A dynamic asset-allocation model is specified in probabilistic terms as a combination of return distributions resulting from multiple pairs of dynamic models and portfolio strategies based on momentum patterns in US industry returns. The nonlinear state space representation of the model allows...

A dynamic asset-allocation model is specified in probabilistic terms as a combination of return distributions resulting from multiple pairs of dynamic models and portfolio strategies based on momentum patterns in US industry returns. The nonlinear state space representation of the model allows...

We study the filter in Ma and Tang (2012) and show that the filter has predictive power for sunspot cycles. The filter is a 12-month simple moving average of the sum of one month, three month and six month differences of the logarithm monthly sunspot number series. We show that this filter leads...