MAILLET Bertrand

In this article, we first generalize the Conditional Auto-Regressive Expected Shortfall (CARES) model by introducing the loss exceedances of all (other) listed companies in the Expected Shortfall related to each firm, thus proposing the CARES-X model (where the ‘X’, as usual, stands for eXtended in the case of large-dimensional problems). Second, we construct a regularized network of US financial companies by introducing the Least Absolute Shrinkage and Selection Operator in the estimation step. Third, we also propose a calibration approach for uncovering the relevant edges between the network nodes, finding that the estimated network structure dynamically evolves through different market risk regimes. We ultimately show that knowledge of the extreme risk network links provides useful information, since the intensity of these links has strong implications on portfolio risk. Indeed, it allows us to design effective risk management mitigation allocation strategies.

No summary available.

We investigate the difference in pricing cross-sectional risky assets performance between tradable and nontradable factors by comparing their misspecification errors—the Hansen–Jagannathan (HJ) distance. By constructing nontradable factors mimicking portfolios (FMPs) and incorporating them into the least-misspecified tradable stochastic dis-count factor (SDF), we provide cross-country empirical evidence that this SDF that combines tradable and nontradable factors dominates others in which nontradable factors further decrease the SDF’s mis-specification errors. Since nontradable FMPs are functions of current tradable factor information about the economic state, FMPs “hedge” the state variable risks, and FMPs’ returns describe the risk premiums.

No summary available.

A new method has been implemented to de-constrain the correlation matrix of equity returns based on a constrained principal component analysis using financial data. Portfolios, named "Fundamental Maximum variance portfolios", are constructed to optimally capture a risk style defined by a financial criterion ("Book", "Capitalization", etc.). The constrained eigenvectors of the correlation matrix, which are linear combinations of these portfolios, are then studied. Thanks to this method, several stylized facts of the matrix have been highlighted among which: i) the increase of the first eigenvalues with the time scale from 1 minute to several months seems to follow the same law for all the significant eigenvalues with two regimes. ii) a _universal_ law seems to govern the composition of all the portfolios "Maximum variance". Thus, according to this law, the optimal weights would be directly proportional to the ranking according to the financial criterion studied. iii) the volatility of the "Maximum Variance_" portfolios, which are not orthogonal, would explain a large part of the diffusion of the correlation matrix. iv) the leverage effect (increase of the first eigenvalue with the fall of the market) exists only for the first mode and does not generalize to the other risk factors. The leverage effect on the beta, the sensitivity of stocks with the "market mode", makes the weights of the first eigenvector variable.

No summary available.

Several recent finance articles use the Omega measure (Keating and Shadwick, 2002), defined as a ratio of potential gains out of possible losses, for gauging the performance of funds or active strategies, in substitution of the traditional Sharpe ratio, with the arguments that return distributions are not Gaussian and volatility is not always the relevant risk metric. Other authors also use Omega for optimizing (non-linear) portfolios with important downside risk. However, we question in this article the relevance of such approaches. First, we show through a basic illustration that the Omega ratio is inconsistent with the Second-order Stochastic Dominance criterion. Furthermore, we observe that the trade-off between return and risk corresponding to the Omega measure, may be essentially influenced by the mean return. Next, we illustrate in static and dynamic frameworks that Omega-based optimal portfolios can be closely associated with classical optimization paradigms depending on the chosen threshold used in Omega. Finally, we present robustness checks on long-only asset and hedge fund databases, that confirm our results.

In this paper, we propose to test an extension of the Capital Asset Pricing Model (CAPM), in which we add a systemic risk factor, measured by an aggregation of the main systemic risk measures, as recently proposed by different authors, in the framework of a parsimonious principal component analysis (PCPA). Our empirical analysis of the U.S. market shows that systemic risk is indeed an important component of compensation on some securities. Finally, we propose an original application for the identification and classification of systemically important financial institutions (SIFIs).

Several recent finance articles use the Omega measure (Keating and Shadwick, 2002), defined as a ratio of potential gains out of possible losses, for gauging the performance of funds or active strategies, in substitution of the traditional Sharpe ratio, with the arguments that return distributions are not Gaussian and volatility is not always the relevant risk metric. Other authors also use Omega for optimizing (non-linear) portfolios with important downside risk. However, we question in this article the relevance of such approaches. First, we show through a basic illustration that the Omega ratio is inconsistent with the Second-order Stochastic Dominance criterion. Furthermore, we observe that the trade-off between return and risk corresponding to the Omega measure, may be essentially influenced by the mean return. Next, we illustrate in static and dynamic frameworks that Omega-based optimal portfolios can be closely associated with classical optimization paradigms depending on the chosen threshold used in Omega. Finally, we present robustness checks on long-only asset and hedge fund databases, that confirm our results.

