Credit Risk Assessment of Farmer Loans: Based on CFPS2018 Data

doi:10.11924/j.issn.1000-6850.casb2020-0572

Abstract

Abstract:

The aim of the study is to improve the accuracy of credit risk measurement of farmer loans, reduce the non-performing rate of banks’ agricultural loans and promote the bank's credit coverage to farmers. We selected default indicators from the four aspects of householder characteristics, assets and liabilities, income and expenditure, and willingness to repay, and used a combination of factor analysis and Logistic regression to build the credit risk assessment model for farmer loans. The empirical analysis took CFPS2018 data as a sample. The asset status of farmers was the most important factor affecting loan default, and the asset-liability ratio was positively correlated with the credit risk of loans. The more households spend on consumption, the more likely their loans will default. Recognition of trust is an important factor, the higher the trust of farmers in cooperation with others, the lower the risk of default. Moreover, the prediction accuracy of the constructed model exceeded 90%, which was universally applicable. We encourage qualified rural small and medium-sized banks to implement internal credit risk rating to improve the credit coverage of farmers. At the same time, banks should strengthen related supporting constructions, such as optimizing the risk organization structure, improving the risk management system, and building the risk control talent team.

Key words: farmer loans, credit risk, probability of default, Logistic model, factor analysis

CLC Number:

S-9
F832.43

Liang Weisen, Fang Wei. Credit Risk Assessment of Farmer Loans: Based on CFPS2018 Data[J]. Chinese Agricultural Science Bulletin, 2021, 37(25): 157-164.

Figures/Tables 5

References 26

[1]	祝树民. 以改革推动农村中小银行公司治理建设[J]. 中国金融, 2020(16):9-11.
[2]	马文勤, 孔荣, 杨秀珍. 农户小额信贷信用风险评估模型构建[J]. 财会月刊, 2009(36):49-51.
[3]	Omobowale O, Oladele O I, Oyewole I K. Analysis of Factors Influencing Loan Default Among Poultry Farmers in Ogun State Nigeria[J]. Journal of Central European Agriculture, 2005,6(4):619-624.
[4]	Abunyuwah I, Blay J K. Accessibility Constraints of Small-Scale Fish Farmers to Formal Credit in the Nzema East Municipality[J]. Journal of Economics and Sustainable Development, 2013,4(1):128-132.
[5]	Mukhopadhyay P, Sengupta. Acceptance of Bengal Fasal Bima Yojona by the Farmers: A Study of Purba Burdwan District of West Bengal, India[J]. International Journal of Asian Business and Information Management, 2019,10(2):126-142. doi: 10.4018/IJABIM URL
[6]	吴晶妹, 张颖, 唐勤伟. 基于农户信用特征的WU’s三维信用评价模型研究[J]. 财贸经济, 2010(9):22-28,63.
[7]	周刚, 牛霞. 基于WU’s三维信用论的农户贷款信用风险研究——山东省实证分析[J]. 征信, 2016,34(2):16-23.
[8]	石宝峰, 王静. 基于ELECTRE III的农户小额贷款信用评级模型[J]. 系统管理学报, 2018,27(5):854-862.
[9]	邹建国, 李明贤. 农业供应链金融视角的农户信用增进研究[J]. 财经理论与实践, 2019,40(6):32-38.
[10]	温涛, 冉光和, 王煜宇, 等. 农户信用评估系统的设计与运用研究[J]. 运筹与管理, 2004(4):82-87.
[11]	胡愈, 许红莲, 王雄. 农户小额信用贷款信用评级探究[J]. 财经理论与实践, 2007(1):30-33.
[12]	谭民俊, 王雄, 岳意定. FPR-UTAHP评价方法在农户小额信贷信用评级中的应用[J]. 系统工程, 2007(5):55-59.
[13]	Hwang R, Chung H, Chu C K. Predicting Issuer Credit Ratings Using a Semiparametric Method[J]. Journal of Empirical Finance, 2010,17(1):120-137. doi: 10.1016/j.jempfin.2009.07.007 URL
[14]	陈永明, 周龙, 李双红. 基于AHP和DEMATEL方法的农户信用评级研究[J]. 征信, 2012,30(5):20-24.
[15]	谭燕芝, 李高娟. 基于Tobit模型的农户信用分析——来自CFPS的证据[J]. 区域经济评论, 2015(1):127-135.
[16]	迟国泰, 潘明道, 程砚秋. 基于综合判别能力的农户小额贷款信用评价模型[J]. 管理评论, 2015,27(6):42-57.
[17]	Dong G, Lai K K, Yen J. Credit Scorecard Based on Logistic Regression with Random Coefficients[J]. Procedia Computer Science, 2010,1(1):2463-2468. doi: 10.1016/j.procs.2010.04.278 URL
[18]	潘纬. 农户信用评价的指标设计与模型识别——基于安徽省金寨县的实践研究[J]. 征信, 2014,32(9):33-37.
[19]	朱莹, 吕德宏. 不同类型农户小额信贷信用风险影响因素研究——基于Logistic-ISM模型的实证分析[J]. 浙江金融, 2015(6):56-61.
[20]	王霞, 吕德宏. 基于多分类有序Logit模型的农户信用等级影响因素[J]. 中国农业大学学报, 2013,18(3):209-214.
[21]	Basri S, Naveen K K. Non-performance of Financial Contracts in Agricultural Lending: A Case Study from Karnataka, India[J]. Agricultural Finance Review, 2016,76(3):362-377. doi: 10.1108/AFR-01-2016-0001 URL
[22]	李莉, 张宗毅. 农户品质对农户信用影响实证分析——基于“农分期”16101个农户贷款数据[J]. 中国农业大学学报, 2019,24(1):206-216.
[23]	Kaiser H F. A second-generation little jiffy[J]. Psychometrika, 1970,35(4):401-415. doi: 10.1007/BF02291817 URL
[24]	胡胜, 雷欢欢, 胡华强. 基于Logistic模型的我国房地产企业信用风险度量研究[J]. 中国软科学, 2018(12):157-164.
[25]	张国政, 陈维煌, 刘呈辉. 基于Logistic模型的商业银行个人消费信贷风险评估研究[J]. 金融理论与实践, 2015(3):53-57.
[26]	付玮琼. 供应链金融视角下中小农业企业信用风险预警及防范研究[J]. 贵州社会科学, 2020(4):158-168.

