Stress Test Model Research Articles

Shock Amplification in an Interconnected Financial System of Banks and Investment Funds

This paper shows how the combined endogenous reaction of banks and investment funds to an exogenous shock can amplify or dampen losses to the financial system compared to results from single-sector stress testing models. We build a new model of contagion propagation using a very large and granular data set for the euro area. Based on the economic shock caused by the Covid-19 outbreak, we model three sources of exogenous shocks: a default shock, a market shock and a redemption shock. Our contagion mechanism operates through a dual channel of liquidity and solvency risk. The joint modelling of banks and funds provides new insights for the assessment of financial stability risks. Our analysis reveals that adding the fund sector to our model for banks leads to additional losses through fire sales and a further depletion of banks’ capital ratios by around one percentage point.

Loan fair values and the financial crisis

PurposeIn 2009, effective the second-quarter, the financial accounting standards board mandated that all banks need to disclose the fair value of loans in their 10-Q filings in addition to their 10-K filings. This paper aims to investigate whether these disclosures reduced the level of information asymmetry about the riskiness of bank loan portfolios during the financial crisis.Design/methodology/approachThe paper examines the impact of these disclosures on the bid-ask spread of a panel of 246 publicly traded bank holding companies. The spread serves as a proxy for information asymmetry and the ratio of the fair value of a bank’s loan portfolio to its book value is a proxy for the credit and liquidity risk associated with the same. The reaction to the first-quarter filing serves as a control to assess the reaction at the time of the second-quarter filing.FindingsThere is a significant negative association between bid-ask spread and the ratio indicating that the fair value information was useful in reducing information asymmetry during the financial crisis. A pattern was observed in the information dissemination related to the fair value of loans that is consistent with the literature that documents a delayed investor reaction to complex financial information.Originality/valueInvestors may use the fair value information to better assess the risk profile of a BHC’s loan portfolio. Also, loan fair values provide managers with data to better implement stress test models and determine optimal capital buffers.

Adjusting loss forecasts for the impacts of government assistance and loan forbearance during the COVID-19 recession

Since the onset of the COVID-19 recession, loss forecasting and stress testing models have dramatically overpredicted losses. As all models are pattern recognisers trained on past events, such an unprecedented event inevitably leads to model errors. Rather than, however, view the models as broken, they are useful in providing an upper bound of what could have happened if government assistance and loan forbearance had not been provided. The present work develops an approach for quantifying the short- and long-term impacts of these government and lender policies in order to create quantitative model overlays. These overlays express the problem via a set of key parameters that can be set via management judgment or simulation studies. Examples of this approach and parameter sensitivity analysis are provided using time series models of National Credit Union Administration and Federal Deposit Insurance Corporation call report data. This paper provides a framework for incorporating simulations, simple to complex, into an existing stress testing framework to better project future losses.

Banking system stress testing and COVID-19: A first summary appraisal

This paper dicusses the special banking system stress tests which central banks for the USA, the UK and the euro area published shortly after the outbreak of COVID-19. Based on the limited quantitative information available on a comparable basis, this paper discusses how the results seem broadly consistent across institutions as well as with previously conducted exercises, once accounting for the respective differences in scenario severity. The paper also discusses how central banks identified specific design features characterising these COVID-19 exercises, employing the typology put forward in the Bank for International Settlements’ (BIS) Financial Stability Institute (FSI) insights #12 — considering three building blocks: governance, implementation and outcomes. In particular, these special exercises set a new, more demanding benchmark for scenario adversity. Top-down modelling was also used to a larger extent to deal with uncertainty via alternative scenarios and to assess the impact of policies. Moreover, only aggregate results were published, with limited information on the results across banks. Finally, the connection with supervisory decisions appeared loose at most. The latter two features are not expected to remain in subsequent regular supervisory system-wide exercises. At the same time, the granularity of information collected, the banks in scope or the treatment of spillovers and feedbacks largely remained in line with standard supervisory stress-testing practices. This admittedly came short of what could be needed to get a more complete assessment of the impact of this specific crisis, from both a microprudential or macroprudential perspective. In particular, sector-specific information was sparse and banks’ credit deleveraging explored only for the UK. Looking ahead, it appeared that stress-testing models can represent a valuable forecasting and monitoring tool, especially when baseline conditions are adverse enough and uncertainty prevails. The review also suggests that bottom-up and top-down system-wide supervisory stress-test results could coexist, thereby giving further concrete relevance to recent proposals on the future of European Union area-wide stress tests.

