Séminaires de recherche

Filtrer

When Managers Speak, Markets Listen: Price Discovery during Analyst/Investor Days

Date : Jeudi | 2026-11-12 à 12h30
Lieu : Salle des thèses

Lien TEAMS : Cliquez ici pour rejoindre le séminaire du LÉO

Olga Kolokolova (Lancaster University)

We study how markets incorporate information released during Analyst/Investor Days, multi-hour corporate disclosure events comprising managerial presentations and a Q&A session. Aligning sentence-timestamped transcripts with tick-level trades and quotes for 2,490 U.S. events from 2014 to 2025, we show that 80% of the event-window return is incorporated before the Q&A begins. Price adjustment occurs through both discrete jumps and gradual drift bursts. Forward-looking and future-oriented statements are associated with larger absolute returns, while jump probabilities vary systematically with disclosure content, increasing after negative-tone speech and decreasing after positive-tone speech. CEO speech is associated with larger returns and higher jump probabilities, while drift bursts show little relation to language and are strongly associated with order flow in the preceding minutes. During presentations, the same order flow is followed by larger price moves and stronger drift bursts.

Infrastructure aid and resource trade: the crossroads of strategy and sustainable development

Date : Jeudi | 2026-11-05 à 12h30
Lieu : Salle des thèses

Lien TEAMS : Cliquez ici pour rejoindre le séminaire du LÉO

Thais Nunez-Rocha (Université d'Orléans)

We investigate whether rich countries use development aid to secure access to natural resources. Our model shows that even an altruistic donor may allocate more infrastructure aid to recipients from whom it receives more resource exports. Using CRS project-level data over 2006-2021, with a 2001-2011 robustness sample, we estimate a fixed-effects ladder of increasing stringency. Resource trade is positively associated with the number of infrastructure projects: a one-unit increase in the resource-trade term is associated with an 8 to 14 percent increase in the expected number of projects across specifications. This association is statistically significant under donor-recipient pair clustering but not under two-way clustering by donor and recipient, nor under a wild cluster bootstrap over donors. A positive value effect appears only in our most demanding, fully saturated specification and is only marginally significant under the bootstrap. No single resource category robustly drives the result, road is the only transport mode with a significant effect, and, in the fully saturated specification, greater oil tanker fleet capacity is linked to a smaller aggregate effect. We interpret our findings as a systematic but fragile allocation pattern, not a causal effect.

Epidemic Risk and the Breakdown of Analyst Information Channels:Evidence from China’s Zero-COVID Policy

Date : Mardi | 2026-11-10 à 12h
Lieu : Salle des thèses

Lien TEAMS : Cliquez ici pour rejoindre le séminaire du LÉO

Sandy Suardi  (University of Wollongong)

This paper examines how regional epidemic risk during China’s COVID-19 Zero-COVID
period affected financial analysts’ earnings forecast accuracy and related behaviors. Using
analyst- and firm-level forecasts for A-share-listed firms from 2020–2021, we show that higher
local epidemic risk significantly reduces forecast accuracy at both annual and quarterly
horizons. Two mechanisms drive this decline. First, strict lockdowns and travel restrictions
impede analysts’ ability to conduct on-site due diligence, with the adverse effects more
pronounced for distant and out-of-province analysts. Second, heightened uncertainty and
operational disruptions amplify information asymmetry, especially for non-SOEs, small firms,
firms with greater earnings management, and firms audited by non-Big 4 auditors. Epidemic
risk also reshapes analyst practices: it reduces coverage, increases forecast dispersion, raises
the likelihood of forecast revisions, and encourages team collaboration as analysts attempt to
mitigate informational frictions. Overall, the findings provide large-scale evidence that
epidemic-induced disruptions fundamentally weakened analysts’ information sets and
forecasting precision. The results highlight the vulnerability of capital-market information
production to external shocks and underscore the importance of robust information channels
when face-to-face information gathering becomes constrained.

Climatic litigations and GHG emissions

Date : Jeudi | 2026-10-15 à 12h30
Lieu : Salle des thèses

Lien TEAMS : Cliquez ici pour rejoindre le séminaire du LÉO

Inmaculada Martínez-Zarzoso (Université de Jaume I)

This paper provides the first empirical evaluation of the impact of climate litigation on greenhouse gas emissions. Using a stochastic version of the Impact, Population, Affluence and Technology framework and a sample of developed and developing countries over the period 1996–2023, we employ panel-data methods with instrumental variables and a staggered DID approach to address endogeneity. Results show that following a decided litigation, emissions decrease by around 1.8% according to the more conservative estimates. This effect is most pronounced in high-income countries. Our findings suggest that the judiciary acts as a critical enforcement mechanism, bridging the gap between legislative intent and climate action.

IA générative et recherche en économie: choc d’offre et rupture

Date : Jeudi | 2026-10-08 à 12h30
Lieu : Salle des thèses

Lien TEAMS : Cliquez ici pour rejoindre le séminaire du LÉO

Christophe Hurlin (Université d'Orléans / LEO)

Ce séminaire propose d’examiner la manière dont l’IA générative transforme les pratiques de recherche en économie, de la rédaction et la programmation à la revue de littérature et à l’évaluation des travaux. Il discutera à la fois des gains de productivité permis par ces outils et des questions qu’ils soulèvent en matière de qualité scientifique, de peer review et d’évaluation de la recherche.