Portail HAL HEC
Not a member yet
4784 research outputs found
Sort by
Reflections of Women Standard Setters in the United States
No full textMotivated by trends toward gender equality on standard-setting boards in the United States, this study interviews women members of the FASB, GASB, and EITF to understand the factors critical to their successful nomination and appointment. Semi-structured interviews were conducted with women standard setters to root our understanding in their own experiences and perceptions. Value emerged as a generalizing theme. Participants perceived value to the board in nomination as associated with professional expertise and ties with professional societies. Most participants perceived value in giving back to the profession by serving as a voice for an important stakeholder group as a critical factor for acceptance. Women standard setters consistently downplay the idea that diversity, equity, and inclusion represent primary decision criteria for board membership. Instead, their reflections imply that it is up to the profession to implement practices that promote the advancement of the most competent professionals from all backgrounds
Machine Learning and IRB Capital Requirements: Advantages, Risks, and Recommendations
No full textThis survey proposes a theoretical and practical reflection on the use of machine learning methods in the context of the Internal Ratings Based (IRB) approach to banks' capital requirements. While machine learning is still rarely used in the regulatory domain (IRB, IFRS 9, stress tests), recent discussions initiated by the European Banking Authority suggest that this may change in the near future. While technically complex, this subject is crucial given growing concerns about the potential financial instability caused by the banks' use of opaque internal models. Conversely, for their proponents, machine learning models offer the prospect of better measurement of credit risk and enhancing financial inclusion. This survey yields several conclusions and recommendations regarding (i) the accuracy of risk parameter estimations, (ii) the level of regulatory capital, (iii) the trade-off between performance and interpretability, (iv) international banking competition, and (v) the governance and operational risks of machine learning models
Equilibrium Data Mining and Data Abundance
No full textWe model of the search for predictors by speculators (active asset managers) and use it to analyze how the improvement in data processing power and the growth in available data (“data abundance”) affect the diversity of trading signals used by speculators, the dispersion of their profits and the similarities of their holdings. Our central message is that data abundance and computing power do not have the same effects. In particular, an improvement in computing power always raises the bar for the quality of predictors that managers consider good enough to exploit while more data lower it when data becomes sufficiently abundant. When this happens, the diversity of speculators’ signals and the dispersion of their trading profits increase in equilibrium while their holdings become less correlated
National Security and Corporate Investment
No full textWe examine whether and how the implementation of national security policies by the Committee on Foreign Investment in the United States (CFIUS) affects corporate investment. CFIUS can block a foreign takeover of a US company when its closing threatens to impair the national security of the United States. On a sample of 41,918 firm-year observations (6,192 US incorporated listed firms) over the 2008–2019 period, we find that CFIUS interventions reduce corporate investment by firms in industries of national security interest. We also document a decrease in corporate investment following the 2016 US presidential election, which increased the likelihood of CFIUS interventions. The effects are more pronounced among firms that are financially constrained and more exposed to foreign investments. Our results may be of interest to regulators who have recently adopted CFIUS-like mechanisms to protect their critical assets from foreign takeovers
Machine Learning et Modèles IRB : Avantages, Risques et Préconisations
No full textL’objectif de cette étude est de proposer une réflexion théorique et pratique sur les enjeux de l’utilisation des méthodes d'apprentissage automatique (Machine Learning) dans le cadre spécifique des modèles de risque de crédit basés sur les notations internes (IRB), permettant le calcul des fonds propres réglementaires des banques. Si le ML est aujourd'hui encore peu utilisé dans le domaine réglementaire (IRB, IFRS9, stress tests), les récentes discussions initiées par l'Autorité Bancaire Européenne (EBA) laissent penser que cet usage pourrait se développer dans le futur. Bien que techniquement complexe, ce sujet est crucial compte tenu des craintes pour la stabilité financière que soulèvent l'utilisation de modèles internes sophistiqués et opaques pour la gestion des fonds propres. A l’inverse, pour leurs défenseurs, ces modèles offrent la perspective de mieux mesurer le risque de crédit et d'ouvrir de nouvelles perspectives en termes d'inclusion financière. De cette étude, ressortent plusieurs conclusions et recommandations concernant (i) les enjeux de concurrence bancaire internationale liés aux données utilisées par ces modèles, (ii) l'amélioration de la précision dans l'estimation des paramètres de risque, (iii) la réduction des fonds propres réglementaires pour les banques, (iv) la nécessité de remettre en cause l'arbitrage entre performance et interprétabilité, et de développer dans ce contexte des modèles de ML nativement interprétables et (v) le défi de la gouvernance, des risques opérationnels et de la formation