COVID-19 implications for banks: evidence from an emerging economy.

The COVID-19 pandemic is damaging economies across the world, including financial markets and institutions in all possible dimensions. For banks in particular, the pandemic generates multifaceted crises, mostly through increases in default rates. This is likely to be worse in developing economies with poor financial market architecture. This paper utilizes Bangladesh as a case study of an emerging economy and examines the possible impacts of the pandemic on the country’s banking sector. Bangladesh’s banking sector already has a high level of non-performing loans (NPLs) and the pandemic is likely to worsen the situation. Using a state-designed stress testing model, the paper estimates the impacts of the COVID-19 pandemic on three particular dimensions—firm value, capital adequacy, and interest income—under different NPL shock scenarios. Findings suggest that all banks are likely to see a fall in risk-weighted asset values, capital adequacy ratios, and interest income at the individual bank and sectoral levels. However, estimates show that larger banks are relatively more vulnerable. The decline in all three dimensions will increase disproportionately if NPL shocks become larger. Findings further show that a 10% NPL shock could force capital adequacy of all banks to go below the minimum BASEL-III requirement, while a shock of 13% or more could turn it to zero or negative at the sectoral level. Findings call for immediate and innovative policy measures to prevent a large-scale and contagious banking crisis in Bangladesh. The paper offers lessons for other developing and emerging economies similar to Bangladesh.Electronic supplementary materialThe online version of this article (10.1007/s43546-020-00013-w) contains supplementary material, which is available to authorized users.

The Impact of Data Aggregation and Risk Attributes on Stress Testing Models of Mortgage Default

Stress testing models have been developed at various levels of data aggregation with or without risk attributes, but there is limited research on the joint impact of these modeling choices. In this paper, we investigate how data aggregation and risk attributes affect the development and performance of stress testing models by studying residential mortgage loan defaults. We develop mortgage default models at various data aggregation levels including loan-level, segment-level, and top-down. We also compare the models with and without risk attributes as control variables. We assess model performance for goodness-of-fit, prediction accuracy, and projection sensitivity for stress testing purposes. We find that the loan-level models do not always win among models with various data aggregation levels, and including risk attributes greatly improves goodness-of-fit and projection accuracy for models of all data aggregation levels. The findings suggest that it is important to consider data aggregation and risk attributes in developing stress testing models.

The impact of data aggregation and risk attributes on stress testing models of mortgage default

Stress testing models have been developed at various levels of data aggregation with or without risk attributes, but there is limited research on the joint impact of these modeling choices. In this paper, we investigate how data aggregation and risk attributes affect the development and performance of stress testing models by studying residential mortgage loan defaults. We develop mortgage default models at various data aggregation levels including loan-level, segment-level, and top-down. We also compare the models with and without risk attributes as control variables. We assess model performance for goodness-of-fit, prediction accuracy, and projection sensitivity for stress testing purposes. We find that the loan-level models do not always win among models with various data aggregation levels, and including risk attributes greatly improves goodness-of-fit and projection accuracy for models of all data aggregation levels. The findings suggest that it is important to consider data aggregation and risk attributes in developing stress testing models.

