Why Profitability Does Not Automatically Become Credit
The lending results show that a few dominant banks can move Mauritania’s aggregate credit through bank-specific lending changes. This chapter asks whether their earnings shocks behave the same way. The comparison matters because a bank can become more profitable without having a larger loan pipeline, and a monthly null result should not be mistaken for proof that profits can never support lending. The task is to separate what the paper measures from the balance-sheet pathways it considers plausible.
Profitability moves with the macro background
The profitability first stage begins with the same general logic as the lending exercise: remove common macroeconomic conditions, selected bank-level characteristics, and persistent differences across banks, then use the residual as a bank-specific profitability innovation. The outcome is not a vague sense that a bank is doing well. It is net income relative to total bank capital, observed at the bank-month level.
Eight fixed-effects specifications are used. They include inflation, the policy rate, M3 growth, and exchange-rate changes, along with bank variables such as liquidity, lending growth, borrower concentration, depositor concentration, and the lending–deposit spread in different combinations. The results are notably different from the lending first stage. Inflation coefficients are positive and significant across the specifications, roughly 2.13 to 2.25. Policy-rate coefficients are also positive and significant, roughly 1.48 to 1.78. Lending growth, liquidity, concentration measures, and the lending–deposit spread are generally imprecise.
The fit statistics tell the same story. Overall R-squared ranges from about 0.443 to 0.465, and within-bank R-squared from about 0.102 to 0.137. These values are higher than the lending models’ within-bank fit reported earlier. In this sample, observable macro conditions and bank characteristics organize more of the month-to-month movement in profitability than they organize the month-to-month movement in lending growth.
That pattern is consistent with several ways earnings can respond quickly to the financial environment. Inflation can change nominal revenues and costs. A policy-rate movement can pass through to asset yields or pricing. Fees, valuation effects, and the performance of concentrated loan exposures can alter net income without creating a corresponding increase in new lending. The paper discusses these interpretations, but the positive coefficients do not establish that inflation or monetary tightening causes every profitability movement, nor do they identify one mechanism as dominant.
The distinction is economically important. A macro-sensitive profitability measure can rise because existing assets produce more income or because prices and margins change. New credit requires another decision: the bank must find or approve a borrower, allocate funding, accept a risk, and carry the loan on its balance sheet. Those steps may occur later, or not at all, even when the income statement improves in the current month.
The profitability residual therefore answers a narrower question than “is the bank earning more?” It captures an earnings movement left after the study’s selected controls. The researchers then weight it by lagged bank size and ask whether it adds information about aggregate credit growth. If profitability is the missing bridge between bank success and intermediation, a size-weighted profitability shock should improve the aggregate credit equation in a way comparable to the lending shock.