Abstract:The financial market is a complex system because it is susceptible to environmental changes.It is difficult to determine the real value of a financial asset simply based on the performance of that financial asset.Understanding the fundamentals of the financial market can help investors better understand the actual value of a financial asset.Optimizing financial assets in the market is a multi-attribute problem for investors to solve.Analytical hierarchy process is an effective method to solve this multi-attribute decision-making problem.Investors often have different interpretations of the influence of external factors on the actual value of financial assets.The varied interpretations are the result of investor's preference and intuition.Since the analytical hierarchy process framework is based on investor's sensations,the fuzzy analytical method can accurately describe the decision-making process for financial asset allocation.Multi-period asset allocation can provide a long-term view of optimal allocation dynamics.It is important to establish an optimal asset allocation program effectively.The essence of the optimal asset allocation model is to simulate all kinds of future market conditions and market statuses.The simulation results can help investors solve the optimal asset allocation scheme by using the financial asset yield performance under different market conditions.In recent years,researchers have focused on multi-period asset allocation models and scenario generation methods.However,all of these models and methods directly describe the market situation generated in the random environment,and extract random points.The investor tends to use language types to represent the future market situation,such asriseanddown.In other words,the investor exhibits the fuzzy uncertainty for future markets.Some scholars have improved the fuzzy judgment matrix in the fuzzy analytical process.However,the improvement of all methods relies on the expert's scoring system to construct the fuzzy judgment matrix.In the absence of experts,the fuzzy analytical hierarchy process has limited application to the asset allocation area.This research ranks the optimality of financial assets based on the relative deviation and the absolute deviation between variable factors.We also map the deviation value of the financial assets into fuzzy numbers which stand for language variables.Mapping can avoid decision maker's subjective judgment resulting using experts' scoring methods,and can expand applications for the fuzzy analytic hierarchy process.We establish a single-period fuzzy asset allocation program and introduce the scenario tree which is a common multi-period decision-making method in fuzzy situations.We calculate the Bayesian probability by setting the fuzzy interval values that represent future market situations.The multi-period fuzzy scenario tree asset allocation program is established in the fuzzy environment.First,we construct a fuzzy hierarchy which is required in the asset allocation decision-making process.The fuzzy assessment scales between the variables of the asset allocation are the triangular fuzzy numbers.The fuzzy comprehensive evaluation value can describe the subjective and qualitative factors in the investment decision-making process.The fuzzy hierarchy can represent the qualitative attribute more intuitively and effectively.The fuzzy analytical hierarchy process can combine fundamental analysis with the quantitative analysis.In order to avoid subjective scoring by experts when making a judgment matrix,we estimate the triangular fuzzy comprehensive evaluation value of the return indices and the risk indices by using the relative deviation and the absolute deviation between variables.Second,we establish the asset allocation model.The objective function of this model can calculate the triangle fuzzy returns with the constraint of triangle fuzzy risks.We also consider the constraint of short time limit,no external financing,and the upper limit of the investment proportion in the fuzzy single-period asset allocation program.Comparing optimal asset allocations,we find that the performance of the single period asset allocation is superior to the market portfolio.In addition,the investment return has higher sensitivity than the investment portfolio.The investment return has lower sensitivity for the risk constraint of lower limit from the sensitivity analysis.Finally,we expand the fuzzy single-period asset allocation program into the fuzzy multi-period asset allocation program.We further calculate the Bayesian probability between the market situations by using the fuzzy logic allocation.In addition,we estimate the coefficients of return between the market portfolio and the market segments based on the special market condition.The market segment returns are also calculated under the special market condition.We construct the multi-period asset allocation model in the fuzzy environment with scenarios trees and solve the optimal solution with the Bellman Optimal theorem.The performance of the fuzzy multi-period asset allocation is superior to the market portfolio.In conclusion,using the absolute deviation and the relative deviation between the factor variables can help map the discrepancy between the factor variables.The fuzzy judgment matrix is established in order to avoid the subjective bias of experts.The fuzzy analytical hierarchy process can be used to effectively dispose the asset allocation in the absence of the experts.Fuzzy asset allocation is effective in the fuzzy environment.Thus,the method can be used in solving the multi-period decision-making tasks.The market performance of the multi-period fuzzy scenario tree asset allocation program outperforms the market performance of the market portfolio.The multi-period fuzzy scenario tree asset allocation program can help determine the real value of a financial asset by incorporating the uncertainty factors of a financial market.

Empirical Study of Generating Scenarios in Multi-Period Financial Asset Allocation

Scenario Generation for Multistage Investment Decision and Its Empirical Research

Knowledge-Based Scenario Tree Generation Methods and Application in Multiperiod Portfolio Selection Problem

A Copula-Based Scenario Tree Generation Algorithm for Multiperiod Portfolio Selection Problems

The Multi-period Fuzzy Analytic Hierarchy Asset Allocation Based on the Scenarios Tree

Comparative Study of Scenario Generation Methods in Portfolio Optimization

Multi-period Portfolio Optimization under Possibility Measures

Multi-Stage International Portfolio Selection with Factor-Based Scenario Tree Generation

Vine copula‐based scenario tree generation approaches for portfolio optimization

Scenario Tree Generation Approaches Using K-means and LP Moment Matching Methods.

Scenario generation model for single stage R&D projects and securities portfolio optimization

Multi - Period Model for Assets Liability Management and Its Application

Using genetic algorithm to solve a new multi-period stochastic optimization model

Multi - Period Financial Asset Allocation Models Based on VaR

Multi-period Investment Decision Problem Based on Time Consistent Generalized Convex Risk Measure and Extremum Scenarios

A Novel Method for Multistage Scenario Generation Based on Cluster Analysis.

Scenario generation by selection from historical data

Bank Asset Liability Management Model Based on Multi-period Stochastic Programming

Evolutionary multi-stage financial scenario tree generation

PERFORMANCE COMPARISON OF SCENARIO-GENERATION METHODS APPLIED TO A STOCHASTIC OPTIMIZATION ASSET-LIABILITY MANAGEMENT MODEL

Personal Financial Planning Model Based on Multi-Stage Stochastic Optimization