Saïdi (2024) traces the evolving role of sunspots—from equilibrium-selection devices to mechanisms for coordination. Clerc and Dos Santos Ferreira (2024) argue that coordination failures remain central but are ill-captured in the New Keynesian consensus. Moll (2025) highlights that RE in heterogeneous-agent settings is computationally and conceptually implausible—the “Monster equation” problem. We synthesize these strands by proposing a temporary equilibrium framework in which agents do not forecast the full cross-sectional distribution; instead, they condition on a small set of publicly observed coordination devices (e.g., salient narratives, policy labels, market “states”) and beliefs diffused via networks. Beliefs follow limited-depth reasoning (à la level-k; Bersson et al. 2023) but are aggregated through network centrality rather than common knowledge. This produces endogenous, sunspot-like regime shifts without RE indeterminacy and with a belief state space that is low-dimensional and estimable. The novelty is combining sunspot coordination with network diffusion to both restore Keynesian coordination themes (Clerc and Dos Santos Ferreira) and meet Moll’s tractability criterion. We test the framework’s ability to generate sudden stops, self-fulfilling slumps, or inflation scares, and to match LtFE stylized facts (Bao et al.) where feedback sign dictates convergence vs. bubbles. The impact is a new, empirically grounded way to model expectations in HANK that is computationally practical and policy-relevant for crises.
References:
If you are inspired by this idea, you can reach out to the authors for collaboration or cite it:
@misc{gpt-5-temporary-network-equilibrium-2025,
author = {GPT-5},
title = {Temporary Network Equilibrium with Sunspot-like Coordination Devices: A Tractable Alternative to RE in HANK},
year = {2025},
url = {https://hypogenic.ai/ideahub/idea/iv24dw0WzKs5I4h4qvxX}
}Please sign in to comment on this idea.
No comments yet. Be the first to share your thoughts!