HURLIN Christophe

Commodities play an essential role in our economies, and futures markets play a central role in determining their prices. The purpose of this thesis is to contribute to our understanding of the behavior of commodity prices, and to produce forecasts based on recent econometric methods. For forecasting, we focus on two different topics for three commodities (oil, wheat, and gold): forecasting prices at a monthly horizon from a large database, and forecasting volatility at a daily horizon using a recent variable selection procedure for conditional volatility. For the explanation, we focus on informational efficiency and information discovery in two different settings: predictive regressions using data on different theories, and an analysis of the effect of changes in open positions of different groups of traders on volatility.

In credit markets, screening algorithms discriminate between good-type and bad-type borrowers. This is their raison d’être. However, by doing so, they also often discriminate between individuals sharing a protected attribute (e.g. gender, age, race) and the rest of the population. In this paper, we show how to test (1) whether there exists a statistical significant difference in terms of rejection rates or interest rates, called lack of fairness, between protected and unprotected groups and (2) whether this difference is only due to credit worthiness. When condition (2) is not met, the screening algorithm does not comply with the fair-lending principle and can be qualified as illegal. Our framework provides guidance on how algorithmic fairness can be monitored by lenders, controlled by their regulators, and improved for the benefit of protected groups.

No summary available.

In the context of credit scoring, ensemble methods based on decision trees, such as the random forest method, provide better classification performance than standard logistic regression models. However, logistic regression remains the benchmark in the credit risk industry mainly because the lack of interpretability of ensemble methods is incompatible with the requirements of financial regulators. In this paper, we pro- pose to obtain the best of both worlds by introducing a high-performance and interpretable credit scoring method called penalised logistic tree regression (PLTR), which uses information from decision trees to improve the performance of logistic regression. Formally, rules extracted from various short-depth decision trees built with pairs of predictive variables are used as predictors in a penalised logistic regression model. PLTR allows us to capture non-linear effects that can arise in credit scoring data while preserving the intrinsic interpretability of the logistic regression model. Monte Carlo simulations and empirical applications using four real credit default datasets show that PLTR predicts credit risk significantly more accurately than logistic regression and compares competitively to the random forest method. JEL Classification: G10 C25, C53.

This PhD thesis includes three essays on the implementation, and where appropriate the improvement, of financial risk measures and bank risk assessment based on machine learning methods. The first chapter develops an elementary formula, called E2C, for estimating credit risk premiums inspired by CreditGrades, and improves its accuracy with a decision tree forest algorithm. Our results highlight the prominent role played by this estimator and the additional contribution of the financial rating and the size of the firm considered. The second chapter infers a one-sided version of the inequality bounding the probability of a unimodally distributed random variable. Our results show that the unimodality assumption for stock returns is generally admissible, allowing us to refine bounds on individual risk measures, to comment on the implications for extreme risk multipliers, and to derive simplified versions of bounds on systemic risk measures. The third chapter provides a decision support tool grouping rated banks by risk level based on an adjusted version of the k-means algorithm. This fully automated process is based on a very large universe of individual and systemic risk indicators synthesized into a subset of representative factors. The results obtained are aggregated by country and region, offering the possibility to study areas of fragility. They highlight the importance of paying particular attention to the ambiguous impact of bank size on systemic risk measures.

This paper proposes an original approach for backtesting systemic risk measures. This backtesting approach makes it possible to assess the systemic risk measure forecasts used to identify the financial institutions that contribute the most to the overall risk in the financial system. Our procedure is based on simple tests similar to those generally used to backtest the standard market risk measures such as value-at-risk or expected shortfall. We introduce a concept of violation associated with the marginal expected shortfall (MES), and we define unconditional coverage and independence tests for these violations. We can generalize these tests to any MES-based systemic risk measures such as the systemic expected shortfall (SES), the systemic risk measure (SRISK), or the delta conditional value-at-risk ([Formula: see text]CoVaR). We study their asymptotic properties in the presence of estimation risk and investigate their finite sample performance via Monte Carlo simulations. An empirical application to a panel of U.S. financial institutions is conducted to assess the validity of MES, SRISK, and [Formula: see text]CoVaR forecasts issued from a bivariate GARCH model with a dynamic conditional correlation structure. Our results show that this model provides valid forecasts for MES and SRISK when considering a medium-term horizon. Finally, we propose an early warning system indicator for future systemic crises deduced from these backtests. Our indicator quantifies how much is the measurement error issued by a systemic risk forecast at a given point in time which can serve for the early detection of global market reversals. This paper was accepted by Kay Giesecke, finance.

Decision trees and related ensemble methods like random forest are state-of-the-art tools in the field of machine learning for credit scoring. Although they are shown to outperform logistic regression, they lack interpretability and this drastically reduces their use in the credit risk management industry, where decision-makers and regulators need transparent score functions. This paper proposes to get the best of both worlds, introducing a new, simple and interpretable credit scoring method which uses information from decision trees to improve the performance of logistic regression. Formally, rules extracted from various short-depth decision trees built with couples of predictive variables are used as predictors in a penalized or regularized logistic regression. By modeling such univariate and bivariate threshold effects, we achieve significant improvement in model performance for the logistic regression while preserving its simple interpretation. Applications using simulated and four real credit defaults datasets show that our new method outperforms traditional logistic regressions. Moreover, it compares competitively to random forest, while providing an interpretable scoring function. JEL Classification: G10 C25, C53.