In this paper, we propose to test an extension of the CAPM, in which we add a systemic risk factor, measured by an aggregation of the main systemic risk measures, as recently proposed by different authors, in the framework of a Parsimonious Principal Component Analysis (PPCA). Our empirical analysis on the US market shows that systemic risk is indeed an important component of the compensation on some securities. We finally propose an original application to identify and classify Systemically Important Financial Institutions (SIFIs).

No summary available.

Following the last severe financial crisis, several measures of systemic risk have been proposed to quantify the state of stress of the financial system. In this paper, we propose an aggregate index for measuring financial systemic risk based on a parsimonious principal component analysis. This methodology allows us to obtain an aggregate index that is more parsimonious and more stable over time. The application of the methodology to twelve measures of aggregate systemic risk using data from the US financial market confirms this property. It also appears that the extreme positive movements of the systemic risk index thus constructed can be considered as anticipations of periods of strong contraction of economic activity.

Following the last severe financial crisis, several measures of systemic risk have been proposed to quantify the state of stress of the financial system. In this paper, we propose an aggregate index of financial systemic risk measure based on a parsimonious principal component analysis. This methodology allows us to obtain an aggregate index that is more parsimonious and more stable over time. The application of the methodology to twelve measures of aggregate systemic risk using data from the US financial market confirms this property. It also appears that the extreme positive movements of the systemic risk index thus constructed can be considered as anticipations of periods of strong contraction of economic activity.

Several recent finance articles employ the Omega measure, proposed by Keating and Shadwick (2002) - defined as a ratio of potential gains out of possible losses - for gauging the performance of funds or active strategies (e.g. Eling and Schuhmacher, 2007. Farinelli and Tibiletti, 2008. Annaert et al., 2009. Bertrand and Prigent, 2011. Zieling et al., 2014. Kapsos et al., 2014. Hamidi et al., 2014), in substitution of the traditional Sharpe ratio (1966), with the arguments that return distributions are not Gaussian and volatility is not, always, the relevant risk metric. Other authors also use the same criterion for optimizing (non-linear) portfolios with important downside risk. However, we wonder in this article about the relevance of such approaches. First, we show through a basic illustration that the Omega ratio is inconsistent with the Strict Inferior Second-order Stochastic Dominance criterion. Furthermore, we observe that the trade-off between return and risk, corresponding to the Omega measure, may be essentially influenced by the mean return. Next, we illustrate in static and dynamic frameworks that Omega-based optimal portfolios can be associated with traditional optimization paradigms depending on the chosen threshold used in the computation of Omega. Finally, we present some robustness checks on long-only asset and hedge fund databases that all confirm our general results.

No summary available.

The measurement of systemic risk of financial institutions has become a key issue for the stability of the financial system. Most of the currently proposed measures are based on the estimation of conditional quantiles, which have the characteristic of being extremely sensitive to the estimation method and to the specification of the risk models used. We propose to correct systemic risk measures based on statistical validation procedures. Our application to covari suggests that model risk is important and that institutions identified as "systemic" differ depending on whether or not model risk is taken into account.

Fraud detection is a key issue for investors and financial authorities. Bernard Madoff's Ponzi scheme is an emblematic illustration of a large-scale fraud, always possible when well orchestrated. Traditional methods of detecting frauds require lengthy and costly investigations, requiring sophisticated financial and legal knowledge and highly trained professionals. Here we continue and extend the intuition of Billio et al [2015], who suggest the use of a performance measure - denoted GUN - to construct a fraud detection index. In order to illustrate the methodology and show its usefulness, we first analyze the Madoff case, and then, on several international equity mutual fund markets marketable in France, the number of funds potentially susceptible to fraud (or proven underperformance). The proposed alert system makes it possible to detect anomalies in several dozen funds, which should therefore be the subject of special attention.

