Nicolas Crouzet (Northwestern University, Kellogg School of Management)
Can the cure kill the patient? Corporate credit interventions and debt overhang abstract
Abstract:
Interventions in corporate credit markets were a major innovation in the policy response to the 2020 recession. This paper develops and estimates a model to quantify their impact on borrowing and investment. Even during downturns, credit interventions can be a bad policy idea, because they exacerbate debt overhang and depress investment in the long run. However, if the downturn is accompanied by financial market disruptions, they initially help forestall inefficient liquidations. These short term benefits quantitatively dominate the long run overhang costs. Additionally, constraining shareholder distributions, and targeting high-leverage firms substantially increases the "bang for the buck" of credit interventions.
t.b.a. • EPF Lausanne, Quartier UNIL - Dorigny, Salle Extranef 126
Euan Spence (University of Bath)
At the interface between semiclassical analysis and numerical analysis of wave-scattering problems
10:00 • ETH Zentrum, Rämistrasse 101, Zürich, Building HG, Room G 43
Eliane Bécache (Université Paris 6)
Abstract:
The PML method is one of the most widely used for the numerical simulation of wave propagation problems set in unbounded domains. However difficulties arise when the exterior domain has some complexity which prevents from using classical approaches. For instance, it is well-known that PML may be unstable for time-domain eslastodynamic waves in some anisotropic materials. More recently, is has also been noticed that standard PML cannot work in presence of some dispersive materials. In some cases, new stable PMLs have been designed.
In this talk, we address the questions of well-posedness, stability and convergence of standard and new models of PMLs in the context of electromagnetic waves for non-dispersive and dispersive materials.
11:00 • Universität Basel, Zoom Seminar
Johannes Hauber (ETH Zürich)
The Cup Product in Morse Cohomology abstract
Abstract:
It is known that in the category of smooth closed manifolds, Morse homology satisfies the Eilenberg-Steenrod Axioms. This means that Morse homology gives a homology theory that is equivalent to, for example, singular homology. Hence, a whole family of questions arise,that is: what are the Morse-theoretic realizations of standard constructions in well-known homology theories? One phenomenon, that can be observed answering those questions, is that the Morse theoretic incarnations of standard constructions are either much easier (e.g. the Morse theoretic incarnation of Poincare duality can be viewed as the reversal of flow lines) or much harder. We will give an outline of the construction of the cup product in Morse cohomology, which is one of the cases in which the Morse theoretic realization is harder to construct than in singular cohomology, and provide an example. One difficulty, among others, that we will encounter in this construction is that moduli spaces of Y-shaped trajectories and their compactness properties play a central role. This construction is only the special case of a broader family of constructions, namely A-infinity operations, on the Morse cochain complex.Our exposition is based on the work of Fukaya and Abouzaid, who introduced these A-infinity operations of which the cup product is the simplest operation, and especially Mescher\'s book "Perturbed Gradient Flow Trees and A-infinity-algebra Structures in Morse Cohomology".
13:00 • Université de Neuchâtel, Rue Emile-Argand 11, Room E213
Prof. Dr. Andras Szenes (Université de Genève)
Diagonal bases and wall-crossings in moduli spaces of vector bundle abstract
Abstract:
The idea of the calculation of the Hilbert function of the moduli spaces of vector bundles on Riemann surfaces goes back to the works of Michael Thaddeus in the early 90’s.I describe joint work with Olga Trapeznikova where this plan is carried out in detail, which uses only basic tools of Geometric Invariant Theory and a combinatorial/analytic device introduced by myself: the diagonal basis of hyperplane arrangements.
13:30 • Université de Genève, Section Mathématiques, Bâtiment: Conseil Général 7-9, Room 06-13
Dr. Steven Charlton (Universität Hamburg)
Generators of multiple t values, and alternating multiple zeta values abstract
Abstract:
Multiple zeta values, and their relatives including the multiple t values, are a prominent but mysterious class of real numbers, which appear in various areas from high energy physics and knot theory, to number theory and the periods of mixed Tate motives. I will review some work by Francis Brown, and some recent work by Takuya Murakami, on how to prove certain elements \\zeta(2\'s and 3\'s), and t(2\'s and 3\'s), generate the space of multiple zeta values. I will then extend Murakami\'s work to show t(1\'s and 2\'s) generate the space of multiple t values and alternating multiple zeta values, and make some progress towards Saha\'s conjecture that t(1\'s and 2\'s, 2 or 3) are a basis for convergent MtV’s.
14:15 • ETH Zentrum, Rämistrasse 101, Zürich, Building HG, Room G 43
Antoine Chambaz (Université de Paris)
Abstract:
Empirical risk minimization (ERM) is the workhorse of machine learning but its model-agnostic guarantees can fail when using data collected in an adaptive fashion, like in the setting of a contextual bandit algorithm for instance. In this setting, and focusing on policy learning, I will present a generic importance sampling weighted ERM algorithm and its regret guarantees, which close an open gap in the existing literature whenever exploration decays to zero. An empirical investigation validates the theory.
This is a joint work with Aurélien Bibaut (Netflix), Nathan Kallus (Cornell University and Netflix), Maria Dimakopoulou (Netflix) and Mark J. van der laan (UC Berkeley).
15:15 • EPF Lausanne, Zoom Seminar