This paper thesis is composed of four self-contained chapters, contributing to the field of nonlinear econometrics. The first chapter focuses on the contribution of nonlinear econometrics through the measurement of financial performance using a dichotomous variable as the independent variable. The next three chapters are based on nonlinear regression models where the dichotomous variable is the dependent variable of the equation. Given the links between financial risk and the macroeconomic context, this part is linked to the theme of optimal allocation via the study of crises and recessions. This class of model (probit / logit) is used in the second chapter to study empirically the role of financial development in the probability of occurrence of banking crises. Then the last two chapters focus on the methodological framework developed by Kauppi and Saikkonen (2008) and Candelon, Dumitrescu and Hurlin (2012 . 2014) about forecasting business cycles from probit / logit models. Thus, the third chapter studies the empirical relationship linking the evolution of the credit spread and the future probability of expansion/recession in an extended data panel while testing the homogeneity of this relationship. Finally, the fourth chapter proposes a theoretical contribution by deriving the response functions of probit / logit models from the approach of Kauppi and Saikkonen (2008). These response functions are then used in an empirical framework to estimate the impact of an exogenous shock on the expansion/recession cycle.

We derive several popular systemic risk measures in a common framework and show that they can be expressed as transformations of market risk measures (e.g. beta). We also derive conditions under which the different measures lead to similar rankings of systemically important financial institutions (SIFIs). In an empirical analysis of US financial institutions, we show that (1) different systemic risk measures identify different SIFIs and that (2) firm rankings based on systemic risk estimates mirror rankings obtained by sorting firms on market risk or liabilities. One-factor linear models explain most of the variability of the systemic risk estimates, which indicates that systemic risk measures fall short in capturing the multiple facets of systemic risk.

In this paper, we identify several shortcomings in the systemic-risk scoring methodology currently used to identify and regulate Systemically Important Financial Institutions (SIFIs). Using newly-disclosed regulatory data for 119 US and international banks, we show that the current scoring methodology severely distorts the allocation of regulatory capital among banks. We then propose and implement a methodology that corrects for these shortcomings and increases incentives for banks to reduce their risk contributions.

This thesis deals with the evaluation and validation of law forecasts. In the first part, we are interested in the contribution of machine learning to quantile and law predictions. To do so, we have tested different machine learning algorithms in a framework of quantile forecasts on real data. We thus try to highlight the interest of certain methods according to the type of data we are confronted with. In the second part, we present some validation tests of law predictions present in the literature. Some of these tests are then applied on real data related to the log-returns of stock market indices. In the third part, we propose a recalibration method allowing to simplify the choice of a particular density forecast compared to others. This recalibration allows us to obtain valid forecasts from a misspecified model. We also highlight conditions under which the quality of recalibrated forecasts, as assessed using CRPS, is systematically improved, or very slightly degraded. These results are illustrated through applications to temperature and price scenarios.

In this article, we propose a reflection on the contribution of Machine Learning techniques and New Data to credit risk modeling. Credit scoring was historically one of the first fields of application of Machine Learning techniques. Today, these techniques make it possible to exploit "new" data made available by the digitalization of customer relations and social networks. The combination of the emergence of new methodologies and new data has thus structurally modified the credit industry and favored the emergence of new players. First, we analyze the contribution of Machine Learning algorithms with a constant information set. We show that there are productivity gains linked to these new approaches but that the gains in credit risk prediction remain modest. Second, we evaluate the contribution of this "datadiversity", whether or not these new data are exploited by Machine Learning techniques. It turns out that some of these data reveal weak signals that significantly improve the quality of the assessment of borrowers' creditworthiness. At the micro level, these new approaches promote financial inclusion and access to credit for the most fragile borrowers. However, Machine Learning applied to these data can also lead to bias and discrimination.

The research work carried out is mainly in the field of banking, monetary and financial economics and is organized around three empirical perspectives. The first perspective is microeconomic and is located at the level of the banking firm. The research conducted analyzes the provisioning behavior of banks for credit losses and evaluates the risk profiles of banks. The second perspective is macroeconomic and is situated at the country level. The research focuses on the procyclicality of credit activities on the one hand, and the occurrence of banking crises on the other. The third perspective introduces an international dimension and thus complements the microeconomic and macroeconomic approaches previously used. The research conducted analyzes the effects of banking regulation and the consequences of the 2007-2008 financial crisis on international banking and financial activities.

This thesis studies income disparities from two different and complementary perspectives: (i) between groups of individuals sharing time periods and (ii) between groups of individuals sharing the same territory. Our objective is to analyze the economic mechanisms that generate both temporal and spatial income disparities and to study the means available to remedy them. The thesis is divided into three chapters in which we mainly use statistical tools on the French case in order to answer the problems raised. The aspect of temporal income disparities is the subject of the first chapter where we study the interaction between fiscal multipliers and public debt insofar as they are the two main tools for smoothing income over time and present both short and long term issues. More specifically, we seek to identify the best instruments and the "best" economic context to initiate a reduction in the public debt ratio. We highlight that the use of fiscal austerity policies during economic recessions is likely to increase the public debt ratio. The second chapter is devoted to the measurement and characterization of the convergence phenomenon between French territories. In particular, we answer the following questions: (i) is there a convergence phenomenon? (ii) is the convergence process linear? (iii) is the convergence uniform across the territory? We answer these three questions in the affirmative, which raises questions about the reasons for spatial heterogeneity in convergence speeds. Finally, the third chapter aims at identifying the determinants of spatial convergence. In particular, we wish to highlight the factors that encourage spatial convergence and, on the contrary, those that slow it down. We then show that the heterogeneity of territories in terms of access to different external markets is an important factor in explaining the different levels of convergence observed across France.

