Publicação
Essays on macroprudential policy
| Resumo: | This thesis aims to add to the existing literature tools for calibrating an macroprudential instrument - the countercyclical capital buffer (CCyB) - and to assess an implemented borrower-based measure - a limit to the debt service-to-income ratio (DSTI) - in both a low and in an increasing and higher interest rate environment. Macroprudential policy became relevant after the 2008 financial crisis, as systemic risk - the accumulation of financial imbalances that led to excessive credit and booming housing prices - was not adequately covered by existing resilience. This interaction between risk and resilience is explored in the first chapters. In the first chapter, the focus is on how ensuring enough resilience in the financial system implies managing a trade-off between expected bank profitability and downside risk in banks’ profitability. To describe this trade-off, a dynamic quantile regression model using bank-level data for Portugal that links future bank profitability to the current cyclical systemic risk environment net of the prevailing level of capital-based resilience (residual cyclical systemic risk) is estimated. Results indicate that an increase in cyclical systemic risk negatively affects the conditional distribution of bank profitability at the medium-term projection horizon (11 to 16 quarters ahead), confirming the findings in the literature. At the same time, the increase of capital-based resilience is found to counter the negative effects of cyclical systemic risk on banks profitability in the same projection horizon. Using these results, a novel calibration rule for the CCyB is defined, which is flexible enough to accommodate different preferences of the policymaker and that factors in the prevailing levels of cyclical systemic risk and capital-based resilience. This rule targets downside risk in banks’ profitability: the CCyB rate should be enough that all existing capital-based resilience (including the calibrated CCyB rate) can absorb the effect of cyclical systemic risk on deviating a measure of downside risk (in this case on the distance between the mean and the 10th percentile of banks profitability distribution in the medium-term) from the policymaker’s target. In the second chapter, evolutionary game theory is used to identify stable strategies of cyclical systemic risk and targeted resilience through macroprudential policy, specifically the CCyB. The model considers that each decision or interaction performed in the financial system can be defined within the dichotomy risk-resilience (either high or low) which receives a payoff. The payoffs are operationalized by considering quantile regressions on banks’ profitability. The payoff of high or low risk activities will be awarded based on their short-term impact on banks’ median profitability and depending on the state of macroprudential policy. The payoff of macroprudential policy will be assessed based on its medium-term impact on the ii tail of the profitability distribution and will depend on the state of cyclical systemic risk. Both specifications also depend on the level of financial stress in the system. Results show that there are two evolutionary stable strategies that depend on the level of financial stress, i.e., two strategies that, when achieved, are robust to small disturbances in the system. When financing conditions are good (financial stress is low), cyclical systemic risk will tend to increase, as agents take advantage of good conditions to engage in riskier activities; from a risk threshold onwards, the benefits of tightening macroprudential policy (of increasing resilience) will outweigh its costs. When financial stress is heightened, cyclical systemic risk will materialize and the costs of having macroprudential policy will be higher than its benefits, indicating a need to loosen the policy. In the third chapter, the impact of the Portuguese DSTI limit on the distribution of the loan service-to-income (LSTI) ratio for new household loans for house purchases is assessed in both low and increasing interest rate environments. The choice for assessing the impact on the LSTI ratio of new loans relies on the fact that they will be both constrained by the DSTI limit and by rising interest rates. The purpose of studying the effects over the distribution is that it is expected that interest rates increases (driven by reference rates) will have a more significant effect in increasing the number of loans given at a higher LSTI ratio, if no restriction was in place, and so it can increase the right tail at a larger magnitude than in the rest of the distribution. However, at the same time, because there is a DSTI limit, the interest rate increase might induce households to take out smaller loans or even exclude them entirely from the credit market, which can offset the previous effect. In order to estimate the two effects, instrumental variable quantile regressions are employed and the exceptions to the DSTI limit, foreseen in the Portuguese Recommendation, are used as a potential counterfactual to explore how differences between similar loans that belong to each one of the two groups (exceptions and non-exceptions) impacts the conditional distribution of the LSTI ratio. Results indicate that, in the absence of the DSTI limit, the LSTI distribution could shift rightward, both before and after interest rates started to increase, reflecting a higher financial burden on borrowers. The impact of the limit is more pronounced after March 2022, demonstrating its efficacy in mitigating the impact of rising interest rates by more stringently restricting higher LSTI ratios. Results throughout the thesis indicate that macroprudential policy can be effective in increasing resilience and curbing systemic risk in the financial system. The multiple tools, either capital-based or borrower-based, allow for a targeted impact depending on how systemic risk is originating in the financial system. |
|---|---|
| Autores principais: | Passinhas, Joana Luzia Monteiro |
| Assunto: | Macroprudential policy Cyclical systemic risk Financial system Evolutionary game theory Quantile regression Política macroprudencial Risco sistémico cíclico Sistema financeiro Teoria de jogos evolutivos Regressão de quantis |
