Skip to content

as-cb/bond-index-replication-python

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

17 Commits
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Dynamic Bond Index Replication

Replicating Bloomberg’s US Corporate Bond Index using the TRACE transaction dataset.
This project constructs investable portfolios of 50–100 bonds out of ~6,000 investable bonds using XGBoost and Kalman Filters, and evaluates out-of-sample tracking performance.


Key Results

Model OOS Tracking Error OOS Benchmark Correlation
XGBoost 2.01% 0.808
Kalman Filter 2.60% 0.935

Data availability

Large files are NOT uploaded to this GitHub repository due to size limits. Download the datasets from the Google Drive links below and place them in the specified folders.

About

This project aims to replicate Bloomberg's US Corporate Bond Index using XGBoost and Kalman Filters. Tracking error has been used to quantify model performance.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors