Julien Pascal

I am really thrilled to see this article finally in press! 👉onlinelibrary.wiley.com/doi/10.1111/jo… The paper is a survey on Heterogeneous Agent (New Keynesian) models [HA(NK)]. It reviews (i) selected key mechanisms in HA(NK)s models and why we should (or not?) care and ... 1/5

LLM hallucination, or correct shared consensus?

behavioral parameters like the MPC aren't stable, so a model that treats them as fixed constants will give wrong predictions precisely when you need it most."

LLM on Macro: "One camp - call it the "IS-LM plus common sense" school - argues that for 90% of policy questions, the old framework with some mental adjustments for expectations and inflation gives you the right answer. People like Blanchard, Krugman operate roughly in this mode"

"Submit paper" (1968) "Get referee reports" (1972) "Accepted for publication" (2026)

Could be improved, but Economics is ranked 1st in terms of "precisely reproduced". (0% "precisely reproduced" in Education, really?🧐)

Economists in 2030 be like:

We added a new and final chapter to Dynamic Programming Vol II (DP2) on approximation and reinforcement learning: dp.quantecon.org. DP1 and DP2 will always remain freely available online -- give me liberty or give me death 🤟💀

Ready to master DSGE modelling? 📊 The Dynare Masterclass 2026 is here! 🚀 Not only for students 👩‍🎓, but also junior/senior economists 🤷🏼! 🗓️ 6-10 July 2026 📍Paris 👉Apply Now: dynare.org/events/dynare-… #Dynare #Dynare2026 #Macroeconomics #CentralBanking #DSGE

Tom and I have finally finished a draft of Dynamic Programming Vol 2! Exhausting but satisfying. New approach to DP theory, advanced material, many applications... dp.quantecon.org

🚀 Big News: Dynare 7.0 is out! Introducing powerful new toolkits for HANK, samplers, perfect foresight and massive performance improvements. Packages are available for download NOW: 👉 dynare.org/download/ A quick summary below 👇 (1/5)

📢 AI Summer School/conference 2026 is open for applications! 3 intensive days for PhD students in economics & finance covering: Deep Equilibrium Nets, PINNs & heterogeneous agent models, LLMs, RAG, agentic AI & NLP Link: sites.google.com/carloalberto.o… @YuchengYang1993 @HECLausanneEcon

Here is a blog post discussing and illustrating the paper with Python code/ 📋Blog post: julienpascal.github.io/post/bc_mc_pea/ n/n

[Practical payoff] The bc-MC-PEA lets you trade off: * exploring more states (M) * integrating shocks more accurately (N) Optimal values for (M*, N*) are derived in the paper. Easy to implement. 5/n

🚨I just wrote a short paper on a generalization of the Parameterized Expectations Algorithm (PEA), recently published in Economics Letters 🙏 📄 Paper: doi.org/10.1016/j.econ… 💻 Replication code: github.com/JulienPascal/b… 🧵⬇️ 1/n

"This was not Keynes's intent, nor is it the view of all of his most eminent followers. Yet if one does not view the revolution in this way, it is impossible to account for some of its most important features." 5/5

Why combine results *and* methods in the same paper? To quote Lucas & Sargent (After Keynesian Macroeconomics, 1978): "The Keynesian Revolution was, in the form in which it succeeded in the United States, a revolution in method." 4/5

I am really thrilled to see this article finally in press! 👉onlinelibrary.wiley.com/doi/10.1111/jo… The paper is a survey on Heterogeneous Agent (New Keynesian) models [HA(NK)]. It reviews (i) selected key mechanisms in HA(NK)s models and why we should (or not?) care and ... 1/5