Global Forecasting School

The Global Forecasting School (GFS) provides a very practical training opportunity for staff and policymakers to learn how to develop analytical frameworks for surveillance systems designed to make sense of a very complex world and to use much more sophisticated nonlinear models to study the policy implications of nonlinearities and uncertainty. The example that the GFS is focusing on is how to think about the risks of falling behind shifts in the Phillips Curve and allowing long-term inflation expectations to ratchet upwards.

The GFS is creating an integrated system to support decision-making under uncertainty. The system includes multiple ingredients: monitoring, data generation, nowcasting, medium and long-term projections and reporting. The US scenario builder is another part of the GFS. It is a front-end application in R, Python, Julia and Java that can be GFS follows 2 main principles:


  • Human capital is the key. It is very important to invest in training and continuous education of people.

  • Never say model says. Nobody should ever expect to get answers just from models. Models can only be some of many tools used by decisionmakers, but they can never give ultimate “correct” answers.

At first, GFS mechanism will be applied to the US economy given its global importance as a financial shock emitter, but later will be extended to other countries.

BPP's Optimal Zip Lines for Central Banks

FPAS Mark II is about creating critical thinkers that design and implement analytical frameworks to help policymakers avoid dark corners, where they fail to anchor long-term inflation expectations or allow excessive credit expansions and asset price bubbles.  Requires much more focus on nonlinearities and uncertainty, as well as trained economists that are good at creating narrative-based scenarios to study the policy implications of uncertainty.  

Recap: What we do at the GFS?

Larry Summers's Critique of Fed

What Does GFS Suggest?

Foundations of FPAS Mark II Framework