一级指标	二级指标	英文	简称
户主特征	城乡分布	urban and rural distribution	UAR
	性别	sex	SEX
	年龄	age	AGE
	健康状况	health status	HEAL
	受教育程度	education level	EDU
	家庭成员数量	number of family members	FNUM
资产负债（还款能力）	现金资产	cash assets	CASH
	房产市值	property market value	PROP
	耐用消费品总值	total value of durable consumer goods	CON
	农用机械总值	total value of agricultural machinery	AGRM
	代偿借贷总额	compensatory loan	COML
家庭收支（还款能力）	农业收入	agricultural income	AGRI
	农业净收益	net agricultural income	NETAGRI
	非农收入	non-agricultural income	NONAGRI
	全年支出	annual expenditure	AEXP
还款意愿	人缘关系	popularity relation	POP
	勤奋认同	diligence identity	DIL
	信任认同	trust identity	TRUST
	风险偏好	risk appetite	RISK

一级指标	二级指标	英文	简称
户主特征	城乡分布	urban and rural distribution	UAR
	性别	sex	SEX
	年龄	age	AGE
	健康状况	health status	HEAL
	受教育程度	education level	EDU
	家庭成员数量	number of family members	FNUM
资产负债（还款能力）	现金资产	cash assets	CASH
	房产市值	property market value	PROP
	耐用消费品总值	total value of durable consumer goods	CON
	农用机械总值	total value of agricultural machinery	AGRM
	代偿借贷总额	compensatory loan	COML
家庭收支（还款能力）	农业收入	agricultural income	AGRI
	农业净收益	net agricultural income	NETAGRI
	非农收入	non-agricultural income	NONAGRI
	全年支出	annual expenditure	AEXP
还款意愿	人缘关系	popularity relation	POP
	勤奋认同	diligence identity	DIL
	信任认同	trust identity	TRUST
	风险偏好	risk appetite	RISK

指标	数据单位	最小值		最大值		均值		标准差
指标	数据单位	被拒	无被拒	被拒	无被拒	被拒	无被拒	被拒	无被拒
UAR	—	0	0	1	1	0.79	0.77	0.41	0.42
SEX	—	0	0	1	1	0.68	0.65	0.47	0.48
AGE	—	1	1	5	5	3.65	3.84	1.12	1.04
HEAL	—	1	1	5	5	2.73	2.86	1.23	1.27
EDU	—	1	1	6	7	2.36	2.24	1.09	1.06
FNUM	个	1	1	11	15	4.69	4.33	1.82	1.91
CASH	万元	0	0	32	450	2.07	2.85	4.60	11.70
PROP	万元	0.2	0.09	515	4012	28.17	29.79	55.98	103.50
CON	万元	0	0	40	100	3.68	2.93	5.95	5.96
AGRM	万元	0	0	70	30	0.93	0.39	5.43	1.33
COML	万元	0	0	200.01	150	5.84	1.58	18.71	5.89
AGRI	万元	0	0	28	66	1.33	1.02	3.23	2.78
NETAGRI	万元	-9.5	-51.5	13.5	45.93	-0.15	-0.13	1.83	2.86
NONAGRI	万元	0	0	101	55	4.86	3.73	10.08	4.12
AEXP	万元	0.1	0	60	50	4.47	3.29	6.15	3.59
POP	—	0	0	10	10	7.16	7.28	1.89	2.02
DIL	—	2	1	5	5	4.07	4.07	0.76	0.71
TRUST	—	0	0	1	1	0.47	0.55	0.50	0.50
RISK	—	0	0	1	1	0.18	0.19	0.39	0.39

