Beyond Reserves: Why Some Banks Still Struggle to Lend Despite Ample Liquidity

In a groundbreaking study conducted under the European Central Bank’s Lamfalussy Fellowship Programme, Felix Corell of Vrije Universiteit Amsterdam and its School of Business and Economics presents a compelling new way to think about banking behavior. His research, titled “Hand-to-mouth banks: deposit inflows and the marginal propensity to lend,” introduces the concept of the Marginal Propensity to Lend out of Deposits (MPLD), a metric that mirrors the Marginal Propensity to Consume (MPC) used in macroeconomics to describe liquidity-constrained households. Just as low-income households tend to spend any unexpected income immediately, some banks, Corell argues, operate “hand-to-mouth” and lend out nearly every dollar of unanticipated deposit inflows. This idea becomes the cornerstone of a richly detailed analysis of how U.S. banks react to funding shocks and how monetary policy interacts with their liquidity conditions.

Before and After QE: A Dramatic Shift in Bank Lending Behavior

Using a novel identification strategy that leverages county-level IRS dividend income data as a proxy for deposit shocks, Corell analyzes the behavior of over 13,000 U.S. banks from 1994 to 2021. Before the onset of Quantitative Easing (QE) in 2008, the average MPLD was strikingly high, about 1.1. This means that for every unexpected dollar deposited, banks lent out more than a dollar, often by activating additional funding sources. It was a sign of a system operating under tight liquidity constraints, especially in real estate lending.

Post-2008, however, the environment changed drastically. The Federal Reserve’s QE programs injected trillions into the banking system, and as a result, the average MPLD dropped sharply to 0.35. While this reduction signals that liquidity constraints had loosened, the still-positive MPLD indicates that many banks remained at least partially constrained. This shift in behavior maps directly onto changes in monetary policy regimes, highlighting the power of central bank actions not just to expand balance sheets but to alter the lending culture itself.

Not All Banks Are Created Equal

One of the study’s most important findings is the significant variation in MPLDs across banks. Some institutions remain highly reactive to deposit inflows, while others barely budge. The determinants of this heterogeneity are as revealing as they are policy-relevant. Banks with higher cash-to-asset ratios, a signal of stronger liquidity buffers, tend to have lower MPLDs, reflecting fewer funding constraints. Similarly, banks with greater deposit market power, as measured by the Herfindahl-Hirschman Index (HHI), also exhibit lower MPLDs, likely because they operate in less competitive environments where deposits are stickier and easier to manage.

Conversely, banks with higher deposit betas, meaning they must raise rates more to retain deposits when interest rates rise, have higher MPLDs. These institutions, more exposed to funding flight, respond more aggressively to deposit inflows. The picture is more complex when it comes to leverage. Moderately leveraged banks tend to lend more when deposits rise, but highly leveraged ones may cut lending to avoid regulatory breaches. This leads to a non-linear relationship between leverage and MPLD, pointing to a delicate balance between capital adequacy and credit growth.

How Banks Use Extra Cash: It’s Not Just About Mortgages

A detailed breakdown of MPLDs by loan type reveals further insights. Before QE, banks transformed deposit inflows into new loans across various categories, real estate, commercial and industrial, consumer, and other loans, in roughly the same proportion as their existing portfolios. But after QE, this proportionality vanished. The decline in MPLD was almost entirely driven by a drop in real estate lending, suggesting that the easy liquidity environment changed banks’ preferences or constraints for mortgage financing. One possible explanation lies in the U.S. mortgage market’s structure, where banks often securitize or offload loans to government-sponsored enterprises, making them less inclined to expand real estate lending with marginal funding.

Interestingly, some banks exhibited MPLDs greater than one, suggesting that deposit inflows not only funded new lending but may have attracted or replaced other funding sources. On the other end, some banks had negative MPLDs, implying that unsolicited deposit inflows led to a reduction in lending, likely a result of leverage constraints that forced balance sheet adjustment.

Why Policymakers Should Care About MPLD

Corell’s findings carry weighty implications for monetary policy and financial stability. In a system rich with reserves, central banks often assume that further injections or withdrawals of liquidity have marginal effects. But this research shows that distribution matters. “Hand-to-mouth” banks with high MPLDs react strongly to liquidity changes, while reserve-rich banks may not respond at all. In a tightening cycle, this could mean that QT (Quantitative Tightening) disproportionately hits the most constrained institutions, curbing credit more than anticipated.

The study also enters the heated debate around central bank digital currencies (CBDCs). If retail CBDCs draw deposits away from traditional banks, high-MPLD institutions, those that depend on deposits to sustain lending, would be most affected. Corell’s framework allows policymakers to identify and monitor these vulnerable banks, helping design CBDCs that avoid unintended disintermediation of credit.

The Marginal Propensity to Lend out of Deposits is more than just an academic concept. It’s a powerful diagnostic tool for understanding the real-time mechanics of banking, liquidity, and credit supply. Felix Corell’s paper opens a new frontier in thinking about how banks behave under different monetary regimes and how central banks can tailor their interventions in an increasingly complex financial ecosystem.