Why Energy Costs Hit Household Credit Harder Than Commercial Loans

Energy shocks can travel far beyond inflation and production costs, eventually reaching household repayment capacity, business revenues and the quality of bank loan portfolios. Still, those pressures may remain difficult to detect when financial risk is measured only through economy-wide lending indicators.

New research shows how sharply the transmission can diverge between borrowers. The study, "Banking-Sector Credit Risk Under Energy Price Shocks: A Borrower-Specific Nonlinear ARDL Analysis of Non-Performing Loans in an Emerging Market," by Mehmet Şuayb Yıldırım and Ismail Onur Baycan, published in the Journal of Risk and Financial Management, uses Türkiye as an emerging-market case to track how increases and decreases in energy prices move through household and commercial credit.

Using monthly banking and energy-price data from January 2005 to June 2026, the researchers find an unusual split: household credit quality deteriorates when real consumer energy prices increase but does not recover correspondingly when they fall, while commercial credit responds more strongly to declining consumer energy prices. Once the portfolios are combined, much of this directional behaviour disappears from the headline non-performing loan ratio.

Household borrowers absorb the shock first and recover slowly

Households have limited room to adjust when essential energy costs rise. Electricity, gas and other household energy expenses draw directly on disposable income, leaving less cash available for loan instalments, credit-card payments and other obligations.

The study estimates that a typical year of real consumer energy-price increases is associated with an increase of about 0.6 percentage points in the household NPL ratio from its sample mean. Larger increases during shock years are associated with considerably greater deterioration. The asymmetry is crucial. Falling real energy prices do not restore household credit quality at the same rate. Once repayments are missed and a loan enters non-performing status, accumulated arrears, damaged credit records and lengthy resolution processes make reversal difficult.

Relief also arrives differently from the original shock. A sharp energy-price increase immediately squeezes the monthly budget, while a decline in the real cost of energy can occur gradually if household tariffs rise more slowly than general inflation. This improves purchasing power over time, but it does not necessarily generate enough immediate cash to cure an already distressed loan.

The study describes this pattern as persistent rather than permanent. Adjustment back toward equilibrium is slow, meaning the effect of a price shock can remain visible in household credit quality well after the initial movement in energy prices.

Commercial loans respond to household purchasing power, not simply input costs

Commercial credit produces a contrasting result when measured against consumer energy prices. Price increases show no significant long-run association with the commercial NPL ratio, while cumulative declines are associated with lower commercial credit risk.

A 1% cumulative decline in the real consumer energy price is associated with an estimated 3.54% reduction in the commercial NPL ratio. The authors treat this coefficient cautiously because the commercial models are more fragile across alternative specifications than the household model.

The more revealing result concerns which price measure carries the association. Producer energy prices do not show the same asymmetric relationship with commercial credit. When consumer and producer energy prices are included together, the commercial response remains attached to the consumer measure.

This shifts the interpretation away from firms' own energy bills. The study argues that lower household energy costs can release purchasing power, supporting demand for goods and services and strengthening firm revenues. Improved cash flow may then help firms service debt, restructure liabilities or return distressed loans to performing status.

The data do not directly track household spending into firm sales, so the demand channel remains an interpretation rather than a demonstrated causal sequence. Even so, the result is consistent with a commercial portfolio that is closely exposed to domestic consumption.

Aggregate NPL ratios can hide concentrated financial stress

The total banking-sector NPL ratio does not show significant asymmetry against any of the seven energy indicators examined. The disaggregated models show why that apparent stability can be misleading.

Household credit deteriorates mainly when consumer energy prices rise. Commercial credit improves mainly when those prices fall. Combining the two portfolios can therefore offset much of the directional response visible at the borrower level.

The researchers test this directly by estimating household and commercial equations jointly against the same real consumer energy-price measure. The two asymmetry gaps have opposite signs, and the hypothesis that they are equal is rejected statistically, including under block-bootstrap testing.

The result has a clear supervisory implication. A sector-wide NPL ratio can remain comparatively stable while one part of the loan book is becoming more vulnerable. Aggregate monitoring may therefore understate the financial consequences of an energy shock if portfolio composition is ignored.

The study also finds that a residual segment of broader corporate and commercial lending does not display the same asymmetry. This reinforces the case for looking beneath the banking-sector average rather than assuming a common response across borrowers.

Financial stability policy needs to look at timing, targeting and portfolio exposure

The findings support a stronger role for consumer energy prices in macroprudential monitoring. Stress tests commonly include growth, interest rates, exchange rates and other macroeconomic variables, but the study suggests that sharp increases in household energy costs can add independent information about credit deterioration.

Timing is especially important for household borrowers. Support delivered before repayment problems become entrenched operates differently from assistance provided after loans have already crossed into non-performing status. The study, hence, gives particular weight to early intervention for indebted households facing a severe energy-price shock.

Longer-term measures address a different source of vulnerability. Residential energy-efficiency investment can reduce exposure to recurring energy costs and weaken the link between energy-price movements and household liquidity. Tariff or income support, by contrast, primarily eases the immediate cash-flow constraint.

The paper does not rank these instruments. Untargeted subsidies can impose fiscal costs and weaken conservation incentives, while targeting itself carries administrative costs. Loan restructuring may improve repayment capacity, but regulatory forbearance can also delay recognition of underlying credit risk.

Commercial policy requires equally careful interpretation. The study finds no strong evidence that broad reductions in producer energy costs would materially improve aggregate commercial delinquency. That does not mean energy-intensive firms are unaffected; sector-specific stress may simply disappear when many different firms are combined in one portfolio.

The analysis also places firm limits on what can be concluded. The long-run elasticities are conditional because evidence for a stable level relationship is not equally strong across models. The commercial result is weaker than the household finding, and the time-series design identifies associations rather than causal effects.

The data operate at portfolio level, so the study cannot observe which household income groups, debt burdens or firm types drive the results. Credit-register data, household budget information and sector-level firm data would be needed to identify those mechanisms directly.

Cross-country testing would also be important because the findings depend on features of Türkiye's economy: household liquidity constraints, administered retail tariffs, inflation, domestic-demand exposure and the speed of loan resolution. Countries with different structures may produce weaker, stronger or entirely different asymmetries.