# Machine Learning for Yield Curve Feature Extraction: Application to Illiquid Corporate Bonds

## TL;DR

This paper introduces a machine learning approach using Denoising Autoencoders to extract features from illiquid corporate bond yields by leveraging a surrogate liquid market, demonstrating superior performance in predicting missing yield parameters.

## Contribution

It applies DAE algorithms to yield curve feature extraction in illiquid markets, combining techniques from image processing to improve yield prediction accuracy.

## Key findings

- DAE outperforms point-in-time inpainting algorithms on unobserved yield surfaces
- Learned features improve yield curve reconstruction in illiquid markets
- Method demonstrates potential for better risk assessment in fixed income

## Abstract

This paper studies an application of machine learning in extracting features from the historical market implied corporate bond yields. We consider an example of a hypothetical illiquid fixed income market. After choosing a surrogate liquid market, we apply the Denoising Autoencoder (DAE) algorithm to learn the features of the missing yield parameters from the historical data of the instruments traded in the chosen liquid market. The DAE algorithm is then challenged by two "point-in-time" inpainting algorithms taken from the image processing and computer vision domain. It is observed that, when tested on unobserved rate surfaces, the DAE algorithm exhibits superior performance thanks to the features it has learned from the historical shapes of yield curves.