Fraud detection is a key issue for investors and financial authorities. Bernard Madoff's Ponzi scheme is an emblematic illustration of a large-scale fraud, always possible when well orchestrated. Traditional methods of detecting frauds require lengthy and costly investigations, requiring sophisticated financial and legal knowledge, and highly trained professionals. Here we continue and extend the intuition of Billio et al [2015], who suggest the use of a performance measure - denoted GUN - to construct a fraud detection index. In order to illustrate the methodology and show its usefulness, we first analyze the Madoff case, and then, on several international equity mutual fund markets marketable in France, the number of funds potentially susceptible to fraud (or proven underperformance). The proposed alert system makes it possible to detect anomalies in several dozen funds, which should therefore be the subject of special attention.

The measurement of systemic risk of financial institutions has become a key issue for the stability of the financial system. Most of the currently proposed measures are based on the estimation of conditional quantiles, which have the characteristic of being extremely sensitive to the estimation method and to the specification of the risk models used. We propose to correct systemic risk measures based on statistical validation procedures. Our application to covari suggests that model risk is important and that institutions identified as "systemic" differ depending on whether or not model risk is taken into account.

No summary available.

No summary available.

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.

No summary available.

No summary available.

The global minimum variance portfolio computed using the sample covariance matrix is known to be negatively affected by parameter uncertainty, an important component of model risk. Using a robust approach, we introduce a portfolio rule for investors who wish to invest in the global minimum variance portfolio due to its strong historical track record, but seek a rule that is robust to parameter uncertainty. Our robust portfolio corresponds theoretically to the global minimum variance portfolio in the worst-case scenario, with respect to a set of plausible alternative estimators of the covariance matrix, in the neighbourhood of the sample covariance matrix. Hence, it provides protection against errors in the reference sample covariance matrix. Monte Carlo simulations illustrate the dominance of the robust portfolio over its non-robust counterpart, in terms of portfolio stability, variance and risk-adjusted returns. Empirically, we compare the out-of-sample performance of the robust portfolio to various competing minimum variance portfolio rules in the literature. We observe that the robust portfolio often has lower turnover and variance and higher Sharpe ratios than the competing minimum variance portfolios.

No summary available.

Money market funds, shareholder behavior and financial stability.

No summary available.

The global minimum variance portfolio computed using the sample covariance matrix is known to be negatively affected by parameter uncertainty, an important component of model risk. Using a robust approach, we introduce a portfolio rule for investors who wish to invest in the global minimum variance portfolio due to its strong historical track record, but seek a rule that is robust to parameter uncertainty. Our robust portfolio corresponds theoretically to the global minimum variance portfolio in the worst-case scenario, with respect to a set of plausible alternative estimators of the covariance matrix, in the neighbourhood of the sample covariance matrix. Hence, it provides protection against errors in the reference sample covariance matrix. Monte Carlo simulations illustrate the dominance of the robust portfolio over its non-robust counterpart, in terms of portfolio stability, variance and risk-adjusted returns. Empirically, we compare the out-of-sample performance of the robust portfolio to various competing minimum variance portfolio rules in the literature. We observe that the robust portfolio often has lower turnover and variance and higher Sharpe ratios than the competing minimum variance portfolios.

Performance measurement is one of the most studied subjects in financial literature. Since the introduction of the Sharpe ratio in 1966, a large variety of new measures has appeared constantly in scientific journals as well as in practitioners' publications. The most complete and significant studies of performance measures, so far, have been written by Aftalion and Poncet, Le Sourd, Bacon, and Cogneau and H übner. A review of the most recent literature led us to collect several dozen measures that we classify into four families. We first present the class of relative measures, starting with the Sharpe ratio. Secondly, we analyse absolute measures, beginning with the most famous one ‐ the Jensen alpha. Thirdly, we study general measures based on specific features of the return distribution, where the pioneering contributions are those of Bernardo and Ledoit, and Keating and Shadwick. Finally, the fourth set concerns a few measures that explicitly take into account the investor's utility functions.

The experience from the global financial crisis has raised serious concerns about the accuracy of standard risk measures as tools for the quantification of extreme downward risks. A key reason for this is that risk measures are subject to a model risk due, e.g. to specification and estimation uncertainty. While regulators have proposed that financial institutions assess the model risk, there is no accepted approach for computing such a risk. We propose a remedy for this by a general framework for the computation of risk measures robust to model risk by empirically adjusting the imperfect risk forecasts by outcomes from backtesting frameworks, considering the desirable quality of VaR models such as the frequency, independence and magnitude of violations. We also provide a fair comparison between the main risk models using the same metric that corresponds to model risk required corrections.

