Abstract
The strategy of simply holding stocks of high momentum, high trailing returns, is amazing for the amount of support that it has gotten from normally skeptical academics. But in practice there have been flies in the ointment. Unhedged pure momentum didn't help at all with the 2007--2008 Lehman Brothers/subprime crash and it would not have helped in 1929 either. Herein momentum is applied to a portfolio of stock funds or of stock-fund-like subportfolios, not stocks. And a simple resort-to-cash cure for momentum failures during panics is specified and tested using a new kind of alpha whose confidence interval is calculable despite the mixed-distribution character of the resultant portfolio returns. A walk-forward procedure is conducted that is more of a simulation than an abstract exercise in mathematical statistics, that provides a dynamically-optimized momentum strategy that would be adaptive to secular changes in the marketplace. That the optimal form of the momentum measure for a portfolio of funds or subportfolios is critically different from a popular form that works with stocks is demonstrated. And it is discovered that if volatility suppression is emphasized and there is a policy of resorting to cash when momentum falters then the optimal target number of funds or subportfolios to hold is a considerable fraction of the number of candidates.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have