2023-07-19-美联储-Modigliani-Miller偏差的综合含义_一种充分的统计学方法_76页_608kb

Aggregate Implications of Deviations from Modigliani-Miller: A Sufficient Statistics Approach Summary

This paper introduces a novel methodology to quantify the aggregate macroeconomic effects of firm-level investment distortions, such as costly external financing or agency costs, without fully microfounding the distortions. The approach leverages two sufficient statistics—the average capital-weighted distortion to the investment rate and its mean squared value—to bound or exactly identify the effects in a general equilibrium model featuring heterogeneous firms, endogenous entry, and capital adjustment costs.

Key findings include:

• GE Effects vs. Partial Equilibrium: Resolving distortions yields markedly smaller aggregate benefits in general equilibrium (GE) compared to partial equilibrium (PE) estimates due to dampening forces like labor supply adjustments and entry effects. For example, a PE change in output of 3.35% shrinks to <1% in GE for external financing frictions.
• Welfare Costs: Removing an external financing friction increases welfare by ~0.13% (GE bounds), while a manager-shareholder overinvestment friction reduces welfare by ~0.04% (GE bounds) due to interactions with monopoly markups and taxes.
• Friction Type Matters: Underinvestment (external financing) boosts welfare, while overinvestment harms it. GE dynamics (e.g., entry/exit) significantly alter the quantified impact.
• Practical Value: The method allows researchers to use widely available empirical estimates from corporate finance (e.g., underinvestment elasticities) to assess macroeconomic policies without tractable full models, highlighting the sensitivity of aggregate outcomes to microeconomic dynamics through sufficient statistics.

This framework enhances the ability to evaluate policy interventions by bridging micro-founded frictions with aggregate outcomes, accounting for real-world complexities like firm heterogeneity and entry dynamics in a computationally tractable way.