Many researchers propose investment strategies to aid investors in buying either growth or value stocks. The authors of this paper have developed two investment strategies--switching investment strategy (SIS) and mixed investment strategy (MIS)--using the theory of hidden Markov models (HMM).

Based on the forecasted risk, switching periods for investments between growth and value stocks are determined by the SIS. On the other hand, optimal proportions of investments between these two stocks are advocated by the MIS.

Erlwein et al. used filtering techniques to develop the SIS strategy. They proposed a 2D process using the return forecasts of Russell 3000 growth and value indices to develop the optimal strategy. The parameters of these processes are governed by a discrete-time Markov chain that switches between regimes (regime-switching parameters). Decisions, to invest in a growth index or a value index, are made using the signals from the HMM forecasts.

Using the optimal parameter values from the HMM filtering method, the optimal proportion for the MIS is calculated. The authors formulated the existing mean-variance theory as a function involving weights by using the Russell 3000 indices, and provided explicit expressions for the optimal weights that will maximize the function.

Existing similar strategies are compared, and the efficiency of the proposed HMM-based strategies is illustrated through tables and graphs.