This post originates from a set of insightful comments delivered by renowned economist John Cochrane at the NBER Asset Pricing conference held at Stanford in early November. The conference agenda and John Cochrane’s comprehensive slides from the event are publicly accessible, offering further context to his remarks. While video footage was initially available, it has since been removed.
Cochrane was invited to discuss the paper “Downward Nominal Rigidities and Bond Premia” by François Gourio and Phuong Ngo. He commends the paper for its clarity and rigor. However, Cochrane’s discussion broadens beyond this specific paper to address more fundamental issues within established economic research, particularly concerning the prevailing new-Keynesian models. He emphasizes that his comments are directed towards the broader field and its common practices, rather than being a critique of Gourio and Ngo’s work itself.
Gourio and Ngo’s paper builds upon Bob Lucas’s seminal 1973 observation about international evidence. Lucas noted that in economies experiencing high inflation, prices tend to be less sticky, leading to a more vertical Phillips curve. Consequently, in such environments, output becomes less sensitive to inflation. This phenomenon suggests that in high-inflation settings, individuals recognize that price fluctuations are primarily driven by aggregate factors rather than relative price shifts, and therefore pay less attention to them. Gourio and Ngo incorporate this concept into a new-Keynesian framework that includes downwardly sticky prices and wages. Their model suggests that periods of low inflation are more likely to fall into this sticky-price regime. They further explore the implications of this idea for bond risk premia, arguing that low-inflation periods can lead to increased correlation between inflation and output. This altered correlation subsequently affects the relationship between nominal bond returns and the discount factor, influencing the term premium.
Cochrane’s commentary addresses two main areas: bond premiums and new-Keynesian models. This post will focus on the latter. Cochrane points out a recurring trend in economic modeling: the continuous addition of layers and complexities to the standard textbook new-Keynesian model. He argues that while individual papers like Gourio and Ngo’s contribute valuable insights, the underlying textbook model itself suffers from significant structural flaws. Despite the existence of known methods to rectify these issues, or the potential for discovering new solutions, the economic research community often continues to build upon this potentially problematic foundation instead of fundamentally addressing its inherent weaknesses.
A primary concern raised by John Cochrane is:
Problem 1: The “Wrong” Sign or Unconventional Implications