Vulnerable asset management? The case of mutual funds

Is the asset management sector a source of financial instability? This paper develops a macroprudential stress test model which enables the quantification of systemic vulnerabilities due to fire sales in this sector. The model incorporates the flow-performance relationship as an additional funding shock in the model of Greenwood et al. (2015). Using data on US equity mutual funds for the period 2003–14, we quantify both fund-specific and system-wide (aggregate) vulnerabilities to fire sales over time. Our main finding is that the aggregate vulnerability, according to this propagation mechanism, is relatively small in comparison with values reported for banks. However, during periods of low market liquidity, the vulnerability of the system can become significant. Our paper also contributes to the ongoing discussion on the SIFI designation of Non-Bank Non-Insurer entities. For this purpose, we explore the determinants of individual funds’ vulnerability to systemic asset liquidations, highlighting the importance of size and portfolio illiquidity. Therefore, regulators should monitor structural vulnerabilities in the fund sector arising through liquidity transformation.

Systemic Stress Test Model in Shared Portfolio Networks

We propose a dynamic model for systemic risk using a bipartite network of banks and assets in which the nodes and links may vary over time. We apply the model to the European sovereign debt crisis and observe that the results closely match real-world events (e.g., the high risk of Greek sovereign bonds and the distress of Greek banks). The model could become complementary to existing stress tests, incorporating the systemic risk contribution of banks and assets in time-dependent networks. In addition, the model provides a simple way of assessing the stability of a system by using the ratio of the log-returns of sovereign bonds and the stocks of major sovereign debt holders as a stability indicator. We also propose a systemic importance ranking, BankRank, for the financial institutions analyzed in this study to assess the contribution of individual banks to the overall systemic risk.

Resilience of Indian banks: Macroeconomic stress test modeling for credit risk

The macro stress testing is an integral part of the credit risk management exercise to judge the strength of banks in hypothetical stress situations. The present paper develops the macroeconomic stress testing model and checks the resilience of banks across different groups of banks in India like the public sector, private, and foreign banks in maintaining the minimum required regulatory capital at the time of stress situations. The stress testing exercise is undertaken by the use of panel data models to evaluate the impact of likely changes in macro parameters on the non‐performing loans under three stress scenarios like baseline, medium, and severe. The impact of these stress testing is substantial for the public sector banks as compared to private and foreign banks. Except for the few banks, all other public sector banks are not able to sustain the macro stress scenarios and became insolvent. In opposite to this, all the foreign banks are able to sustain all the stress scenarios. The private banks are also able to withstand the assumed crisis, except few banks failed to do so. The study may be useful for studying the preparedness of banks to face the crisis period and also helps them to get themselves equipped with the required capital to meet those circumstances.

A FAVAR modeling approach to credit risk stress testing and its application to the Hong Kong banking industry

In October 2018, the Basel Committee on Banking Supervision (BCBS) published its stress testing principles. One of these principles is about stress testing model validation, aided by business interpretation, benchmark comparison and backtesting. In this paper, a credit risk stress testing model based on the factor-augmented vector autoregressive (FAVAR) approach is proposed to project credit risk loss under stressed scenarios. Inherited from both factor analysis (FA) and the vector autoregressive (VAR) model, the FAVAR approach ensures that the proposed model has many appealing features. First, a large number of model input variables can be reduced to a handful of latent common factors to avoid the curse of dimensionality. Second, the dynamic interrelationship among macroeconomic variables and credit risk loss measures can be studied without exogeneity assumptions. Moreover, the application of the impulse response function facilitates the multiperiod projection of credit risk loss in response to macroeconomic shocks. All of these features make the proposed modeling framework a potentially handy solution to fulfilling the BCBS requirement of quantitative adequacy assessment of banks’ internal stress testing results with a benchmark model. The scope of its application can also extend to impairment modeling for International Financial Reporting Standard 9, which requires the projection of credit risk losses over consecutive periods under different macroeconomic scenarios.