Constant proportion portfolio insurance" (CPPI) is nowadays one of the most popular techniques for portfolio insurance strategies. It simply consists of reallocating the risky part of a portfolio with respect to market conditions, via a leverage parameter - called the multiple - guaranteeing a predetermined floor. We propose to introduce a conditional time-varying multiple as an alternative to the standard unconditional CPPI method, directly linked to actual risk management problematics. This "ex ante" approach for the conditional multiple aims to diversify the risk model associated, for example, with the expected shortfall (ES) or extreme risk measure estimations. First, we recall the portfolio insurance principles, and main properties of the CPPI strategy, including the time-invariant portfolio protection (TIPP) strategy, as introduced by Estep and Kritzman (1988). We emphasize the existence of an upper bound on the multiple, for example to hedge against sudden drops in the market. Then, we provide the main properties of the conditional multiples for well-known financial models including the discrete-time portfolio rebalancing case and Lévy processes to describe the risky asset dynamics. For this purpose, we precisely define and evaluate different gap risks, in both conditional and unconditional frameworks. As a by-product, the introduction of discrete or random time portfolio rebalancing allows us to determine and/or estimate the density of durations between rebalancements. Finally, from a more practical and statistical point of view due to trading restrictions, we present the class of Dynamic AutoRegressive Expectile (DARE) models for estimating the conditional multiple. This latter approach provides useful complementary information about the risk and performance associated with probabilistic approaches to the conditional multiple.

No summary available.

“Constant proportion portfolio insurance” is a popular technique among portfolio insurance strategies: the risky part of a portfolio is reallocated with respect to market conditions, via a fixed parameter (the multiple), guaranteeing a predetermined floor. We propose here to use a conditional time-varying multiple as an alternative. We provide the main properties of the conditional multiples for some mainstream cases, including discrete-time rebalancing and an underlying risk asset driven by the Lévy process, while evaluating conditional and unconditional gap risks. Finally, we evaluate the use of a dynamic autoregressive expectile model for estimating the conditional multiple in such a context.

“Constant proportion portfolio insurance” is a popular technique among portfolio insurance strategies: the risky part of a portfolio is reallocated with respect to market conditions, via a fixed parameter (the multiple), guaranteeing a predetermined floor. We propose here to use a conditional time-varying multiple as an alternative. We provide the main properties of the conditional multiples for some mainstream cases, including discrete-time rebalancing and an underlying risk asset driven by the Lévy process, while evaluating conditional and unconditional gap risks. Finally, we evaluate the use of a dynamic autoregressive expectile model for estimating the conditional multiple in such a context.

In this thesis, we examine the important contributions of dependence modeling by copula theory in the context of problems related to the management of financial portfolios and structured credit products.The first part of this thesis is devoted to the management of financial portfolios. The first part of this thesis is devoted to the management of financial portfolios. We first study the relationship that can be established between the level of diversification of the portfolio and the choice of the copula that best describes the dependency structure. The objective is to identify a characteristic in the portfolios allowing a simpler selection of the appropriate copula. In a second chapter, we propose to study the impact of a misspecification of the copula model on the estimates of conventional risk measures such as Value-at-Risk and Expected-Shortfall. The idea is to verify the usefulness of developing these estimates under the true copula model. In a third chapter, we study the impact of a misspecification of the copula model in the context of an optimal portfolio allocation problem. The main objective is to identify the sensitivity of investors, according to their degree of risk aversion (losses), for one or the other component of the copula model. We propose to establish a bridge between the teachings of behavioral finance theories and the modeling of dependence by the copula theory.The second part of the thesis deals with structured credit products. We study, in a first chapter, the contribution of an actuarial model, using copula functions in the modeling of the dependence structure between default times, in the context of the estimation process of risk measures. Finally, in a last chapter we revisit the notion of the "Diversity Score", developed by the rating agency Moody's in order to assign the quality of structured credit products in terms of diversification. We discuss the analogy of this measure with that of the copula approach, and we demonstrate its adequacy with some families of copula functions.