No summary available.

No summary available.

No summary available.

Reproducibility is key for building trust in research, yet it is not widespread in economics. We show how external certification can improve reproducibility in economics research. Such certification can be conducted by a trusted third party or agency, which formally tests whether a given result is indeed generated by the code and data used by a researcher. This additional validation step significantly enriches the peer-review process, without adding an extra burden on journals or unduly lengthening the publication process. We show that external certification can accommodate research based on confidential data. Lastly, we present an actual example of external certification.

No summary available.

We propose a new approach for comparing Loss Given Default (LGD) models which is based on loss functions defined in terms of regulatory capital charge. Our comparison method improves the banks' ability to absorb their unexpected credit losses, by penalizing more heavily LGD forecast errors made on credits associated with high exposure and long maturity. We also introduce asymmetric loss functions that only penalize the LGD forecast errors that lead to underestimate the regulatory capital. We show theoretically that our approach ranks models differently compared to the traditional approach which only focuses on LGD forecast errors. We apply our methodology to six competing LGD models using a unique sample of almost 10,000 defaulted credit and leasing contracts provided by an international bank. Our empirical findings clearly show that model rankings based on capital charge losses differ drastically from those based on the LGD loss functions currently used by regulators, banks, and academics.

We propose a new approach for comparing Loss Given Default (LGD) models which is based on loss functions defined in terms of regulatory capital charge. Our comparison method improves the banks' ability to absorb their unexpected credit losses, by penalizing more heavily LGD forecast errors made on credits associated with high exposure and long maturity. We also introduce asymmetric loss functions that only penalize the LGD forecast errors that lead to underestimate the regulatory capital. We show theoretically that our approach ranks models differently compared to the traditional approach which only focuses on LGD forecast errors. We apply our methodology to six competing LGD models using a sample of almost 10,000 defaulted credit and leasing contracts provided by an international bank. Our empirical findings clearly show that models' rankings based on capital charge losses differ from those based on the LGD loss functions currently used by regulators, banks, and academics.

No summary available.

This PhD thesis is composed of three chapters, one article and two papers and is mainly related to the field of empirical financial econometrics. It analyzes the dependence and the link between the financial markets and the commodity markets, in particular the energy market. The distributions and correlations of the variables belonging to both markets are studied in order to determine their effects on each other and to analyze their trends to give a better insight into their behavior with respect to crises and abrupt events in the economy. These variables are represented by some financial indices (SP500, Euro stoxx 50, Msci China) as well as by the main commodity indices (SP GSCI, Brent Oil, Natural Gas, Precious Metals). We choose to model their correlation over time and to take into account the non-linearity and instability that can affect them. For this purpose, the copula function approach has been used to model their distributions in an efficient way. In the first chapter, we examine the dependence and co-movements between carbon dioxide emission prices and energy indices such as coal, natural gas, Brent oil and the global energy index. The second chapter analyzes the interactions and relationships between the oil market and two major financial markets in Europe and the United States represented by the Euro stoxx 50 and the SP500. The last chapter analyzes the multivariate dependence between commodity indices of different sectors with financial indices using the Regular Vine copula model.

No summary available.

We identify several shortcomings in the systemic-risk scoring methodology currently used to identify and regulate Systemically Important Financial Institutions (SIFIs). Using newly-disclosed regulatory data for 119 US and international banks, we show that the current scoring methodology severely distorts the allocation of regulatory capital among banks. We then propose and implement a methodology that corrects for these short-comings and increases incentives for banks to reduce their risk contributions. Unlike the current scores, our adjusted scores are mainly driven by risk indicators directly under the control of the regulated bank and not by factors that are exogenous to the bank, such as exchange rates or other banks' actions.

This thesis addresses two new debates on central banking that emerged after the 2008 financial crisis: the debate on financial losses on central banks' balance sheets, and the debate on the high level of bank rates relative to market rates after the crisis. The first two chapters are part of the first debate. The link between the financial soundness of central banks and inflation is studied empirically in the first chapter, using a large panel of 82 countries. Theoretically, this link is potentially present when the government does not financially support the central bank and the latter can therefore only rely on itself to improve its financial situation. The results of the first chapter show that in practice this is indeed the case: deteriorations in central bank balance sheets are accompanied by higher inflation when the central bank has no fiscal support. The results do not show a link in a general context, as the theory suggests. The second chapter analyzes and conceptualizes the argument that a central bank may end a fixed or quasi-fixed exchange rate regime out of fear of future financial losses. The analysis is then applied to the case of the floor rate implemented by the Swiss Central Bank (SNB) between 2011 and 2015 vis-à-vis the euro. This argument has been put forward by many to explain the end of the floor rate policy in Switzerland, without any research prior to this one assessing its relevance. The empirical estimates in Chapter 2 show that this argument had credibility: they show that in credible scenarios, breaking the peg with the euro 17 months later, the SNB would have incurred a considerable loss, exceeding a threshold perceived as limiting by many central bankers. The last chapter of this thesis focuses on the spread between deposit rates and the market rate in the eurozone (EURIBOR), which became significantly positive after the crisis, leading some to speak of the "over-remuneration" of deposits. This chapter argues that the majority of this gap is not explained by abnormal deposit behavior, as some have argued, but rather by a loss of relevance of EURIBOR. Constructing an alternative to EURIBOR, this chapter concludes that bank risk has had a primary influence on the level of deposit remuneration in the post-crisis world.