| Ano: | 2025 |
| País: | Portugal |
| Tipo de documento: | tese de doutoramento |
| Tipo de acesso: | acesso aberto |
| Instituição associada: | Universidade de Lisboa |
| Idioma: | inglês |
| Origem: | Repositório da Universidade de Lisboa |
| Resumo: | This thesis aims to add to the existing literature tools for calibrating an macroprudential instrument - the countercyclical capital buffer (CCyB) - and to assess an implemented borrower-based measure - a limit to the debt service-to-income ratio (DSTI) - in both a low and in an increasing and higher interest rate environment. Macroprudential policy became relevant after the 2008 financial crisis, as systemic risk - the accumulation of financial imbalances that led to excessive credit and booming housing prices - was not adequately covered by existing resilience. This interaction between risk and resilience is explored in the first chapters. In the first chapter, the focus is on how ensuring enough resilience in the financial system implies managing a trade-off between expected bank profitability and downside risk in banks’ profitability. To describe this trade-off, a dynamic quantile regression model using bank-level data for Portugal that links future bank profitability to the current cyclical systemic risk environment net of the prevailing level of capital-based resilience (residual cyclical systemic risk) is estimated. Results indicate that an increase in cyclical systemic risk negatively affects the conditional distribution of bank profitability at the medium-term projection horizon (11 to 16 quarters ahead), confirming the findings in the literature. At the same time, the increase of capital-based resilience is found to counter the negative effects of cyclical systemic risk on banks profitability in the same projection horizon. Using these results, a novel calibration rule for the CCyB is defined, which is flexible enough to accommodate different preferences of the policymaker and that factors in the prevailing levels of cyclical systemic risk and capital-based resilience. This rule targets downside risk in banks’ profitability: the CCyB rate should be enough that all existing capital-based resilience (including the calibrated CCyB rate) can absorb the effect of cyclical systemic risk on deviating a measure of downside risk (in this case on the distance between the mean and the 10th percentile of banks profitability distribution in the medium-term) from the policymaker’s target. In the second chapter, evolutionary game theory is used to identify stable strategies of cyclical systemic risk and targeted resilience through macroprudential policy, specifically the CCyB. The model considers that each decision or interaction performed in the financial system can be defined within the dichotomy risk-resilience (either high or low) which receives a payoff. The payoffs are operationalized by considering quantile regressions on banks’ profitability. The payoff of high or low risk activities will be awarded based on their short-term impact on banks’ median profitability and depending on the state of macroprudential policy. The payoff of macroprudential policy will be assessed based on its medium-term impact on the ii tail of the profitability distribution and will depend on the state of cyclical systemic risk. Both specifications also depend on the level of financial stress in the system. Results show that there are two evolutionary stable strategies that depend on the level of financial stress, i.e., two strategies that, when achieved, are robust to small disturbances in the system. When financing conditions are good (financial stress is low), cyclical systemic risk will tend to increase, as agents take advantage of good conditions to engage in riskier activities; from a risk threshold onwards, the benefits of tightening macroprudential policy (of increasing resilience) will outweigh its costs. When financial stress is heightened, cyclical systemic risk will materialize and the costs of having macroprudential policy will be higher than its benefits, indicating a need to loosen the policy. In the third chapter, the impact of the Portuguese DSTI limit on the distribution of the loan service-to-income (LSTI) ratio for new household loans for house purchases is assessed in both low and increasing interest rate environments. The choice for assessing the impact on the LSTI ratio of new loans relies on the fact that they will be both constrained by the DSTI limit and by rising interest rates. The purpose of studying the effects over the distribution is that it is expected that interest rates increases (driven by reference rates) will have a more significant effect in increasing the number of loans given at a higher LSTI ratio, if no restriction was in place, and so it can increase the right tail at a larger magnitude than in the rest of the distribution. However, at the same time, because there is a DSTI limit, the interest rate increase might induce households to take out smaller loans or even exclude them entirely from the credit market, which can offset the previous effect. In order to estimate the two effects, instrumental variable quantile regressions are employed and the exceptions to the DSTI limit, foreseen in the Portuguese Recommendation, are used as a potential counterfactual to explore how differences between similar loans that belong to each one of the two groups (exceptions and non-exceptions) impacts the conditional distribution of the LSTI ratio. Results indicate that, in the absence of the DSTI limit, the LSTI distribution could shift rightward, both before and after interest rates started to increase, reflecting a higher financial burden on borrowers. The impact of the limit is more pronounced after March 2022, demonstrating its efficacy in mitigating the impact of rising interest rates by more stringently restricting higher LSTI ratios. Results throughout the thesis indicate that macroprudential policy can be effective in increasing resilience and curbing systemic risk in the financial system. The multiple tools, either capital-based or borrower-based, allow for a targeted impact depending on how systemic risk is originating in the financial system. |
|---|