Economic and Statistical Analysis of Sustainability Indicators of Banking Sectors of the Leading Countries

The last decade was characterized by the improvement of the financial systems’ monitoring mechanism at the national and international levels aimed at making effective decisions on financial stabilisation. The purpose of this study is to assess the sustainability of banking sectors of the leading economically developed and developing countries for the period 2009-2018, taking into account the processes occurring in the global economy and international financial markets.The research applied the International Monetary Fund methodology for assessing the stability of depository institutions and used the IMF database containing financial soundness indicators of the banking sectors in the studied countries. The economic and statistical analysis of financial soundness indicators of the banking sectors made it possible to explain the influence of economic conditions on the level and dynamics of the banks’ performance indicators as well as to identify the strengths and problems of the banking systems.The results of the study indicate an increase in the sustainability of the developed countries’ banking sectors, which is confirmed by the growing trends of capital adequacy and the quality of bank loan portfolios. The main problem of the developed countries is stably low profitability which, in the long term, could lead to a decrease in the banks’ capitalisation and their inability to maintain economic growth.A positive trend in the banking sectors of the developing countries is the recovery of capital adequacy, whereas the negative trend is a decrease in the quality of loan portfolios and profitability indicators. High credit risks and insufficient capitalisation represent the vulnerabilities of the developing countries’ banking sectors, while the increased volatility of financial soundness indicators, especially liquidity, is caused by the impact of the external trade and financial conditions.The results of this research could be used by analysts and regulators in macroeconomic calculations and when developing supervisory stress testing models, as well as by bank managers for performing internal stress tests and strategic business planning.

COVID-19 Implications for Banks: the Case of An Emerging Economy

The COVID-19 pandemic is damaging economies across the world including financial markets and institutions in all possible dimensions. Particularly for banks, the pandemic generates multifaceted crisis, mostly through increases in default rates. This is likely to be worse in developing economies with poor financial market architecture. As a case of emerging economies, this paper considers Bangladesh and examines the possible impacts of the pandemic on the country’s banking sector. Bangladesh’s banking sector already has a high level of non-performing loans (NPLs) and the pandemic is likely to worsen the situation. Using a state-designed stress testing model, the paper estimates the impacts of COVID-19 pandemic on three particular dimensions – firm value, capital adequacy, and interest income under different NPL shock scenarios. Findings suggest that all banks are likely to see a fall in risk-weighted asset values, capital adequacy ratio, and interest income at the individual bank and sectoral levels. However, estimates show that larger banks are relatively more vulnerable. The decline in all three dimensions will be disproportionately larger if NPL shocks become larger. Findings further show that a 10% NPL shock could force capital adequacy of all banks to go below the minimum BASEL-III requirement, while a 13% or more shock could turn it to zero or negative at the sectoral level. Findings call for an immediate and innovative policy measures to prevent a large-scale and contagious banking crisis in Bangladesh. The paper offers lessons for other developing and emerging economies similar to Bangladesh.

Macroprudential Liquidity Stress Test: An Application to Indonesian Banks

Abstract This paper develops a macroprudential liquidity stress test model for Indonesian banks. Our model incorporates two factors driving liquidity runs: (i) idiosyncratic factors; and (ii) macroeconomic factors. We estimate this model using a sample of 113 banks over the period of January 2011 to June 2018, and dynamic panel data estimators. We establish significant transmission channels from macroeconomic and idiosyncratic (bank idiosyncratic risks) factors to liquidity runs. By using the macroeconomic scenario transmission, we find the liquidity stress test to be more consistent with the solvency stress test.

A Liquidity Risk Stress-Testing Framework with Basel Liquidity Standards

We present a macro stress-testing model for banks' market and funding liquidity risks with a survival period of one year. The model follows the main principles of the Basel standards LCR and NSFR. Besides, the model takes into account the impact of both bank-specific and market-wide scenarios and includes second- round effects of shocks due to banks' feedback reactions. The presented methodology is then applied to a sample of Czech banks. This allows us to monitor the sensitivity of their liquidity position to the combination of shocks under consideration.