We propose a bootstrap-based test of the null hypothesis of equality of two firms? conditional Risk Measures (RMs) at a single point in time. The test can be applied to a wide class of conditional risk measures issued from parametric or semi-parametric models. Our iterative testing procedure produces a grouped ranking of the RMs, which has direct application for systemic risk analysis. Firms within a group are statistically indistinguishable form each other, but significantly more risky than the firms belonging to lower ranked groups. A Monte Carlo simulation demonstrates that our test has good size and power properties. We apply the procedure to a sample of 94 U.S. financial institutions using ?CoVaR, MES, and %SRISK. We find that for some periods and RMs, we cannot statistically distinguish the 40 most risky firms due to estimation uncertainty.

This thesis seeks to link business cycles and portfolio management. The first chapter builds a theoretical framework between business cycles and risk premia. It highlights the importance of turning points in the growth cycle, better known as the output gap. The next two chapters aim to detect these turning points in real time. The first approach focuses on a simple and easily understood non-parametric machine learning method called adaptive vector quantization. The second approach uses more complex machine learning methods, known as ensemble learning: random forests and boosting. Both approaches allow us to create efficient investment strategies in real time. Finally, the last chapter develops an asset allocation method based on different hierarchical clustering algorithms. Empirical results demonstrate the interest of this attempt: the created portfolios are robust, diversified and lucrative.

The paper tests to what extent the intensity of the relationship between a firm and a financial analyst improves or degrades the accuracy of the forecasts produced by this analyst on this firm. Using a sample of earnings per share (EPS) forecasts for 208 French firms, we regress the analysts' forecast error on a set of observable variables. We then decompose the fixed effect of the regression and use the firm-analyst pair effect as a measure of the intensity of the relationship. We show that a small (large) pair effect is associated with a small (large) forecast error, suggesting that a close relationship between a firm and an analyst tends to bias the analyst's forecast. Experienced analysts who specialize in tracking large-cap firms, however, seem less prone to this bias.

In this paper we study various MIDAS models for which the future daily variance is directly related to past observations of intraday predictors. Our goal is to determine if there exists an optimal sampling frequency in terms of variance prediction. Via Monte Carlo simulations we show that in a world without microstructure noise, the best model is the one using the highest available frequency for the predictors. However, in the presence of microstructure noise, the use of very high-frequency predictors may be problematic, leading to poor variance forecasts. The empirical application focuses on two highly liquid assets (i.e., Microsoft and S&P 500). We show that, when using raw intraday squared log-returns for the explanatory variable, there is a “high-frequency wall” – or frequency limit – above which MIDAS-RV forecasts deteriorate or stop improving. An improvement can be obtained when using intraday squared log-returns sampled at a higher frequency, provided they are pre-filtered to account for the presence of jumps, intraday diurnal pattern and/or microstructure noise. Finally, we compare the MIDAS model to other competing variance models including GARCH, GAS, HAR-RV and HAR-RV-J models. We find that the MIDAS model – when it is applied on filtered data –provides equivalent or even better variance forecasts than these models. JEL: C22, C53, G12 / KEY WORDS: Variance Forecasting, MIDAS, High-Frequency Data. RÉSUMÉ.

This doctoral thesis, composed of three chapters, contributes to the development of the problematic on the selection of GARCH-type volatility models. The first chapter proposes a simulation study on model selection in the specific framework of regime-switching models. Simulation experiments are proposed to highlight the inefficiency of the usual selection criteria in particular cases, which can lead to misspecification during model selection. The second chapter proposes a test of the Lagrange multiplier of misspecification in univariate GARCH models. The null hypothesis assumes that the data generating process is a linear GARCH model while under the alternative hypothesis it corresponds to an unknown functional form that is linearized using a Taylor expansion. The test is illustrated in an empirical application on exchange rates. The last chapter studies the impact of oil prices on the sovereign credit default swap spreads of two oil exporting countries: Venezuela and Russia. Using recent data, we find that oil price returns impact Venezuela's sovereign CDS spreads directly while it goes through the exchange rate channel for Russia. This chapter employs advanced statistical methods, including the use of Markov regime-switching models. Finally, the appendix provides a manual for the MSGtool (Matlab) toolbox which provides a collection of functions for studying Markovian regime-switching models. The toolbox is very user-friendly.

We review the extensive literature on systemic risk and connect it to the current regulatory debate. While we take stock of the achievements of this rapidly growing field, we identify a gap between two main approaches. The first one studies different sources of systemic risk in isolation, uses confidential data, and inspires targeted but complex regulatory tools. The second approach uses market data to produce global measures which are not directly connected to any particular theory, but could support a more efficient regulation. Bridging this gap will require encompassing theoretical models and improved data disclosure.

No summary available.

No summary available.

No summary available.

