A5078872448- Financial Markets and Investment Strategies
- Stock Market Forecasting Methods
- Market Dynamics and Volatility
- Complex Systems and Time Series Analysis
- Media Influence and Politics
- Misinformation and Its Impacts
- Monetary Policy and Economic Impact
- Blockchain Technology Applications and Security
- COVID-19 Pandemic Impacts
- Auditing, Earnings Management, Governance
- Spam and Phishing Detection
- COVID-19 epidemiological studies
- Advanced Text Analysis Techniques
- Forecasting Techniques and Applications
- Data-Driven Disease Surveillance
- Agriculture and Rural Development Research
- European Monetary and Fiscal Policies
- Sentiment Analysis and Opinion Mining
- Smart Materials for Construction
- Icing and De-icing Technologies
- Opinion Dynamics and Social Influence
- Radio, Podcasts, and Digital Media
- Big Data Technologies and Applications
- Energy, Environment, Economic Growth
- Veterinary Orthopedics and Neurology
Université de Montpellier
2023-2024
Université Paris 1 Panthéon-Sorbonne
2016-2024
Centre d'Économie de la Sorbonne
2017-2024
Centre National de la Recherche Scientifique
2024
Inserm
2024
Centre Hospitalier Universitaire de Montpellier
2023-2024
CentraleSupélec
2021
Groupe d'Analyse et de Théorie Economique Lyon St Etienne
2020
Proteogenomics Research Institute for Systems Medicine
2019
École Nationale de la Statistique et de l'Administration Économique
2019
We construct a novel database containing hundreds of thousands geotagged messages related to the COVID-19 pandemic sent on Twitter. create daily index social distancing—at state level—to capture distancing beliefs by analyzing number tweets keywords such as “stay home”, safe”, “wear mask”, “wash hands” and “social distancing”. find that an increase in Twitter day t-1 is associated with decrease mobility t. also orders, cases, precipitation temperature contribute reducing human mobility....
We propose a crypto-specific lexicon to quantify the investor sentiment and use it predict cryptocurrency market returns. The new achieves better accuracy (32\% higher) than traditional financial when applied an out-of-sample classification setting. empirical results reveal that positively predicts excess CRIX returns with daily in(out)-of-sample $R^2$ of 2.74\% (3.15\%) without significant evidence on return reversal. are robust inclusion alternative indices, technical indicators,...
We consider several economic uncertainty indicators for the US and UK before during COVID-19 pandemic: implied stock market volatility, newspaper-based policy uncertainty, twitter chatter about subjective future business growth, disagreement among professional forecasters GDP growth. Three results emerge. First, all show huge jumps in reaction to pandemic its fallout. Indeed, most reach their highest values on record. Second, peak amplitudes differ greatly – from a rise of around 100%...
Abstract This paper analyses whether the systematic disclosure of criminals’ origins in press affects natives’ attitudes towards immigration. It takes advantage unilateral change reporting policy announced by German newspaper Sächsische Zeitung July 2016. Combining individual-level panel data from Socio-Economic Panel 2014 to 2018 with 402,819 crime-related articles newspapers and those newspapers’ market shares, we find that systematically mentioning criminals increases relative salience...
Ultra-protective ventilation is the combination of low airway pressures and tidal volume (Vt) combined with extra corporeal carbon dioxide removal (ECCO2R). A recent large study showed no benefit ultra-protective compared to standard in ARDS (Acute Respiratory Distress Syndrome) patients. However, reduction Vt failed achieve objective less than or equal 3 ml/kg predicted body weight (PBW). The main our was assess feasibility ultra-low (Vt ≤ PBW) facilitated by ECCO2R acute respiratory...
Social media can help investors gather and share information about stock markets. However, it also presents opportunities for fraudsters to spread false or misleading statements in the marketplace. Analyzing millions of messages sent on social platform Twitter small capitalization firms, we find that an abnormally high message activity is associated with a large price increase event day followed by sharp reversal over next week. Our findings are consistent patterns pump-and-dump scheme,...
We build a novel database of around 285,000 notes from the Twitter CommunityNotes program to analyze causal influence appending contextual informationto potentially misleading posts on their dissemination. Employing difference indifference design, our findings reveal that adding context below tweet reduces thenumber retweets by almost half. A significant, albeit smaller, effect is observedwhen focusing number replies or quotes. Community Notes also increase by80% probability deleted its...
We use crowd-sourced assessments from X’s Community Notes program to examine whether there are partisan differences in the sharing of misleading information. Unlike previous studies, misleadingness here is determined by agreement across a diverse community platform users, rather than fact-checkers. find that 67% more posts Republicans flagged as compared Democrats. These results not base rate artifacts, we no meaningful over-representation among X users. Our findings provide strong evidence...
We use crowd-sourced assessments from X’s Community Notes program to examine whether there are partisan differences in the sharing of misleading information. Unlike previous studies, misleadingness here is determined by agreement across a diverse community platform users, rather than fact-checkers. find that 67% more posts Republicans flagged as compared Democrats. These results not base rate artifacts, we no meaningful over-representation among X users. Our findings provide strong evidence...
We use crowd-sourced assessments from X’s Community Notes program to examine whether there are partisan differences in the sharing of misleading information. Unlike previous studies, misleadingness here is determined by agreement across a diverse community platform users, rather than fact-checkers. find that 131% more posts Republicans flagged as compared Democrats. These results not base rate artifacts, we no meaningful over-representation among X users. Our findings provide strong evidence...
We use crowd-sourced assessments from X’s Community Notes program to examine whether there are partisan differences in the sharing of misleading information. Unlike previous studies, misleadingness here is determined by agreement across a diverse community platform users, rather than fact-checkers. find that 131% more posts Republicans flagged as compared Democrats. These results not base rate artifacts, we no meaningful over-representation among X users. Our findings provide strong evidence...
The revolution, which is quite recent, brought about by digital convergence and connected objects, has enabled a homogenisation of data types would historically have been considered as different, for example: data, texts, sound, still images, moving images.This encouraged the Big Data phenomenon, volume includes two related parameters: quantity frequency acquisition; can extend far exhaustivity be up to including real time.This Special Issue features series articles that examine its uses...
We consider several economic uncertainty indicators for the US and UK before during Covid-19 pandemic: implied stock market volatility, newspaper-based policy uncertainty, twitter chatter about subjective future business growth, disagreement among professional forecasters GDP growth. Three results emerge. First, all show huge jumps in reaction to pandemic its fallout. Indeed, most reach their highest values on record. Second, peak amplitudes differ greatly — from a rise of around 100%...
We develop a field-specific dictionary to measure the stance of European Central Bank (ECB) monetary policy (dovish, neutral, hawkish) and state Eurozone economy (positive, negative) through content ECB press conferences. In contrast with traditional textual analysis, we propose novel approach using term-weighting contiguous sequence words (n-grams) better capture subtlety central bank communication. find that quantifying communication our weighted lexicon helps predict future decisions when...
Nous étudions l’hypothèse d’efficience des marchés au niveau intra-journalier en analysant les réactions du marché aux tweets négatifs et rapports publiés sur Internet par un vendeur à découvert. À l’aide d’études d’événements, nous constatons que investisseurs peuvent générer profit anormal, faible mais significatif, passant ordres très rapidement suite informations publiques publiées réseaux sociaux. La réaction est plus forte lorsqu’une entreprise mentionnée pour la première fois Twitter,...
<titre>Résumé</titre>Les blogs permettent aux chercheurs de sortir leur tour d'ivoire, s'adresser à un public plus large et mettre en débat leurs travaux. Les échanges contradictoires restent cependant trop rares dans la blogosphère francophone.