In this thesis, we propose to study the longevity risk and its impact on insurers and pension plan providers. In the first part, we focus on the different mortality models. Based on this literature review, we propose an extension of the so-called CBD model, using Lévy processes, which will take into account the effects of jumps in the mortality curve. In the second part of the thesis, this new model will be used as a basis for the valuation of longevity derivatives. We use as valuation measures the so-called Wang and Esscher transforms that we will have previously defined and justified as being martingale measures equivalent to the historical measure. Finally, we propose a new contract called "mortality collar" which, by its definition, allows an efficient hedge against longevity/mortality risk, both for an insurer and for a pension fund. We provide an in-depth analysis of this risk management tool, both in its mechanism and in its valuation.

Performance measurement is one of the most studied subjects in financial literature. Since the introduction of the Sharpe ratio in 1966, a large variety of new measures has appeared constantly in scientific journals as well as in practitioners' publications. The most complete and significant studies of performance measures, so far, have been written by Aftalion and Poncet, Le Sourd, Bacon, and Cogneau and H übner. A review of the most recent literature led us to collect several dozen measures that we classify into four families. We first present the class of relative measures, starting with the Sharpe ratio. Secondly, we analyse absolute measures, beginning with the most famous one - the Jensen alpha. Thirdly, we study general measures based on specific features of the return distribution, where the pioneering contributions are those of Bernardo and Ledoit, and Keating and Shadwick. Finally, the fourth set concerns a few measures that explicitly take into account the investor's utility functions.

We study in this article the problem of model risk in VaR computations and document a procedure for correcting the bias due to specification and estimation errors. This practical method consists of “learning from model mistakes”, since it dynamically relies on an adjustment of the VaR estimates – based on a back‐testing framework – such as the frequency of past VaR exceptions always matches the expected probability. We finally show that integrating the model risk into the VaR computations implies a substantial minimum correction to the order of 10–40% of VaR levels.

Following the recent crisis and the revealed weakness of risk management practices, regulators of developed markets have recommended that financial institutions assess model risk. Standard risk measures, such as the value-at-risk ( VaR), emerged during the 1990s as the industry standard for risk management and become today a key tool for asset allocation. This paper illustrates and estimates model risk, and focuses on the evaluation of its impact on optimal portfolios at various time horizons. Based on a long sample of US data, the paper finds a non-linear relation between VaR model errors and the horizon that impacts optimal asset allocations.

Following the recent crisis and the revealed weakness of risk management practices, regulators of developed markets have recommended that financial institutions assess model risk. Standard risk measures, such as the value‐at‐risk (VaR), emerged during the 1990s as the industry standard for risk management and become today a key tool for asset allocation. This paper illustrates and estimates model risk, and focuses on the evaluation of its impact on optimal portfolios at various time horizons. Based on a long sample of US data, the paper finds a non‐linear relation between VaR model errors and the horizon that impacts optimal asset allocations.

No summary available.

Following the recent crisis and the revealed weakness of risk management practices, regulators of developed markets have recommended that financial institutions assess model risk. Standard risk measures, such as the Value-at-Risk (VaR), emerged over recent decades as the industry standard for risk management and have today become a key tool for asset allocation. We illustrate and estimate model risk, and focus on the evaluation of its impact on optimal portfolios at various time horizons. Based on a long sample of U.S. data, we find a non-linear relation between VaR model errors and the horizon that impacts optimal asset allocations.

No summary available.

No summary available.

No summary available.

This thesis is essentially interested in the criteria defining market performance. It proposes the study of different measures of fund performance compatible with the hypothesis of rational choice, in the sense of Von Neumann and Morgenstem (1947), of individual investors. The first chapter of part one aims to give a quick overview of the main candidate processes for a good representation of stock market returns. None of them corresponds, however, to all the stylized facts highlighted in the literature. The second chapter concerns an empirical study of stock returns on high frequency data on the Parisian stock market, in a distorted time scale - the time-volume -. The third and fourth chapters attempt to provide an evaluation, as complete as possible, of the methods of forecasting returns based on the precepts of technical analysis. The first, second and third chapters of the second part of this thesis provide a review of the literature on performance measures. The first chapter reports on performance measures related to valuation models and the biases associated with them. The second chapter deals with the main measures without reference to an explicit evaluation model. The third chapter compares the main performance measures and presents the findings of the main empirical studies on fund performance. The last chapter proposes two new measures: one, for managed funds, is a generalization of Sharpe's measure to the case of non-perfect financial markets. The other, for primary securities, is essentially an extension of Treynor's measure to the case of many factor risks.