Network Productivity and Loan-Income Adjustment after Monetary Policy Normalization: Evidence from Japanese Regional Banks
This study utilizes a multi-period network DEA model to analyze 96 Japanese regional banks from 2019 to 2025, revealing that while global productivity increased following monetary policy normalization, the widening gap between loan-to-deposit and loan-yield growth rates in 2024–2025 primarily reflects stronger loan-income performance driven by repricing and portfolio shifts rather than independent operational efficiency gains.
Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer
For decades, Japan's central bank kept interest rates at or below zero, a policy that made money cheap and abundant but squeezed the profits of local banks. These regional banks act as the vital middlemen of the economy: they take money from savers in the form of deposits and turn it into loans for local businesses and families. In a world of near-zero rates, the profit margin on these loans is razor-thin, forcing banks to operate with extreme efficiency just to survive. When the Bank of Japan finally ended this era of negative rates in March 2024, it set off a chain reaction. The question for economists was not just whether banks would make more money, but how the two distinct parts of their job would react. One part is the act of lending itself—converting deposits into loans. The other part is the act of earning income from those loans, which depends on the interest rates charged. Because existing loans take time to adjust to new rates while new loans can be priced immediately, these two activities might not move in sync. Understanding this lag is crucial, as it reveals whether banks are simply collecting higher fees or actually becoming better at their core business of connecting savers to borrowers.
A researcher at Soka University, Hirohisa Shimura, set out to watch this adjustment unfold in real time. He examined the financial records of 96 regional banking entities across Japan, tracking them from the fiscal year 2019 through 2025. To do this, he used a sophisticated analytical tool that treats a bank not as a single black box, but as a two-step factory. In the first step, the bank uses deposits to create loans. In the second step, it uses those loans to generate interest income. By separating these stages, the study could see if the bank was getting better at making loans, getting better at earning money from them, or if both were happening at different speeds. The analysis covered the period just before the policy change, the year the change happened, and the two full years that followed, allowing the researcher to compare how the banks performed before and after the shift in monetary policy.
The results showed a clear and surprising pattern. Before the interest rates began to rise, the two stages of banking moved in lockstep; when lending activity improved, income generation improved by a similar amount. However, once the policy normalized, a gap began to open between the two. In the first year after the change, the difference was small. But in the second year, the gap widened significantly. The banks did not necessarily become better at the physical act of lending; the volume of loans relative to deposits grew only slightly. Instead, the dramatic improvement came from the second stage: the income generated per loan jumped sharply. This suggests that the banks were not necessarily processing more business, but were successfully collecting higher interest payments on the loans they already held or issued.
This divergence happened across almost every bank in the study, with 91 out of 96 entities showing a larger boost in income generation than in lending volume. The effect was not uniform across the country, however. In the second year after the policy shift, the gap between lending and earning performance varied noticeably by region. Some areas, like the Tokai and Kinki regions, saw a massive widening of this gap, while others, such as Chugoku-Shikoku, saw a much smaller change. Interestingly, the traditional classification of banks into "Regional Bank I" and "Regional Bank II" did not explain these differences; both types of banks reacted in almost the same way. The variation seemed tied more to local market conditions than to the size or type of the bank itself.
The study offers a careful, descriptive look at what happened, rather than a definitive proof of cause and effect. The researcher notes that the sharp rise in income per loan could be due to several factors: existing loans being repriced at higher rates, new loans carrying higher interest, or a shift in the types of loans the banks held. The data used in the study cannot separate these possibilities. What is clear is that the overall productivity of these banks increased, but that increase was driven almost entirely by the earnings side of the equation, not by a surge in lending activity. This finding highlights that when a major economic policy shifts, the different parts of a financial system do not adjust at the same speed. The first year may show a modest reaction, but the full picture of how banks adapt to a new normal often only becomes visible in the second year, revealing a complex rebalancing where earning power outpaces the volume of lending.
Drowning in papers in your field?
Get daily digests of the most novel papers matching your research keywords — with technical summaries, in your language.