A Python/Jupyter notebook project to understand the Yield Curve and its potential for forecasting a recession. Yield curve rates between 1990 and present are from the U.S. Department of the Treasury
- Git clone the repository
- Create a virtual environment and run
$ pip install -r requirements.txt - Update the
docker-compose.ymlvolumes to point to your local instance and cd to where the repo is stored - Run the following:
$ docker-compose build $ docker-compose up
You can access the project at http://localhost:8888/notebooks/yield%20curve.ipynb and you'll need to provide a token that appears in the docker logs. To bring down the project, run $ docker-compose down and to destroy the containers, run $ docker-compose kill.
A deeper dive into U.S. Department of the Treasury Yield Curve data and it's predictive capaiblities.
In an efficiently performing market, long-term bonds have higher bond yield rates than shorter-term bonds, T-notes, and T-bills as the market expects greater risk in investing in long-term bonds (a lot can happen in 30 years). However, when the yield curve inverts, the bond yield rates for shorter-term bonds are higher than long-term bond yield rates. An Inverted Yield Curve is used as one predictor of a recession as it captures the nervousness of investors about the near term market outlook.
In my analysis, an Inverted Yield Curve occurs when the ratio of long-term bond rates (i.e. 30 years, 10 years) versus short-term bonds (6 months, 1 year, 3 years, etc.) is between 0 and 1. The yield curve last inverted between 2006 and 2007.
The ratio of 10 year bonds/6 month bonds and 10 year/1 year bonds inverted in May 2019