The complexity of financial markets and the resurgence of particularly severe crises are contributing to the evolution and questioning of so-called standard econometric models for explaining and forecasting financial dynamics. The warning given jointly by prudential managers and researchers aims at encouraging the development of more complex, non-linear models largely inspired by other disciplines. We argue in this thesis that a chaos-stochastic approach to financial chronicles is likely to lead to better results. The relevance of this association is evaluated for market risk in two distinct analytical frameworks. We show the interest of a synthesis of chaotic models and GARCH specifications with or without Markovian regime shifts (MRS) for the modeling and forecasting of the Value-at-Risk of Eurozone stock indices. This study shows better results for the chaos-stochastic models and, in the case of the MRS-GARCH specifications, a better adequacy of the chaotic model of Lasota(1977) for the Southern European indices, which are particularly more volatile than those of Northern Europe for which we recommend the Mackey-Glass(1977) model. This combination allows us, in a bivariate framework, to better understand the links between the different stock markets of the euro zone. We introduce two new specifications that integrate the issues related to correlation breaks: the first one allows us to distinguish, through a sub-period analysis, the interdependence relationships from the contagion phenomena and the second one proposes, in a unified framework, to integrate the correlation breaks. This dual analysis highlights the driving role of the Franco-German index pair, the existence of two distinct spheres made up of Northern European indices on the one hand and Southern European countries on the other, and the intensification of certain relationships between indices following the sovereign debt crisis. We note and insist on the relevance of a chaotic model on average to account for part of the volatility wrongly attributed to GARCH effects.

This paper introduces a new class of models for the Value-at-Risk (VaR) and Expected Shortfall (ES), called the Dynamic AutoRegressive Expectiles (DARE) models. Our approach is based on a weighted average of expectile-based VaR and ES models, i.e. the Conditional Autoregressive Expectile (CARE) models introduced by Taylor (2008a) and Kuan et al. (2009). First, we briefly present the main non-parametric, parametric and semi-parametric estimation methods for VaR and ES. Secondly, we detail the DARE approach and show how the expectiles can be used to estimate quantile risk measures. Thirdly, we use various backtesting tests to compare the DARE approach to other traditional methods for computing VaR forecasts on the French stock market. Finally, we evaluate the impact of several conditional weighting functions and determine the optimal weights in order to dynamically select the more relevant global quantile model.

This article proposes intraday high-frequency risk (HFR) measures for market risk in the case of irregularly spaced high-frequency data. In this context, we distinguish three concepts of value-at-risk (VaR): the total VaR, the marginal (or per-time-unit) VaR and the instantaneous VaR. Since the market risk is obviously related to the duration between two consecutive trades, these measures are completed with a duration risk measure, i.e. the time-at-risk (TaR). We propose a forecasting procedure for VaR and TaR for each trade or other market microstructure event. Subsequently, we perform a backtesting procedure specifically designed to assess the validity of the VaR and TaR forecasts on irregularly spaced data. The performance of the HFR measure is illustrated in an empirical application for two stocks (Bank of America and Microsoft) and an exchange-traded fund based on Standard & Poor's 500 index. We show that the intraday HFR forecasts capture accurately the volatility and duration dynamics for these three assets. Copyright © 2015 John Wiley & Sons, Ltd.

We present CoMargin, a new methodology to estimate collateral requirements in derivatives central counterparties (CCPs). CoMargin depends on both the tail risk of a given market participant and its interdependence with other participants. Our approach internalizes trading externalities and enhances the stability of CCPs, thus, reducing systemic risk concerns. We assess our methodology using proprietary data from the Canadian Derivatives Clearing Corporation that include daily observations of the actual trading positions of all of its members from 2003 to 2011. We show that CoMargin outperforms existing margining systems by stabilizing the probability and minimizing the shortfall of simultaneous margin-exceeding losses.

We propose a bootstrap-based test of the null hypothesis of equality of two firms' conditional Risk Measures (RMs) at a single point in time. The test can be applied to a wide class of conditional risk measures issued from parametric or semi-parametric models. Our iterative testing procedure produces a grouped ranking of the RMs which has direct application for systemic risk analysis. A Monte Carlo simulation demonstrates that our test has good size and power properties. We propose an application to a sample of U.S. financial institutions using CoVaR, MES, and SRISK, and conclude that only SRISK can be estimated with enough precision to allow for meaningful ranking.

This thesis consists of three empirical papers on central bank institutional design.Chapter 1 contributes to the debate on the importance of central bank independence (CBI) in lowering inflation rates. It stresses the relevance of employing indices of central bank independence computed dynamically in two ways. First, it recomputes the evolution of the Grilli et al. (1991) index of CBI and shows that the timing of large legislative reforms is closely related to inflation rate dynamics. Using unit root tests with endogenous structural breaks, I find that reforms that modify the degree of CBI represent structural breaks in the inflation rate dynamics. Second, employing the dynamic Grilli et al. (1991) index of independence confirms the negative relationship between CBI and inflation in a sample of 10 advanced economies.Chapter 2 presents a new and comprehensive database of central bank institutional design for 65 countries over the period 1972--2014. This chapter describes in detail the sources of information and the coding rules used to create a new index of central bank independence. It also compares this new index with the classical measures of CBI and highlights the new aspects of central bank institutional design included in this database such as financial independence and accountability. An important innovation of this new index is its dynamic nature. This enables an investigation of the endogenous determination of the level of independence of central banks and suggests several instruments for the CBI index. Using an instrumental variable approach, this chapter provides strong support for a causal, negative CBI-inflation nexus.Chapter 3 uses a political economy framework to investigate the drivers of reforms in central bank institutional design. Using the new CBI index developed in Chapter 2, this Chapter investigates the determinants of central bank reforms in a sample of 65 countries over the period 1972--2014. The results obtained suggest that the incentives generated by initial reforms which increased the level of independence, as well as a regional convergence, represent important drivers of reforms in central bank design. At the same time, an external pressure to reform, such as obtaining an IMF loan or joining a monetary union, also increases the likelihood of reforms, while government changes or crises episodes have little impact.

