World-System (1950-2100) Trump vs. the US Federal Reserve

 

Jan 12, 2026. The New York Times has reported that the Fed Changes Course and Takes On Trump’s Political Fight. Trump has been pressing the US Federal Reserve (the Fed), making threats and finally having the Chair, Jerome Powell, charged with lying to Congress--all of this over interest rates (the Fed Funds Rate, FFR). Trump wants interest rate lower, the Federal Reserve Board of Governors is not sure (given Trump's economic policies).

It turns out that both Trump and the Fed are right. Trump's erratic economic policies are a threat to economic stability and interest rates are too high. However, there are bigger problems for the Fed on the horizon.

In this post I'm going to address the issue of interest rates (others have taken on the critique of Trumponomics). 

ChatGPT itemizes the following problems for the US Federal Reserve:

The Fed is expected to prevent inflation, increase employment, stabilize prices, prevent recessions, protect financial markets, remain politically independent and react to structural economic forces (e.g., shocks and system model inputs) using one main tool: the Federal Funds Rate (FFR), which indirectly affects interest rates. At the same time, the Fed is the US Central Banker charged with regulating the US Banking System. Since communications by the Fed can have rather immediate consequences for financial markets, the Fed is also expected to maintain its credibility.**

The FED policy actions with the FFR are displayed in the graphic at the beginning of the post. In the short-run (year-to-year) Fed actions are a random walk and cannot be predicted.* In the long run, the attractor path driven by the World System (WL20 model) shows that the FFR is expected to stabilize around 2025 at an historically low level.  The prediction is based on an expected peak for the World System (WL20 model) in 2035.

In addition to the problems enumerated by ChatGPT, the FED faces the added problems of (1) reduced range for FFR policy options at low levels of interest rates (the ZLB, from Google AI) 

and (2)  appropriate policy actions when the World System is in peak-and-decline mode. There will be intense pressure on the FED to keep interest rates low in the hope that it will stimulate economic growth. But the Fed will be fighting strong World-System forces that are unlikely to respond to FFR manipulations. 

Notes

* Policy actions that are a random walk and cannot be predicted prevent investors for "setting up" on the Fed. In spite of the RW and relying on the BAU model, the Federal Reserve of Atlanta tries to predict the probability of rate cuts and rate increases. Currently low:

** The Fed performs a wide range of other functions to include maintain economic statistics, conducting research and interacting with other Central Banks in the World-System. And, Fed history, from its founding during the Great Depression to the present, has not been without controversy.

*** An attractor path for the FFR is established by simulating the WL20 model staring from initial conditions in 1950. Typically, econometric models are simulated resetting initial conditions every year. The bootstrap prediction intervals show the 98% confidence bands; Fed FFR actions are well outside the 98% interval. The attractor path AIC is computed using deviations from the long-run attractor path (rather than year-to-year predicted deviations, see below).

The Financial Forecasting Center also forecasts the prime interest rate here, the inflation rate (here) and the unemployment rate (here). Other forecasts from the Atlanta Fed can be found here. Other forecasts from my models for the US Economy can be found here.

AIC Statistics

I have fit four models to find the "best model" for predicting the  Federal Funds Rate (FFR). I use the Akaikie Information Criterion (AIC) to compare models (the smaller the AIC, the better). The best year-to year model is a Random Walk (RW) with the Business-as-usual (BAU) model a close second. In other words, in the short-run, FED decisions about the FFR cannot be predicted.*

For establishing an attractor path,*** the best models use the state of the World-System (WL20 model) or the state of US Economy (USL20 model) to predict the FFR. Of these models, World-System Input is the best and is presented in the graphic at the beginning of this post. Using either the WL20 model or the USL20 model, the FFR is expected to decline over time.

FFR Model W Input

The WL20 Input model is stable. The coefficients in the input matrix (G) are somewhat larger than those for USL20 input (see below), meaning the Fed reacts more strongly to World System forces.

FFR Model US Input

The USL20 Input model is also stable with somewhat smaller coefficients in the Input Matrix (G).