Forecasting Short Rates with Policy Rules and Time-Series Models

Summary

The note outlines several ways to think about future short-term interest rates. It emphasizes the role of central bank decisions, since short-maturity government yields tend to track policy rates such as the Effective Federal Funds Rate or the midpoint of the Federal Reserve’s target range. A Taylor-rule approximation links policy decisions to inflation and the output gap, offering a macroeconomic framework for forecasting policy moves.

It also describes a statistical alternative: an autoregressive model of order one, or an Ornstein–Uhlenbeck process, which can capture the persistence commonly seen in short rates. The cited paper reportedly finds that the current policy rate and employment are strong predictors of the Effective Federal Funds Rate. Federal funds futures offer another source of expectations, though their prices reflect risk-neutral rather than necessarily real-world expectations. The note gives a high-level overview rather than a forecast specification or a comparison of predictive performance across regimes.

Key ideas

• Short-term rates are strongly influenced by central bank policy actions.
• A Taylor-rule approximation relates policy settings to inflation and the output gap.
• Persistent short rates can be modeled with an AR(1) process or an Ornstein–Uhlenbeck process.
• The cited study identifies the current policy rate and employment as useful predictors of the Effective Federal Funds Rate.
• Federal funds futures reflect risk-neutral market expectations, which may differ from real-world expectations.

Tags

# What is the short rate a function of?


# What is the short rate a function of?












According to the expectations hypothesis we get the long rates via the expected short rates and then adding a term premium.

One therefore needs to consider how the short rate might change. I suppose the central bank essentially set this rate. This is done with the purpose to steer the economy.

To get an idea of how the short rate might evolve one might then consider the macro outlook right? i.e the future short rate is a function of growth in GDP, employment and inflation?

Is this the right idea?

## Answer by fes (score 1, accepted)

https://quant.stackexchange.com/a/60021

Yes short term rates tend to be highly dependent on central bank actions. For example the short end of Treasury yield curve is highly correlated with the Effective Federal Funds Rate (FFR) or the midpoint of the target range set by Fed. One way to forecast short rates would therefore be to predict CB actions. CB policy functions on the other hand are often approximated by a Taylor rule that is assumed to depend on inflation and output gap.

From a more econometric angle a standard model for short rates is an AR(1) / Ornstein-Uhlenbeck-process. This works fairly well because short rates tend to be very persistent. This paper studies the performance of different variables in predicting the FFR. It finds that the current FFR and employment are the best predictors. FFR futures can also be used with the caveat that they represent risk neutral market expectations.

Shown in full with attribution under the source's licence. Licence: CC BY-SA 4.0 (Stack Exchange)

This summary was written by Stratmill's research agent from the original; it is not a copy of the source.