PhD Workshop: Topics in Economic Theory
2026-27 Term 1

The topics for this course change every year.
I cover term 1; the second part is during term 2 and is taught by Ran Spiegler.

Course Description

Many of the most important decisions unfold over time. In part 1 of this course, we study how people make such decisions when faced with uncertainty, limited information, and evolving opportunities.

We focus on three fundamental challenges.
First: stopping. From accepting a job offer to selling a house or ending a search, timing is often the decision itself.
Second: searching. From browsing products to developing a technology, individuals must decide not just when to choose, but need to search for what is feasible.
Third: learning from others. People rarely make decisions in isolation – they observe, imitate, and respond to what others do.
Understanding these dynamics is key to understanding real-world behaviour in a variety of contexts, from product and technology adoption to the diffusion of misinformation.

This course brings together classic and modern models that address these questions, combining theoretical rigour with economic insight. While technically rooted, the emphasis is on understanding how decisions are made – and sometimes mis-made – in a variety of contexts, such as labour markets, online platforms, financial settings, and social environments.

The course will mostly focus on continuous-time models with a wide variety of applications and has two main objectives. First, to develop tools to analyse continuous-time problems, which should be part of the toolkit of any economic theorist, but also those interested in cutting-edge research in macroeconomics (labour search), econometrics (sequential testing), and finance (option pricing). Second, to foster theoretical research and inspire theory-driven applied research by providing an overview of a collection of models spanning multiple topics of interest, in which time is of the essence.

In the final part of the course, students will present and discuss their research ideas. The goal is to spark questions and inspire research.

This course is aimed at PhD students interested in topics and methods, and in theory, behavioural economics, and dynamic decision-making more broadly. A good grasp of probability theory, real analysis and optimisation is assumed. The course is open to students from outside UCL; please send me an email if you are interested.

Meeting Times and Location
Monday, 9:00-11:00. Drayton House, room 321.
Meeting Dates: 5 October – 14 December 2026. No class on 30 November.

Topics (work in progress)

1. Optimal Stopping in Continuous Time: Stochastic Processes. Consumer search, pricing, and complexity. Exercising an option.
   Lecture(s): 5, 12, and 19 October.
   References: Branco Sun Villas-Boas (2012 MnSc), Fudenberg Strack Strzalecki (2018 AER), Gonçalves (2026 WP), Cripps Gonçalves (2026 WP), Dixit & Pindyck (1994 ch. 5), Oksendal (2003), Pham (2009).
2. Optimal Control in Continuous Time: Search intensity, directed attention, conflict.
   Lecture(s): 26 October, 2 and 16 November.
   References: Moscarini Smith (2001 Ecta), Urgun Yariv (2025 JPE), Barilla Gonçalves (2024 TE), Liang Mu Syrgkanis (2023 Ecta), Oksendal (2003), Pham (2009), Crandall Ishii Lions (1992).
3. Learning in Games 1: Fictitious Play and Best-Response Dynamics.
   Lecture(s): 23 November (time willing).
   References: Fudenberg and Levine (1998 Book, Ch. 2-4; 2009 ARE).
   Slides. Handouts.
4. Learning in Games 2: Approachability, Calibration, and Adaptive Algorithms.
   Lecture(s): 23 November (time willing).
   References: Blackwell (1956); Maschler, Solan, and Zamir (2013 Book, Ch. 14); Foster and Vohra (1997 GEB, 2018 GEB); Foster and Hart (2021 JPE; 2023 TE) Hart and Mas-Colell (2003 AER); Hart (2005 Ecta).
   Slides. Handouts.
5. Project Presentations.
   Lecture(s): 7 and 14 December.
   Presentations by students.

Materials will be posted below as the course progresses. Here is a list of papers: (suggestions of papers to be added are welcome!)

Paper Presentations
Most weeks, every member of the class will be required to work in a group of at most 3 people to prepare a 15 minute presentation on an assigned paper. One group will be selected at random to give the presentation at the start of the class.

• 14 October: Laiho Murto Salmi (2025 TE) Gradual learning from incremental actions.
• 21 October: McClellan (2022 Ecta) Experimentation and Approval Mechanisms.
• 28 October: Bolton Harris (1999 Ecta) Strategic Experimentation.
• 4 November: Zhong (2022 Ecta) Optimal Dynamic Information Acquisition or Bloedel Zhong (2026 WP) The Cost of Optimally Acquired Information.
• 18 November: Che Hörner (2018 QJE) Recommender Systems as Mechanisms for Social Learning or Keppo Smith Davydov (2008 REStud) Optimal Electoral Timing: Exercise Wisely and You May Live Longer.
• 25 November: Abreu Gul (2000 Ecta) Bargaining and Reputation and Faingold Sannikov (2011 Ecta) Reputation in Continuous-Time Games.

Project Presentations
The project presentations are to be structured as follows: 20min project idea (motivation, model sketch, target results, proof intuition if available); 10min discussion.
2 December

9 December