We show how to reverse-engineer banks' risk disclosures, such as Value-at-Risk, to obtain an implied measure of their exposures to equity, interest rate, foreign exchange, and commodity risks. Factor Implied Risk Exposures (FIRE) are obtained by breaking down a change in risk disclosure into a market volatility component and a bank-specific risk exposure component. In a study of large US and international banks, we show that (1) changes in risk exposures are negatively correlated with market volatility and (2) changes in risk exposures are positively correlated across banks, which is consistent with banks exhibiting commonality in trading.

Systemic risk played a key role in the spread of the last global financial crisis. many measures of systemic risk have been developed to assess the contribution of a financial institution to system-wide risk. However, many questions regarding the ability of these measures to identify systemically important financial institutions (SIFIs) have been raised since systemic risk has multiple facets and some of them are difficult to identify, such as similarities between financial institutions.The general objective of this thesis in finance is therefore (i) to propose an empirical solution to identify SIFIs at the national level, (ii) to compare theoretically and empirically different measures of systemic risk and (iii) to measure changes in banks' risk exposures.First, chapter 1 proposes an adjustment of three market-based measures of systemic risk designed in an international framework to identify SIFIs at the national level. Second, chapter 2 introduces a common model in which several measures of systemic risk are expressed and compared. It is theoretically established that these systemic risk measures can be expressed in terms of traditional risk measures. Empirical application confirms these results and shows that these measures are not able to capture the multidimensional nature of systemic risk. Finally, Chapter 3 presents the Factor Implied Risk Exposures (FIRE) methodology for decomposing a change in a bank's risk measure into two components, the first representing market volatility and the second representing the bank's risk exposure. This chapter empirically illustrates that changes in risk exposures are positively correlated across banks, which is consistent with the fact that banks have similarities in their market positions.

No summary available.

In this paper we study various MIDAS models in which the future daily variance is directly related to past observations of intraday predictors. Our goal is to determine if there exists an optimal sampling frequency in terms of volatility prediction. Via Monte Carlo simulations we show that in a world without microstructure noise, the best model is the one using the highest available frequency for the predictors. However, in the presence of microstructure noise, the use of ultra high-frequency predictors may be problematic, leading to poor volatility forecasts. In the application, we consider two highly liquid assets (i.e., Microsoft and S&P 500). We show that, when using raw intraday squared log-returns for the explanatory variable, there is a "high-frequency wall" or frequency limit above which MIDAS-RV forecasts deteriorate. We also show that an improvement can be obtained when using intraday squared log-returns sampled at a higher frequency, provided they are pre-filtered to account for the presence of jumps, intraday periodicity and/or microstructure noise. Finally, we compare the MIDAS model to other competing variance models including GARCH, GAS, HAR-RV and HAR-RV-J models. We find that the MIDAS model provides equivalent or even better variance forecasts than these models, when it is applied on filtered data.

Traditionally, financial crisis Early Warning Systems (EWSs) have relied on macroeconomic leading indicators when forecasting the occurrence of such events. This paper extends such discrete-choice EWSs by taking the persistence of the crisis phenomenon into account. The dynamic logit EWS is estimated using an exact maximum likelihood estimation method in both a country-by-country and a panel framework. The forecasting abilities of this model are then scrutinized using an evaluation methodology which was designed recently, specifically for EWSs. When used for predicting currency crises for 16 countries, this new EWS turns out to exhibit significantly better predictive abilities than the existing static one, both in- and out-of-sample, thus supporting the use of dynamic specifications for EWSs for financial crises.

No summary available.

As most Exchange-Traded Funds (ETFs) engage in securities lending or are based on total return swaps, they expose their investors to counterparty risk. In this paper, we estimate empirically such risk exposures for a sample of physical and swap-based funds. We find that counterparty risk exposure is higher for swap-based ETFs, but that investors are compensated for bearing this risk. Using a difference-in-differences specification, we uncover that ETF flows respond significantly to changes in counter-party risk. Finally, we show that switching to an optimal collateral portfolio leads to substantial reduction in counterparty risk exposure.

We derive several popular systemic risk measures in a common framework and show that they can be expressed as transformations of market risk measures (e.g., beta). We also derive conditions under which the different measures lead to similar rankings of systemically important financial institutions (SIFIs). In an empirical analysis of US financial institutions, we show that (1) different systemic risk measures identify different SIFIs and that (2) firm rankings based on systemic risk estimates mirror rankings obtained by sorting firms on market risk or liabilities. One-factor linear models explain most of the variability of the systemic risk estimates, which indicates that systemic risk measures fall short in capturing the multiple facets of systemic risk.

No summary available.

No summary available.

We present an evaluation of the main empirical approaches used in the literature to estimate the contribution of public capital stock to growth and private factors' productivity. Based on a simple stochastic general equilibrium model, built as to reproduce the main long-run relations observed in US post-war historical data, we show that the production function approach may not be reliable to estimate this contribution. Our analysis reveals that this approach largely overestimates the public capital elasticity, given the presence of a common stochastic trend shared by all non-stationary inputs.

