Cross-section without factors: a string model for expected returns

Area of Science:

• Quantitative Finance
• Asset Pricing Theory
• Financial Econometrics

Background:

• Traditional asset pricing models often rely on common factors to explain expected returns.
• The need for alternative models that capture complex inter-asset relationships is growing.
• Existing models may not fully account for the interconnectedness of asset returns.

Purpose of the Study:

• To develop a novel asset pricing model based on a "string" concept, linking asset returns through correlations.
• To investigate the role of granular exposure and correlation premiums in asset pricing.
• To identify unique properties of large stocks within this new framework.

Main Methods:

• Formulation of a new asset pricing model incorporating a "string" of asset returns.
• Application of no-arbitrage restrictions to define expected returns based on cross-asset exposures.
• Analysis of the model's predictions for large stocks and their hedging properties.

Main Results:

• The proposed "string" model posits that expected returns are influenced by an asset's exposure to all other asset returns, termed a correlation premium.
• Large stocks exhibit higher connectivity during market downturns.
• These large stocks function as correlation hedges, negatively contributing to the correlation premium.

Conclusions:

• The "string" model offers a new perspective on asset pricing, emphasizing inter-asset correlations.
• Large stocks play a crucial role as hedges, potentially reducing portfolio risk.
• The model demonstrates competitive performance compared to established linear factor models.