我国银行业信用风险的宏观压力测试实证研究
商业银行信用风险管理商业银行信用风险管理
作为压力测试方法在宏观经济分析中的运用,宏观压力测试应运而生。随着各国金融监管当局对金融风险的日益重视,宏观压力测试逐渐成为检验一国金融体系脆弱性的首选工具。本文的研究目的就是为了探索我国银行体系信贷违约率与宏观经济因素之间的关系,并在此基础上运用宏观压力测试评估我国银行体系的抗风险能力。
本文在借鉴和分析国外成熟模型的基础上,构建了适合我国经济环境的宏观压力测试模型。首先使用Logit变换和一阶差分将不良贷款比率转化为“不良贷款指标”〖∆y〗_t,以指标〖∆y〗_t作为因变量与宏观经济因素进行多元线性回归分析,尽可能使这一指标能够充分反映各宏观经济指标所提供的信息。在宏观经济变量的选择方面,参考国内外学者实证研究,结合我国宏观经济环境,并逐个对各个变量进行单变量回归分析,最终,选取了狭义货币M1增速和房地产开发投资增速作为宏观经济变量来构建模型。本文的一个创新之处是对官方公布的银行业不良贷款比率进行了修正。我国分别在2005年和2008年进行了两次大规模的不良资产剥离,使商业银行的不良贷款比率大幅下降,但这样的数据有人为操纵之嫌。因此本文将被剥离的不良资产重新计入不良资产余额,并重新计算不良贷款比率,以使模型的回归结果更加贴近现实。
本文研究发现货币M1增速与银行不良贷款比率呈负相关关系,房地产投资增速与不良贷款比率正相关,即:货币M1增长越慢,房地产投资增长越快,均会导致银行不良贷款比率的上升。并在此基础上构建了三种极端恶劣的宏观经济环境,模拟了我国银行业在各种压力情境下的损益分布。
本文最后提出了相关政策建议。
关键词:不良贷款比率;宏观压力测试;蒙特卡罗模拟
[英文摘要]: As stress-testing in the use of macroeconomic analysis, macroeconomic stress tests came into being. Since the authorities take importance to the financial risk, macroeconomic stress-testing is becoming the first choice to test the vulnerability of a country's financial system. The purpose of this paper is to explore the relationship between the credit default rates of Chinese banking system and macroeconomic factors, and on the basis of previous research, carrying stress tests to assess the risk resistance capacity of Chinese banking system.
After a review of relative study and mature model constructed for macroeconomic stress-testing, this paper has designed a model suitable for China. First, making Chinese non-performing loans as default rates, and calculate default rates index y_t as the inverse of the Logit function, and then establish the annual differences 〖∆y〗_t which called “default index”, it is determined by the macroeconomic factors under consideration of multiple linear regression analysis. With a full consideration of empirical studies refer to China and foreign research, this paper test each variable one by one using univariate estimations, finally, selects money M1’s growth rate and real estate investment’s growth rate as the macroeconomic variables to build the model. One innovation of this paper is to correct the bad loans rate published by official. China has conducted two large scale non-performing assets stripping in 2005 and 2008 respectively, make the default rates of banking system fell sharply, but this data should be suspected. This paper recalculate the non-performing assets included the stripped assets and compute the modified default rates, hopping the regression results of the model would be more realistic.
This study found that M1’s growth rate shows a negative correlation with bank default rates, and real estate investment’s growth rate has a positive relation with default rates, that means: the slower growth in M1, the faster growth in real estate investment, will lead to default rates increase. And on this basis, three macroeconomic extreme environments are set to simulate the Chinese banking sector’s profit and loss distribution under the crisis scenario.
In the last, this paper puts forward relevant policy recommendations.
Keywords: Default rates; Macroeconomic Stress-testing; Monte Carlo Simulation
[参考文献]: [1] Wilson T C. Portfolio Credit Risk[J]. FRBNY Economic Policy Review, 1998, (10): 71-82.
[2] Boss T. A Macroeconomic Credit Risk Model for Stress Testing the Austrian Credit Portfolio[J]. Financial Stabiliy Report. ONB, 2002, (4): 64-82.
[3] Virolainen K. Macro stress-testing with a maroeconomic credit risk model for Finland[J]. Bank of Finland. Mimeo, 2004, 19(1): 1-66.
[4] 毛晓威, 巴曙松. 巴塞尔委员会资本协议的演变与国际银行业风险管理的新进展[J]. 国际金融研究, 2001, (4): 45-51.
[5] 李雪莲, 李永辉. 巴塞尔协议对银行经营行为的影响分析[J]. 财经研究