This paper proposes intraday High Frequency Risk (HFR) measures for market risk in the case of irregularly spaced high-frequency data. In this context, we distinguish three concepts of value-at-risk (VaR): the total VaR, the marginal (or per-time-unit) VaR, and the instantaneous VaR. Since the market risk is obviously related to the duration between two consecutive trades, these measures are completed with a duration risk measure, i.e., the time-at-risk (TaR). We propose a forecasting procedure for VaR and TaR for each trade or other market microstructure event. We perform a backtesting procedure specifically designed to assess the validity of the VaR and TaR forecasts on irregularly spaced data. The performance of the HFR measure is illustrated in an empirical application for two stocks (Bank of America and Microsoft) and an exchange-traded fund (ETF) based on Standard and Poor's (the S&P) 500 index. We show that the intraday HFR forecasts accurately capture the volatility and duration dynamics for these three assets.

We identify a potential bias in the methodology disclosed in July 2013 by the Basel Committee on Banking Supervision (BCBS) for identifying systemically important financial banks. Contrary to the original objective, the relative importance of the five categories of risk importance (size, cross-jurisdictional activity, interconnectedness, substitutability/financial institution infrastructure, and complexity) may not be equal and the resulting systemic risk scores are mechanically dominated by the most volatile categories. In practice, this bias proved to be serious enough that the substitutability category had to be capped by the BCBS. We show that the bias can be removed by simply standardizing each input prior to computing the systemic risk scores.

We study whether financial analysts' concern for preserving good relationships with firms' managers motivates them to issue pessimistic or optimistic forecasts. Based on a dataset of one-yearahead EPS forecasts issued by 4 648 analysts concerning 241 French firms (1997-2007), we regress the analysts' forecast accuracy on its unintentional determinants. We then decompose the fixed effect of the regression and we use the firm-analyst pair effect as a measure of the intensity of the firm-analyst relationship. We find that a low (high) firm-analyst pair effect is associated with a low (high) forecast error. This observation suggests that pessimism and optimism result from the analysts' concern for cultivating their relationship with the firm's management.

In this paper, we identify several shortcomings in the systemic-risk scoring methodology currently used to identify and regulate Systemically Important Financial Institutions (SIFIs). Using newly-disclosed regulatory data for 119 US and international banks, we show that the current scoring methodology severely distorts the allocation of regulatory capital among banks. We then propose and implement a methodology that corrects for these shortcomings and increases incentives for banks to reduce their risk contributions.

The interest of non-linear models lies, on the one hand, in a better consideration of the non-linearities characterizing macroeconomic and financial series and, on the other hand, in a more information-rich forecast.At this level, the originality of the intervals (asymmetric and/or discontinuous) and the forecasting densities (asymmetric and/or multimodal) offered by this new form of modelling suggests that an improvement in forecasting relative to linear models is possible, and that sufficiently powerful evaluation tests are needed to verify this possible improvement. These tests generally consist in checking distributional hypotheses on the violation processes and the probabilistic transforms associated with each of these forms of forecasting. In this thesis, we have adapted the GMM framework based on orthonormal polynomials designed byBontemps and Meddahi (2005, 2012) to test the fit to certain probability laws, an approach already initiated by Candelon et al. (2011) in the context of Value-at-Risk assessment. In addition to the simplicity and robustness of the method, the tests developed have good properties in terms of size and power. The use of our new approach in the comparison of linear and non-linear models in an empirical analysis confirmed the idea that the former are preferred if the objective is the calculation of simple point forecasts while the latter are the most appropriate to account for the uncertainty around them.

For several decades, energy prices have been subject to increasing volatility, which has weighed heavily on the entire economy. Compared to the prices of other commodities (such as precious metals or agricultural products), the evolution of energy products has appeared exceptionally uncertain, both in the long and short term. In a global economic context, this phenomenon acquires all its importance as the damage to the real economy of a strong variation in commodity prices can be significant. This thesis therefore focuses on the underlying causes of these fluctuations. More specifically, by uniting the different fields of energy economics, econometrics, finance and psychology, it seeks to understand the phenomenon of commodity financialization and the close relationship between financial markets and commodity markets. This reflection is articulated in three themes: on the one hand, the relationship between the prices of different energies and their financial properties is analyzed, on the other hand, the emotional and behavioral aspects of the markets are studied, and finally, the direct links between stock markets and commodity markets are addressed.

We present an evaluation of the main empirical approaches used in the literature to estimate the contribution of public capital stock to growth and private factors' productivity. Based on a simple stochastic general equilibrium model, built as to reproduce the main long-run relations observed in US post-war historical data, we show that the production function approach may not be reliable to estimate this contribution. Our analysis reveals that this approach largely overestimates the public capital elasticity, given the presence of a common stochastic trend shared by all non-stationary inputs.

No summary available.

No summary available.

No summary available.

