Design Of FOF Combination Scheme Based On GBDT Algorithm

The FOF fund is called the fund in the fund.At the beginning of its establishment,as a highlight of financial innovation,its development prospects were favored by many investors.The two core links in constructing a FOF fund are fund optimization and asset allocation.A high-quality fund coupled with an excellent asset allocation plan can make the asset portfolio achieve good performance.In the empirical research of FOF portfolios,large-scale asset allocation is the focus of many scholars' research.Although the asset allocation plan is the core part of constructing a FOF portfolio,it lacks a screening plan for high-quality funds and still cannot form a good performance.FOF combination.In addition,there are some problems in the research of large-scale asset allocation schemes.First,the importance of the choice of the underlying asset for asset allocation is ignored.There are more than one well-known index among similar assets,and different indexes are used to build risk parity portfolios and combinations.Performance may vary widely.Secondly,considering the risk parity portfolio after asset diversification,although a better risk-reward ratio can be achieved,the portfolio return is not high.Finally,the selection of high-quality funds in the current FOF fund portfolio requires fund managers to do their best,consumes a lot of manpower and material resources,and the screening efficiency is low.Therefore,a quantitative investment framework that is in line with the public investment logic and easy to understand the model framework is required to initially screen the funds to improve screening efficiency.Based on some problems in the current scheme of constructing FOF funds,this paper proposes a practical solution: training the indicator data through the GBDT algorithm,and optimizing the parameter adjustment.It is concluded that when the models perform optimally,the indicators affect the return on assets.Contribution level,and calculate the comprehensive value of each fund as a screening index for high-quality funds.After selecting the optimal fund for each holding period,use the risk parity method,equal weight method,the most decentralized method,and the minimum variance method to configure the asset weights separately,and conduct a comprehensive comparative analysis of their returns and risks through performance indicators In order to select the optimal asset allocation plan;based on different investment needs,on the constructed plan,bonds are used to increase leverage and increase the risk budget to enhance portfolio returns.The results show that the combination of the GBDT algorithm and the risk parity strategy can better reflect the advantages of the risk parity method itself,and its overall performance is better than the other three investment strategies.In terms of performance,the new scheme's FOF portfolio returns is at the upper level in the market and the investment effect is relatively stable,which can well solve the above problems.At the same time,the income enhancement scheme adopted in this paper can also effectively enhance the portfolio income.Among them,bonds plus leverage is a cost-effective yield enhancement solution.While increasing yields,volatility has only increased slightly;increasing the risk budget can greatly increase returns,but also greatly increase the risk of the portfolio.More suitable for risk appetite investors.The empirical results show that the scheme in this paper,including the subsequent income enhancement scheme,can be effectively applied to the construction of FOF funds to meet different investment needs.At the same time,the use of machine learning to screen funds is a relatively new approach in this field,which provides a new idea for building FOF portfolios in the future.