Non-local magnon transconductance in extended magnetic insulating films.\\Part II: two-fluid behavior
No full textThis review presents a comprehensive study of the spatial dispersion of propagating magnons electrically emitted in extended yttrium-iron garnet (YIG) films by the spin transfer effects across a YIGPt interface. Our goal is to provide a generic framework to describe the magnon transconductance inside magnetic films. We experimentally elucidate the relevant spectral contributions by studying the lateral decay of the magnon signal. While most of the injected magnons do not reach the collector, the propagating magnons can be split into two-fluids: \textit{i)} a large fraction of high-energy magnons carrying energy of about , where is the lattice temperature, with a characteristic decay length in the sub-micrometer range, and \textit{ii)} a small fraction of low-energy magnons, which are particles carrying energy of about , where is the Kittel frequency, with a characteristic decay length in the micrometer range. Taking advantage of their different physical properties, the low-energy magnons can become the dominant fluid \textit{i)} at large spin transfer rates for the bias causing the emission of magnons, \textit{ii)} at large distance from the emitter, \textit{iii)} at small film thickness, or \textit{iv)} for reduced band mismatch between the YIG below the emitter and the bulk due to variation of the magnon concentration. This broader picture complements part I \cite{kohno_SD}, which focuses solely on the nonlinear transport properties of low-energy magnons
The Ecosystem Penalty: Value Creation Technologies and Incentive Misalignment
No full textWhen are the incentives of a business ecosystem’s participants aligned with its growth? How is the type of complementarities between ecosystem components affecting this alignment? Developing a formal model of value creation and value capture in an ecosystem, we find that alignment is typically imperfect compared to an integrated benchmark, highlighting an “ecosystem penalty” whereby participants’ returns to value creation are lower than that of the ecosystem. Contrary to conventional wisdom, ecosystems with strong synergies between components can exhibit increasing misalignment when its participants have strong ex ante capabilities, while those with weaker synergies can be well-aligned when the orchestrator is strong and complementors are close substitutes. Ecosystems where value creation is constrained by its weakest component exhibit both the best and the worst alignment
Money and Taxes Implement Dynamic Optimal Mechanisms
No full textWe analyze dynamic capital allocation and risk sharing between a principal and many agents, who privately observe their output. Incentive compatibility requires that agents bear part of their idiosyncratic risk. The larger the agents' risk exposure, the larger the rents the principal can extract from them. The optimal mechanism can be implemented as the equilibrium of a market where agents exchange goods for money, needed to pay taxes. Inflation affects agents' portfolio choice between risky capital and safe money. To implement the optimal mechanism, the principal targets an inflation rate such that agents' risk exposure is the same in equilibrium and in the mechanism
The Origins of Limited Liability: Catering to Safety Demand with Investors' Irresponsibility
No full textLimited liability is a key feature of corporate law. Using data on asset prices and capital flows in mid-19th century England, I argue that its liberalization was not decided to relax firms' financing constraints, but to satisfy investors' demand for "safe" stores of value. Limited liability eliminated adverse selection about the quality of other shareholders; stocks could be held to store wealth in diversified portfolios, without extended forms of responsibility. Prices of newly issued stocks are consistent with this hypothesis. Thus, the quest for "safe" stores of value explains not only features of debt markets, but also of equity markets
Are People Willing to Pay for Reduced Inequality?
No full textWould consumers be willing to pay more for goods for which there is less inequality inwages across those involved in their production? In incentive-compatible behaviouralchoice studies on representative samples of the English and US populations, we find significantlypositive willingness to pay for such inequality reductions in over 80% of subjects.Whilst it varies with political leaning and the extent of the inequality reduction, willingnessto pay is positive across the political spectrum and for all studied inequality differences. Itis higher for more intuitive and informative inequality-reporting formats. Our findingshave policy implications for both governments and firms. On the one hand, they suggestthe promise of universal provision of product-level inequality information as a toolfor moderating income inequality. On the other, they highlight the potential relevance ofinequality reporting for firms’ marketing strategies