What if we could find the top 10 stocks that have a high correlation.
1. Invest in them for one month.
2. Then again, find the top 10 stocks that have a Hi-C. Some socks will remain, some will be replaced.
3. Rebalance the portfolio and repeat steps 1 & 2...etc, etc.
Not exactly using high correlation, but VectorVest suggests a similar strategy of buying the 10 stocks with the highest VV rating, then replacing those next month that have fallen out of the list from the prior month. I suppose you could do the same with Value Line or any similar service that allows ranking of some sort or another. Would need to test how such a thing would work in the 'real world - in other words as something falls out of the top 1o are you still making a profit on it or are you losing money?
Best,
AIMster