指标	数据单位	最小值		最大值		均值		标准差
指标	数据单位	被拒	无被拒	被拒	无被拒	被拒	无被拒	被拒	无被拒
UAR	—	0	0	1	1	0.79	0.77	0.41	0.42
SEX	—	0	0	1	1	0.68	0.65	0.47	0.48
AGE	—	1	1	5	5	3.65	3.84	1.12	1.04
HEAL	—	1	1	5	5	2.73	2.86	1.23	1.27
EDU	—	1	1	6	7	2.36	2.24	1.09	1.06
FNUM	个	1	1	11	15	4.69	4.33	1.82	1.91
CASH	万元	0	0	32	450	2.07	2.85	4.60	11.70
PROP	万元	0.2	0.09	515	4012	28.17	29.79	55.98	103.50
CON	万元	0	0	40	100	3.68	2.93	5.95	5.96
AGRM	万元	0	0	70	30	0.93	0.39	5.43	1.33
COML	万元	0	0	200.01	150	5.84	1.58	18.71	5.89
AGRI	万元	0	0	28	66	1.33	1.02	3.23	2.78
NETAGRI	万元	-9.5	-51.5	13.5	45.93	-0.15	-0.13	1.83	2.86
NONAGRI	万元	0	0	101	55	4.86	3.73	10.08	4.12
AEXP	万元	0.1	0	60	50	4.47	3.29	6.15	3.59
POP	—	0	0	10	10	7.16	7.28	1.89	2.02
DIL	—	2	1	5	5	4.07	4.07	0.76	0.71
TRUST	—	0	0	1	1	0.47	0.55	0.50	0.50
RISK	—	0	0	1	1	0.18	0.19	0.39	0.39

指标	1	2	3	4	5	6	7	8	9	10	11
UAR	-0.039	-0.189	-0.056	0.027	0.025	0.187	0.273	0.545	0.301	-0.289	-0.052
SEX	0.016	-0.032	-0.001	-0.031	-0.002	0.837	-0.060	-0.009	0.007	0.019	-0.016
AGE	0.031	-0.533	0.032	0.072	0.008	0.250	-0.020	-0.085	-0.167	0.046	0.045
HEAL	-0.074	0.448	-0.029	0.106	-0.127	0.157	0.472	0.062	-0.139	-0.002	-0.235
EDU	0.019	0.379	-0.002	-0.064	0.015	0.249	-0.188	-0.108	-0.034	-0.056	0.167
FNUM	-0.126	-0.088	-0.021	-0.079	0.825	0.023	0.072	-0.023	0.047	-0.085	-0.012
CASH	-0.123	-0.008	0.026	0.662	0.056	0.009	-0.052	0.083	0.033	-0.032	0.006
PROP	0.029	-0.068	-0.031	0.591	-0.113	-0.044	0.068	-0.053	0.035	-0.030	-0.030
CON	0.059	0.086	0.024	0.087	0.352	-0.047	-0.069	0.043	-0.068	0.122	0.005
AGRM	0.009	0.093	0.014	-0.009	-0.014	-0.095	-0.190	0.774	-0.172	0.161	0.056
COML	0.483	-0.104	-0.050	-0.109	-0.303	0.050	0.094	0.028	0.079	-0.108	-0.011
AGRI	0.095	-0.015	0.580	-0.016	-0.034	0.014	-0.001	0.059	-0.005	-0.041	0.007
NETAGRI	-0.088	-0.038	0.552	0.014	0.018	-0.019	0.029	-0.079	0.019	-0.016	-0.027
NONAGRI	0.334	0.018	-0.117	0.069	0.092	-0.007	0.002	-0.036	0	0.012	0.009
AEXP	0.389	-0.024	0.144	-0.055	0.011	0.003	-0.005	-0.024	0.005	0.008	-0.024
POP	0.045	-0.035	0.03	-0.043	0.055	-0.107	0.734	-0.097	-0.027	0.039	0.118
DIL	-0.022	-0.048	-0.035	-0.029	-0.017	0.020	0.035	0.044	0.089	0.916	-0.036
TRUST	-0.014	-0.038	-0.017	0.001	-0.010	-0.016	0.070	0.033	0.025	-0.044	0.949
RISK	0.026	0.032	0.013	0.032	0	-0.013	-0.062	-0.062	0.885	0.095	0.019