Periodic Reporting for period 1 - HousingAndMortgages (Welfare, redistribution and financial stability in housing and mortgage markets)
Período documentado: 2022-09-01 hasta 2025-02-28
Worldwide, housing wealth represents the most important asset in households’ balance sheet, and mortgage debt is the most important liability. In the past two decades, several countries have experienced house price bubbles, excessive households’ leverage, and increasing risk-taking behavior of banks. To limit these phenomena, and in response to the 2008 financial crisis, new regulations have been introduced in housing and mortgage markets. The lack of disaggregate and extensive micro-level data on housing and mortgages has so far prevented researchers from carefully investigating these events and the effectiveness of regulation.
This project will bridge this gap. I will assemble a unique and extensive platform of datasets on housing and mortgage markets, combining disaggregate information on housing transactions, buyers and sellers, and loan level data on mortgages for two European countries, the Netherlands and Norway. These data will be used to develop and estimate novel structural econometric models of demand and supply in housing and mortgage markets. These models will serve to investigate three main questions of concern to policymakers, and to evaluate the welfare effects of existing and alternative regulations via counterfactual simulations.
First, I will investigate a novel demand channel of housing and mortgages driven by the rise of accommodation sharing platforms such as Airbnb. While these platforms provide extra income to households renting their property, they also fuel housing bubbles and affect mortgage markets. Second, I will evaluate the role of mortgage securitization in reducing lenders’ funding costs, quantify its effect on lenders’ risk-taking behavior, and propose regulations to balance this trade-off. Last, I will document the distributional effects of leverage regulations, that have helped to reduce credit risk, but have also disproportionally penalized low-income households and first-time buyers, worsening income and wealth inequality.
This project will bridge this gap. I will assemble a unique and extensive platform of datasets on housing and mortgage markets, combining disaggregate information on housing transactions, buyers and sellers, and loan level data on mortgages for two European countries, the Netherlands and Norway. These data will be used to develop and estimate novel structural econometric models of demand and supply in housing and mortgage markets. These models will serve to investigate three main questions of concern to policymakers, and to evaluate the welfare effects of existing and alternative regulations via counterfactual simulations.
First, I will investigate a novel demand channel of housing and mortgages driven by the rise of accommodation sharing platforms such as Airbnb. While these platforms provide extra income to households renting their property, they also fuel housing bubbles and affect mortgage markets. Second, I will evaluate the role of mortgage securitization in reducing lenders’ funding costs, quantify its effect on lenders’ risk-taking behavior, and propose regulations to balance this trade-off. Last, I will document the distributional effects of leverage regulations, that have helped to reduce credit risk, but have also disproportionally penalized low-income households and first-time buyers, worsening income and wealth inequality.
The main objectives of the grant include the completion and dissemination of three research papers. After two years, relative to the Description of Action, the progress is the following.
The paper described as Work Package 3 is completed, it has been extensively presented at conferences and seminars, and it has been submitted to a journal.
The paper described as Work Package 1 is in progress. The reduced-form descriptive analysis has been completed. The data collection for the structural model has been completed. The model is currently being finalized, and once that is done, it will be possible to proceed with estimation and counterfactuals. The presentation and dissemination of the paper will likely begin in Spring 2025, with the target of submitting to a journal at the end of 2025 or early 2026.
The paper described as Work Package 2 is in progress. The data collection is currently being carried out, integrating the existing data produced for Work Package 1 with new data on securitization. The model is in its early-stage definition, and once the data collection is completed, the descriptive statistics from the data will be used to finalize the layout of the model. The presentation and dissemination of the paper will likely begin in Spring 2026, with the target of submitting to a journal at the end of 2026 or early 2027.
So far, the most significant achievements of the project have been the following:
(1) The hiring of two excellent team members who are collaborating on all work packages of the project. Yushi Peng, hired as a Post-Doc, and Marc Stam, hired as a PhD student.
(2) The dissemination and presentation of the results of Work Package 3 to diverse audiences, including top academic departments and several central banks.
(3) The organization of a conference on financial intermediation in Tilburg in June 2024, including papers strongly related to the focus of the project.
(4) The submission of a paper based on Work Package 3 to one of the top 5 journals in economics.
(5) The completion of the data collection for all work packages.
The paper described as Work Package 3 is completed, it has been extensively presented at conferences and seminars, and it has been submitted to a journal.
