The Financial Accelerator Mechanism: Time-Varying Frequency-Dependent Evidence From the C&I Loan Market

Empirical identification of the financial accelerator mechanism using aggregate data has proven difficult due to measurement challenges and the unobservability of the external finance premium, defined as the wedge between the cost of external finance and the opportunity cost of internal funds. We address this challenge by developing a novel empirical strategy which exploits the frequency-dependent properties of the financial accelerator and uses survey-based credit standards—rather than interest rate spreads—as a proxy for the external finance premium. The time-frequency analysis of U.S. commercial and industrial loan market data—comprising loan growth, rate spreads, credit standards, and aggregate net worth—over a period marked by significant regulatory and structural changes in the financial system, provides three main results. First, credit standards provide a more informative proxy for the external finance premium than loan rate spreads. Second, the relationship between net worth and credit conditions is frequency-dependent, with the strongest effects concentrated at business cycle frequencies (4–10 years). Third, the financial accelerator becomes empirically relevant only after the early 1990s, coinciding with the onset of financial deregulation and increasing agency frictions. These findings suggest that the strength and observability of the financial accelerator mechanism are both frequency- and time-dependent, becoming more pronounced in the post-deregulation era of U.S. financial history.