Known as Early Warning Systems (EWS), financial crisis prediction models are called upon to play a decisive role in the orientation of economic policies at both the microeconomic and macroeconomic levels. However, in the wake of the global financial crisis, major questions are being raised about their real predictive capacity. This applied econometrics thesis aims at proposing (i) a method for systematically evaluating the predictive capabilities of EWS and (ii) new EWS specifications to improve their performance. This work is divided into four chapters. The first one proposes an original test for evaluating predictions by confidence intervals based on the assumption of binomial distribution of the violation process. The second chapter proposes an econometric evaluation strategy of the predictive capabilities of EWS. We show that this evaluation should be based on the determination of an optimal threshold on the predicted probabilities of crisis occurrence as well as on the comparison of models.The third chapter reveals that the dynamics of crises (persistence) is an essential element of the econometric specification of EWS. The results show in particular that dynamic logit models have much better predictive capabilities than static models and Markovian models. Finally, in the fourth chapter we propose an original multivariate dynamic probit model that allows us to analyze the causality patterns between different types of crises (banking, exchange rate and debt). The empirical illustration clearly shows that the switch to trivariate modeling significantly improves the forecasts for countries experiencing all three types of crises.

This thesis analyzes convergence within a monetary union through the dynamics of prices and equilibrium exchange rates. In the first chapter we present the general characteristics of currency areas, as well as the history of the ones we study: the EMU and the CFA zone. The second chapter deals with price convergence in the euro area using smooth transition models. Convergence is non-linear, and the speed of adjustment is different across countries. This is explained by differences in the evolution of price competitiveness, labor market rigidities, but also specialization patterns. The third chapter assesses the validity of absolute Purchasing Power Parity in EMU through unit root and cointegration tests in second and third generation panels. Overall, the price dynamics appear heterogeneous and dependent on the periods of evolution of the EMU as well as on the groups of countries considered. The fourth chapter links external imbalances to the sovereign debt crisis that EMU has been experiencing since 2009. We show that when a country belonging to a monetary union faces an external imbalance vis-à-vis another member country, the corresponding interest rate differential tends to increase. Moreover, when these imbalances persist, they can trigger a balance of payments crisis. Finally, the last chapter looks at the sustainability of the CFA zone. By comparing the CFA zone to a sample of other Sub-Saharan African countries, we show that despite its failure to meet the optimality criteria, the CFA zone has favored internal and external balances and facilitated adjustments at both the aggregate and individual levels. This suggests that this union is sustainable.

Over the last twenty years, panel data econometrics has undergone a profound renewal linked to the emergence of major time series issues, such as non-stationarity, cointegration and causality. However, it is clear that there is still very little work on taking into account non-linearity and more specifically the change of panel regimes. In this context, the general problem of this thesis consists in highlighting the interest of modeling regime shifts on panel data. Panel regime-switching models are a direct extension of the threshold models proposed for time series. However, the introduction of the individual dimension fundamentally enriches their economic interpretation. In this thesis, we study more specifically three specifications: threshold models with abrupt transition (PTR), threshold models with smooth transition (PSTR), autoregressive models with smooth transition (PSTAR). We propose three applications of these models successively on the threshold effects in Okun's law, the Feldstein Horioka (1980) paradox and the non-linear relationship between power consumption and temperature. These applications highlight various advantages of using threshold modeling in panel data. In particular, these models provide a simple parametric solution to account for both the nonlinearity and the individual heterogeneity of the parameters. The slope coefficients are also allowed to change over time. Furthermore, we show that the increase in the information set allows us to identify threshold effects that would not have been possible to identify in time series.

This thesis contributes to the literature on the two main research axes related to Value-at-Risk (VaR): forecasting and backtesting of VaR. Concerning the first axis, the thesis develops a methodology for forecasting VaR at intraday horizons. The methodology conditions non-linearly the returns on the existence of news on the market, measured by the innovation of the transactional intensity. To this end, we introduce two new specifications: a volatility model that links the asymmetric response of volatility to returns to the existence of news and a regime-switching model in the tail of the distribution driven by the existence of news. Empirical applications demonstrate that this methodology is a flexible risk management tool in markets with increasingly large intraday price fluctuations. For the second axis, we introduce two new approaches for VaR backtesting. First, we develop a test that extends the existing ones by checking the validity of VaR for several confidence levels. The test statistic is an extension to the multivariate framework of the Ljung-Box test for the VaR violation process. Second, we use the (geometric) distribution of the duration between two consecutive violations under the assumption of VaR validity and introduce a new test statistic, derived from the moment conditions induced by the orthonormal polynomials associated with the geometric distribution. It is shown that both new tests have good properties at finite distance.

The thesis proposes both a positive and normative analysis of the productive contribution of infrastructure. The first chapter is devoted to the definition of infrastructure. This study shows that the notions of public capital and infrastructure capital are often confused and that most OECD countries have experienced a sharp decline in their infrastructure investments since the mid-1970s. The second chapter provides a synthesis of the literature on the evaluation of returns to public capital. This summary highlights the sensitivity of the econometric results to the specification of the models and to the inclusion of stochastic non-stationarity. The third chapter is devoted to an evaluation of the production function approach, carried out conditionally on a stochastic dynamic general equilibrium model. This chapter specifies the nature of the biases in the estimation of the contribution of public capital when it is understood through a production function. The fourth chapter evaluates the stabilizing properties of public investment. In an endogenous growth model, different public decision rules are evaluated. The spectral analysis of the shock response functions indicates that the more productive the expenditures are, the better their stabilizing properties. The fifth chapter identifies a set of rules for setting the optimal public investment rate. This study shows that the influence of the tax structure. It also shows how the level of infrastructure congestion affects the optimal rate of investment. The sixth chapter proposes a study of the discount rate of public projects in a growth model with tax distortions. The analysis shows that the discount rate is highly dependent on the nature of growth. The link between the discount rate and tax distortions is then examined.