THE ECONOMIC EXPLAINABILITY OF MACHINE LEARNING AND STANDARD ECONOMETRIC MODELS-AN APPLICATION TO THE U.S. MORTGAGE DEFAULT RISK

This study aims to bridge the gap between two perspectives of explainability−machine learning and engineering, and economics and standard econometrics−by applying three marginal measurements. The existing real estate literature has primarily used econometric models to analyze the factors that affect the default risk of mortgage loans. However, in this study, we estimate a default risk model using a machine learning-based approach with the help of a U.S. securitized mortgage loan database. Moreover, we compare the economic explainability of the models by calculating the marginal effect and marginal importance of individual risk factors using both econometric and machine learning approaches. Machine learning-based models are quite effective in terms of predictive power; however, the general perception is that they do not efficiently explain the causal relationships within them. This study utilizes the concepts of marginal effects and marginal importance to compare the explanatory power of individual input variables in various models. This can simultaneously help improve the explainability of machine learning techniques and enhance the performance of standard econometric methods.

Analyzing Characteristics and Implications of the Mortgage Default of Agricultural Land Management Rights in Recent China Based on 724 Court Decisions

The transfer of rural land contractual management rights belongs to the recessive transition of land use. The mortgage of rural land management rights is a way of rural land circulation, and has an important impact on the transformation of land use. Rural land management rights mortgage loans can enable farmers to obtain more credit funds, which is conducive to agricultural development and Rural Revitalization. However, with the development of rural land mortgage financing, the associated risk has become increasingly prominent. The most typical risk is the default risk of farmers’ mortgage loans. Based on court decisions regarding rural land mortgage default during 2014–2020, this paper analyzes the characteristics of farmers’ default in different periods and locations. The empirical results reveal that the time and space of rural land mortgage default cases are widely distributed in China, especially in Heilongjiang Province. In the default judgement, the loan amount of CNY 50,000 to CNY 100,000 and the loan periods of 1 year accounted for the highest proportion. When making mortgage loan policies for rural land management rights, financial institutions should give farmers the most preferential treatment regarding the amount, term and interest rate of loans. Farmers’ social security should be improved, and agricultural insurance should be strengthened. Meanwhile, the credit review of small and short-term loan farmers should be heightened.

Buildings’ Energy Efficiency and the Probability of Mortgage Default: The Dutch Case

We investigate the relationship between building energy efficiency and the probability of mortgage default. To this end, we construct a novel panel data set by combining Dutch loan-level mortgage information with provisional building energy ratings provided by the Netherlands Enterprise Agency. Using the logit regression and the extended Cox model, we find that building energy efficiency is associated with a lower probability of mortgage default. There are three possible channels that might drive the results: (i) personal borrower characteristics captured by the choice of an energy-efficient building, (ii) improvements in building performance that could help to free-up the borrower’s disposable income, and (iii) improvements in dwelling value that lower the loan-to-value ratio. We address all three channels. In particular, we find that the default rate is lower for borrowers with less disposable income. The results hold for a battery of robustness checks. This suggests that the energy efficiency ratings complement borrowers’ credit information and that a lender using information from both sources can make superior lending decisions than a lender using only traditional credit information. These aspects are not only crucial for shaping future energy policy, but also have implications for the risk management of European financial institutions.

Differential Replication for Credit Scoring in Regulated Environments

Differential replication is a method to adapt existing machine learning solutions to the demands of highly regulated environments by reusing knowledge from one generation to the next. Copying is a technique that allows differential replication by projecting a given classifier onto a new hypothesis space, in circumstances where access to both the original solution and its training data is limited. The resulting model replicates the original decision behavior while displaying new features and characteristics. In this paper, we apply this approach to a use case in the context of credit scoring. We use a private residential mortgage default dataset. We show that differential replication through copying can be exploited to adapt a given solution to the changing demands of a constrained environment such as that of the financial market. In particular, we show how copying can be used to replicate the decision behavior not only of a model, but also of a full pipeline. As a result, we can ensure the decomposability of the attributes used to provide explanations for credit scoring models and reduce the time-to-market delivery of these solutions.