The traditional regime-based asset allocation model originated from Merrill Lynch in 2004. These categories of model explore the relationship between macroeconomic variables and asset returns. they point out that there are categories of assets with relatively excellent performance in different regimes of the economic cycle, so an investor should allocate different categories of assets in different economic regime. However, due to special political and economic system of China, regime-based allocation model in China often has confused results. To study how to allocate assets when economic cycle changes, this paper studies how to optimize the method of regime-based asset allocation model to improve the effectiveness and practical value of its application in China's capital market.
Methodologically, firstly, this paper takes indicators of economic growth and inflation rate as standards to divide economic regimes, such as GDP/PMI/CPI. Secondly, considering the actual situation of China, this paper chooses 3% as the threshold for CPI inflation, 6% and 50 as the threshold for GDP and PMI growth rates, respectively, to divide the economy into four regimes. Thirdly, this paper utilizes the corresponding tactical asset allocation theory to study the expected return rates of stocks, bonds, commodity and cash under different economic regimes, and compares the level of yield.
Results illustrate that taking the comprehensive PMI index and CPI index as the two- dimensional indicators of economic cycle are better than those of taking the growth rate of industrial added value or GDP as the dimensional indicators of economic cycle, both in terms of the degree of matching with the theory of asset allocation model and growth rate of the total net value of the optimal investment strategy. This paper also compares the optimal asset in different economic regimes in China and US, the results shows that cash and bond return are the main reason of differences. The main underlying reason is that China is an emerging market while the United States is a mature market, and the economic cycle and interest rate of emerging markets like China is relatively weak.
Finally, considering the defects of traditional regime-based asset allocation model, especially the obvious shortage of China's economic data statistics, the original model can only represent the current economic state with lagging historical data. Therefore, this method is difficult to predict the changes of the economic regime. Therefore, this paper utilizes a simple M-logit model to predict the future economic regime through historical economic data, and then carries out asset allocation based on the forecasted economic state. The research results show that the asset allocation results combined with HP filter and M-logit forecasting model have higher rate of return.
