing monthly sp500 returns against the momentum factor returns on Kenneth French's site. You'll get a small but highly significant coefficient. This is factor beta, not an additional "positive expectancy" over equity risk premium in general. Yes, stocks fall down into mid-cap and small-cap indices all the time. Interestingly, when GE (eventually) drops from the sp500 it will still be in broader indices. So if the "adding winners, dropping losers equals positive expectation" thesis were correct, you would expect the sp500 to out-perform those broader indices. This does not appear to be the case IRL.
Kevin, based on the argument earlier, I honestly think we are dealing with basic arithmetic here, not a misunderstanding of the momentum factor.
There is no reason to expect that the broader index would out-perform the senior index based on this one factor only, there are many factors that influence the holdings of stocks, and these potentially have far more influence. Nevertheless, the continual re-constitution of stock indices gives all of them a positive buoyancy factor which is absent in e.g. an individual stock or in a commodity price or in a forex value.
Look at like in golf, a better player gives strokes to a lesser player or maybe a better analogy is with horse racing, the better horse carries more weight. Indexes are allowed to get rid of that dead weight increasing their chances to "win".