Research On Asset Dynamic Allocation In Global Market

Since 2016,some events such as Britain exiting from the EU,US presidential election,and the inclusion of RMB in the SDR have all caused different effects on the global economy.The accelerating process of financial and economic globalization has made the links between countries increasingly closer.The financial markets of all countries in the world become more open,creating more investment opportunities for international investors.For international investors,in order to effectively avoid risks and improve portfolio performance,globalization of asset allocation has become a trend.More and more investors are beginning to pay attention to the globalization and diversification of asset allocation.In order to diversify the asset allocation and achieve the purpose of diversifying risks,global asset allocation not only requires the allocation of assets in different countries,but also requires the deployment of multi-category high-dimensional assets.In the process of global asset allocation,the correlation between assets is a key factor,because the correlation between asset returns determines the characteristics of the risk characteristics of the asset portfolio to a certain extent,which in turn determines the performance of the asset portfolio.However,the actual financial market is an ever-changing environment.Therefore,accurately measuring the dynamic characteristics of asset correlation during the global asset allocation process is an important factor affecting the investment portfolio decision.In the actual investment process,the high-dimensional investment portfolio has a strong practical significance,but due to the limitations of existing computing capabilities and algorithms,there are still certain challenges in the modeling of high-dimensional assets.To effectively solving the difficulty in the modeling of high-dimensional asset and the high cost of calculation not only have theoretical significance,but also have a guiding role in real investment decisions.This paper takes the stock indexes,ten-year treasury bonds and foreign exchange data of major markets in the world from January 1,2006 to June 30,2017 as samples,starting from the dynamic characteristics of the correlation between high-dimensional assets with DECO model,this paper find that all kinds of assets show positive equicorrelations,and the equicorrelations between assets changes with time,that is,the correlation between assets has a time-varying nature.Specifically,the level of equicorrelations among stock markets is the highest,and that of bond assets is the lowest,the equicorrelations of foreign exchange assets is more volatile than stock assets and bonds.At the same time,this paper also finds that the average correlation decreases with the number of asset classes increases,which indicates that the risk of portfolio can be effectively reduced through asset diversification.In addition,this article also examines the impact of several well-known economic and financial crises on the average correlation between assets and finds that,except for the correlation between bond assets,which shows the opposite trend,the average correlation level shows a clear upward trend during the crisis and remains high during the entire crisis period.And as the impact of the crisis dissipates,its level of correlation gradually declines to a lower level.Subsequently,this paper conducts global asset allocation based on dynamic correlation analysis of high-dimensional assets.Specifically,this paper adopts a"rolling window" to build a portfolio.The DECO model and the further developed Block-DECO model are used to determine the investment weight of the portfolio and compare it with the equal weight portfolio,and find that the DECO portfolio and Block-DECO portfolio performance is superior to equal weight portfolios in most cases.In addition,the performance of the Block-DECO portfolio is always better than DECO portfolio.This is reflected in the higher average yield,lower standard deviation,and higher Sharpe ratio,higher Treynor ratio,higher Jensen ratio and higher information ratio.Therefore,we believe that the investment portfolio constructed based on dynamic correlation analysis of high-dimensional assets has strong applicability and effectiveness.What's more,the more detailed division of high-dimensional assets,the more applicability and effectiveness of that.The main contribution of this paper is to focus the research perspective on multiple developed markets and multiple emerging markets at the same time.At the same time,this paper expanding the scope of assets from stocks and bonds to foreign exchange assets,considering these three types of assets together,taking into account the common risks of these assets and considering the mutual influence among these assets,it is more in line with the actual globalization.The investment process is also more conducive to the construction of a diversified portfolio.This paper uses DECO model to analyze the correlation between financial assets and solve the problem of high-dimensional asset estimation.DECO model can not only grasp the dynamic characteristics of asset correlation,but also improve the accuracy of high-dimensional series estimation.This article provides systematic explanations and recommendations for high-dimensional asset allocation in a dynamic environment.