Abstract: Classical green credit researches always have the limitations of simple models, non-dynamic parameters and only obtaining Nash equilibrium points. To address the above issues, a data-driven multi-objective optimization algorithm is studied, where governments can promote banks to implement green credit. The optimal strategy problem of green credit is first illustrated based on a data-driven multi-objective optimization algorithm framework, and then the optimal strategy Markov state transition model is established based using historical data. Finally a multi-objective particle swarm optimization algorithm is introduced to solve the optimal long-term total income strategy of the government and the bank. Unlike the traditional methods by which the approximate models and game theory are used, the proposed method can obtain experience from historical data, thereby formulating a strategy with more long-term benefits and avoiding the “short insight” phenomenon. The analysis results show that the benefits of green credit will not appear in a short period of time, which makes the government consider the return period of green credit benefits to make long-term judgments when making decisions.

Key words: green credit, data driven, multi-objective optimization, government, bank

摘要： 传统的绿色信贷研究中存在着模型简单、非动态参数以及只能获取纳什均衡点的局限性。为改善这些局限性,研究了一种基于数据驱动多目标优化算法的政府促进银行实施绿色信贷的策略计算方法。首先针对绿色信贷的最优策略求解问题建立数据驱动的多目标优化算法框架,再基于历史数据建立算法框架中的最优策略马可夫状态转移模型,最后使用多目标粒子群优化算法对政府和银行的长远总收益进行最优策略求解。与传统的基于近似模型及博弈论的方法不同,本文提出的方法可以获得历史数据中的经验,从而制定出具有更加长远收益的策略,避免了传统方法中的“短视”现象。分析结果表明,绿色信贷的收益不会在短时间内显现,因此政府在做决策时,必须根据绿色信贷收益的回报周期作出长远的判断。

关键词: 绿色信贷, 数据驱动, 多目标优化, 政府, 银行