Forecasting Corporate Bond Index Returns
This paper forecasts US corporate bond returns using 180 firm characteristics and 65 macroeconomic variables via shrinkage techniques.
What it examines
This paper examines how combining firm-level characteristics and macroeconomic variables can forecast U.S. corporate bond index returns using shrinkage methods including PLS, PCA, and IWC. It aims to improve asset allocation by delivering statistically and economically significant predictions for both investment-grade and high-yield bond indexes.
What it concludes
The study shows that integrating numerous firm characteristics and macroeconomic data yields robust forecasts for corporate bond returns. Its findings enable improved asset allocation and trading strategies, highlighting key predictors such as profitability and liquidity. Future research may refine these techniques for broader market applications and risk management.
Evidence objects
Researchers integrated 180 firm-level characteristics and 65 macroeconomic variables to forecast U.S. corporate bond returns, using dimension reduction methods like $$\text{PLS}$$ and $$\text{IWC}$$ that achieved exceptional out-of-sample accuracy and performance.
key_findings bullet 1 · key_findings · validation V0
A comprehensive integration of predictors revealed aggregate profitability, leverage, growth, liquidity, and default risk forecasts for both investment-grade and high-yield bonds, while variable importance projection and naive averaging reduced overfitting.
key_findings bullet 2 · key_findings · validation V0
Robust statistical tests, multivariate regressions, and out-of-sample evaluations confirmed results despite concerns over market regime shifts and limited external applicability, urging research to optimize fixed income predictability and asset allocation.
key_findings bullet 3 · key_findings · validation V0
Focusing on corporate bond index returns, the paper addresses an overlooked fixed income niche by proposing asset allocation strategies enriched with robust empirical insights. Utilizing a vast set of predictors ($245$ measures) and shrinkage techniques, its innovative quantitative finance perspective captures interest despite largely extending established methods with moderate originality.
key_findings bullet 4 · key_findings · validation V0
Raw abstract and provenance
- … Second, our paper contributes to the literature that investigates how cross-sectional characteristics can be used to predict the time series of market returns. Recent studies …
Source row: 890 · abstract type: snippet