Disclaimer:The views expressed herein are solely those of the author and do not represent the views of Banco de Portugal or the Eurosystem.
Applied and Policy Research
This paper estimates recovery parameters for corporate loans in Portugal using granular data from Banco de Portugal’s Credit Register. To correct for sample selection bias present in recovery databases, it models jointly recoveries and the time to recovery. Covering the period from 2009 to 2024, the dataset provides monthly estimates of expected recovery, uncertainty, and time to recovery. These estimates are generally stable, with greater variability after September 2018 due to a more comprehensive loan characterization and improved data granularity. Recovery parameter estimates can be used in applications such as credit risk monitoring, loan pricing, and regulatory capital estimation. The methodology can be applied to similar credit registers.
This paper presents a comprehensive dataset of recovery rates for all defaulted banking loans to firms and individuals in Portugal from 2009 to 2024, based on Banco de Portugal’s Credit Register. The dataset includes the recovery rates and time to recovery for over 3.2 million exposures. From September 2018 onward, it contains loan-level data on cumulative recoveries recorded monthly from the time of default.
Are lenders in the Portuguese financial system more likely to have large losses now than in the past? If a lender has large losses, is it more likely that another one will as well? How has that likelihood changed over time? We address these and other questions using granular credit exposure data in the period between 2009 and 2023. Our findings indicate that the risk of large losses is lower in 2023 than in 2012. Behind this result is a reduction in the borrowers’ default probabilities, a decline in the share of credit to firms accompanied by a rise in the share of credit to households and, to a more limited extent, an increase in loan recoveries. Additionally, we find that the risk of multiple lenders experiencing large losses simultaneously has decreased during the period of analysis. But, if one lender has large losses, the risk that another one will also face large losses has been rising since 2019. This result is driven by an increase in credit to sectors that have high default risk correlation.
This paper estimates econometric models of default risk for individuals obtaining credit in Portugal using data from Banco de Portugal’s Credit Register. We estimate monthly default probabilities for mortgage and consumer loans over three, six, and twelve-month horizons. The models combine cross-sectional and time series components. The cross-sectional component captures default risk heterogeneity across individuals by relating default risk to loan and borrower characteristics. The time series component captures time variation in aggregate default risk by linking it with macroeconomic variables. Our findings indicate that the model’s performance in distinguishing between defaulting and non-defaulting borrowers is on par with or superior to existing literature. The results also show a close alignment between average default probabilities and actual default rates across various borrower characteristics and lending institutions.
... was part of a research team that develops portfolio analytics for credit institutions. I worked on the estimation of models used for stress testing and IFRS 9 impairments. I developed approaches for reverse stress testing, the measurement of settlement risk, and the back-allocation to individual instruments of portfolio measures of risk (e.g. economic capital). I m familiar with models used to estimate the value distribution of credit portfolios, and with the estimation of the different inputs of these models: Correlations and transition probabilities. I m also familiar with portfolio optimization. I m knowledgeable in IFRS 9 and in the Basel regulation on capital for credit risk, counterparty credit risk (SA-CCR), and market risk (FRTB).
Publications
We analyze the effects of sovereign loan guarantees on financial stability in Portugal using a DSGE model. Sovereign loan guarantees decrease the default rate of banks and increase credit. On the other hand, guarantees increase the leverage and default rate of firms. These effects are larger the lower the sensitivity of the capital of banks to capital requirements. Behind these results are the reduction in regulatory risk-weights and the transfer of loan losses from banks to the sovereign brought by sovereign loan guarantees. A decomposition of the impact of sovereign loan guarantees suggests that insuring banks against loan losses can complement and enhance conventional macroprudential policy.
We analyze the pricing of firm loans originated by the largest banks operating in Portugal between September 2018 and December 2019. On average loans are overpriced in the short-run and underpriced in the long-run. Underpricing is lower in the subsample of loans originated in 2019. Loans with maturity longer than a year, loans to the construction, real estate, and transportation and storage sectors, and loans to high credit risk borrowers are on average underpriced. Loans to other borrowers are generally overpriced. Borrowers and sectors with underpriced loans are potential sources of fragility in the financial sector, especially in the medium to long-run.
How does financial development affect the magnitude of the business cycles fluctuations? We examine this question in a general equilibrium model with heterogeneous agents and endogenous credit constraints based on Kiyotaki (1998). We show that there is a hump-shaped relationship between the degree of financial frictions and the amplification of unexpected productivity shocks. This non-monotonic relation is due to the fall in financial frictions having two opposite effects on the response of output. One effect is the reallocation of productive inputs between agent types, which, while active, increases with the fall in financial frictions. The other effect is the change in the demand of inputs, which decreases with the fall in financial frictions. At low levels of financial development the reallocation effect dominates and a fall in financial frictions increases the amplification of productivity shocks. In contrast, at higher levels of financial development, a fall in financial frictions decreases the shock amplification because the reallocation effect disappears while the effect on the demand of inputs is still present.
This paper evaluates the impact on consumer welfare of the constraints on increasing interest rates present in the Credit Card Act. We develop a model of consumer financing in a setting where information asymmetry between a consumer and a lender arises over time and triggers adverse selection. We find the competitive equilibrium of this setting and show that restrictions on increasing the interest rate, as in the Credit Card Act, are welfare decreasing for a large set of parameters. The Card's restrictions lead to higher credit card interest rates for low credit-quality consumers and lower credit limit for high credit-quality consumers; these negative effects on welfare are only partially offset by lower up-front fees. We find similar results when we extend our analysis to settings in which consumers have limited rationality.
Working Papers
This paper analyzes the effects of incentive-pay and accounting discretion on earnings management. I develop a dynamic model of earnings reporting in which the cost of earnings management is endogenous, and in which a privately informed manager trades-off incentives to manipulate earnings
which increases the firm's stock-price and his current period payoff against
incentives to be honest which improves his reputation and future payoff.
I find that more incentive-pay and more accounting discretion can lead to
less earnings management. Incentive-pay and accounting discretion make a
manager's per-period payoff more sensitive to reports. This effect increases
both the manager's current period payoff to manipulating earnings, and the
manager's payoff to having a better reputation. When having a better
reputation dominates the benefit of increasing the current period payoff,
the manager does less earnings management.
Work-in-Progress
Reputation Concerns, Optimal Contracts, and Performance Measures, with Xiaojing Meng and Haresh Sapra
Old Papers
Financial Analysts,
Trade Generating Incentives and Reputation (November, 2008)