Benchmarking operational risk stress testing models

Benchmarking operational risk stress testing models

Automated Analysis of Microscale Cardiac Contractile Dynamics from Intravital Multiphoton Microscopy in Mouse Ventricle

BackgroundMultiphoton microscopy (MPM) has enabled in vivo time‐lapse imaging of the heart that shows motion of cells within the tissue with micrometer resolution. We developed automated analysis techniques to quantify cellular motion from in vivo cardiac MPM images throughout the cardiac cycle.MethodsIntravital cardiac MPM was performed on ventilated, male and female, 4–6 month‐old, C57Bl6 mice (n=9). MPM image volumes (100 μm deep) of the beating mouse left ventricle were acquired at 30 frames per second while recording the electrocardiogram and respiratory pressure (Fig. a). Image volumes were reconstructed by assembling lines acquired nearest to a specified point in the cardio‐respiratory phase space. Bulk‐tissue motion was calculated from the three‐dimensional translational transformation that best aligned images at each point of the cardiac cycle to a reference image at the most stable portion of the cardio‐respiratory phase space (respiratory 50%, cardiac 80%). After bulk‐tissue motion was subtracted, local tissue deformation was extracted via non‐rigid registration of the reconstructed images to the reference image and divergence was calculated from the resulting vector field. Analysis methods were applied to images before and after epinephrine injection (50 μg intramuscular), and following focal sterile injury by laser irradiation (FSI).ResultsVolumes were reconstructed in 50 intervals across the cardiac cycle that show vasculature (intravenous Texas‐red dextran, red) and cardiomyocytes and leukocytes (rhodamine 6G, cyan) moving across the field of view (Fig. b–c). Automated analysis of bulk‐tissue motion (Fig. d–e) indicated maximum displacements occurring at 16 % (anterior‐posterior direction), 36 % (superior‐inferior) and 38 % (epi‐endocardial) of the cardiac cycle. The calculated epi‐endocardial motion amplitude correlated with stroke volume (R2=0.35) measured by echocardiography in the same animal one day before MPM imaging. Automated analysis of local tissue deformation indicated a maximum absolute deformation occurring at 10 % of the cardiac cycle (Fig. f). Comparison to manual feature tracking across the cardiac cycle validated the automated measurements. Rigid and non‐rigid motion analysis techniques were applied to epinephrine stress test and focal sterile injury (FSI) lesion models showing, respectively, increased and decreased bulk‐tissue and local deformation and divergence.ConclusionsWe have shown and applied novel, fast, and accurate techniques for characterizing cardiac motion from in vivo cardiac MPM to study the performance of contractile cells in the heart during health and disease.Support or Funding InformationAmerican Heart Association (17POST33680127), National Institutes of Health (5R21EB024694‐03), and National Science Foundation (DBI‐1707312).Figure 1

The accuracy of alternative supervisory methodologies for the stress testing of credit risk

Following the financial crisis approaches such as regulatory capital ratios have been supplanted by stress testing as a primary tool for managing systemic risks. Financial institutions are mandated to perform stress testing to forecast performance over stress scenarios. While in parallel supervisors conduct their tress tests to set minimum regulatory capital, nothing is revealed regarding the accuracy of such models. We investigate a modelling framework rather close to that of the regulators, projecting financial statement line items for an aggregated 'average' bank. We assess the accuracy of alternative stress testing models, simple single equation versus complex multiple equation approaches. Results show inaccuracies in stress test model forecasts for multi-equation models that do not account for the dependency structure amongst the model variables. This highlights the public policy need for reconsidering the existent regulatory stress tests, and the need for supervisory models to be subject to model validation standards.

Banking Euro Area Stress Test Model

Banking Euro Area Stress Test Model

Model and estimation risk in credit risk stress tests

This paper deals with stress tests for credit risk and shows how exploiting the discretion when setting up and implementing a model can drive the results of a quantitative stress test for default probabilities. For this purpose, we employ several variations of a CreditPortfolioView-style model using US data ranging from 2004 to 2016. We show that seemingly only slightly differing specifications can lead to entirely different stress test results—in relative and absolute terms. That said, our findings reveal that the conversion of a shock (i.e., stress event) increases the (non-stress) default probability by 20–80%—depending on the stress test model selected. Interestingly, forecasts for non-stress default probabilities are less exposed to model and estimation risk. In addition, the risk horizon over which the stress default probabilities are forecasted and whether we consider mean stress default probabilities or quantiles seem to play only a minor role for the dispersion between the results of the different model specifications. Our findings emphasize the importance of extensive robustness checks for model-based credit risk stress tests.