The paper described as Work Package 1 is in progress. The reduced-form descriptive analysis has been completed. The data collection for the structural model has been completed. The model is currently being finalized, and once that is done, it will be possible to proceed with estimation and counterfactuals. The presentation and dissemination of the paper will likely begin in Spring 2025, with the target of submitting to a journal at the end of 2025 or early 2026.
The paper described as Work Package 2 is in progress. The data collection is currently being carried out, integrating the existing data produced for Work Package 1 with new data on securitization. The model is in its early-stage definition, and once the data collection is completed, the descriptive statistics from the data will be used to finalize the layout of the model. The presentation and dissemination of the paper will likely begin in Spring 2026, with the target of submitting to a journal at the end of 2026 or early 2027.
So far, the most significant achievements of the project have been the following:
(1) The hiring of two excellent team members who are collaborating on all work packages of the project. Yushi Peng, hired as a Post-Doc, and Marc Stam, hired as a PhD student.
(2) The dissemination and presentation of the results of Work Package 3 to diverse audiences, including top academic departments and several central banks.
(3) The organization of a conference on financial intermediation in Tilburg in June 2024, including papers strongly related to the focus of the project.
(4) The submission of a paper based on Work Package 3 to one of the top 5 journals in economics.
(5) The completion of the data collection for all work packages.
The project has achieved the following advancements in developing novel methodologies.
Work Package 3 has developed a structural model of housing demand and supply that extends the state-of-the-art approach developed by Bayer et al. (2016), introducing five new features. First, we allow two types of real estate investors to demand and supply housing products: financially constrained households, who mostly own and exchange a single housing product and face transaction costs, and financially unconstrained investors, who own portfolios of properties and face no transaction costs. Second, we distinguish between owners and renters among households, allowing renters to become owners and vice versa. Third, all households can stay in their current property, so our analysis not only focuses
on those transacting in the housing market during a particular sample period. Fourth, due to household balance sheet data, we can explicitly model households’ heterogeneous aordability constraints. Last, we model the equilibrium pricing in housing markets via a market clearing condition that incorporates households’ and investors’ demand and supply of properties.
Work Package 1 is extending the state-of-the-art mortgage demand and supply models, as in Benetton (2021), by allowing households to choose not only a mortgage product, but a bundle of housing and mortgage product.
Work Package 2 is the extending the state-of-the-art literature on mortgage demand and supply and securitization, as in Buchak et al. (2022). While Buchak et al. (2022) model mortgage demand and lenders’ pricing and securitization decision, where the latter is just the lender’s choice of how many mortgages to keep on the balance sheet vs securitizing, our more detailed data allows us to go one step further. We can in fact model lenders’ decision to include a mortgage in a specific mortgage pool, of which we observe the identifier, and can then trace which investor purchased that pool and at what conditions. This allows us to model more extensively lenders’ securitization decisions and investors’ demand for securitized mortgages.
Work Package 3 has developed a structural model of housing demand and supply that extends the state-of-the-art approach developed by Bayer et al. (2016), introducing five new features. First, we allow two types of real estate investors to demand and supply housing products: financially constrained households, who mostly own and exchange a single housing product and face transaction costs, and financially unconstrained investors, who own portfolios of properties and face no transaction costs. Second, we distinguish between owners and renters among households, allowing renters to become owners and vice versa. Third, all households can stay in their current property, so our analysis not only focuses
on those transacting in the housing market during a particular sample period. Fourth, due to household balance sheet data, we can explicitly model households’ heterogeneous aordability constraints. Last, we model the equilibrium pricing in housing markets via a market clearing condition that incorporates households’ and investors’ demand and supply of properties.
Work Package 1 is extending the state-of-the-art mortgage demand and supply models, as in Benetton (2021), by allowing households to choose not only a mortgage product, but a bundle of housing and mortgage product.
Work Package 2 is the extending the state-of-the-art literature on mortgage demand and supply and securitization, as in Buchak et al. (2022). While Buchak et al. (2022) model mortgage demand and lenders’ pricing and securitization decision, where the latter is just the lender’s choice of how many mortgages to keep on the balance sheet vs securitizing, our more detailed data allows us to go one step further. We can in fact model lenders’ decision to include a mortgage in a specific mortgage pool, of which we observe the identifier, and can then trace which investor purchased that pool and at what conditions. This allows us to model more extensively lenders’ securitization decisions and investors’ demand for securitized mortgages.