When Losses Turn into Loans: Blattner, Farinha & Rebelo (2023)

What this is. The paper’s core results, the natural-experiment identification design, the loan-loss underreporting algorithm, and the firm-level empirical specifications with their defining equations: enough to know what it found and how, without reading all 42 pages. To replicate or extend it, read the full source at the original.

TL;DR

Distressed banks respond to ratio-based capital shortfalls not only by cutting overall credit but by distorting the composition of credit supply: they reallocate lending toward firms whose loan losses they have been underreporting, thereby delaying the recognition of those losses and protecting their reported capital ratios. Using the October 2011 European Banking Authority (EBA) capital exercise as a natural experiment affecting a subset of large Portuguese banks, Blattner, Farinha, and Rebelo develop a bunching-based algorithm to measure loan-loss underreporting at the monthly firm-bank level, show that exposed banks increase credit supply to underreported borrowers by about 2 percentage points per quarter while cutting credit to all other firms, and trace this credit reallocation through to a widening of capital and labor wedges that accounts for roughly 22 percent of the decline in aggregate allocative efficiency in Portugal in 2012.

The paper extends the Japan zombie-lending evidence of Peek and Rosengren (2005) to a causal quasi-experimental setting focused on credit composition. It extends the zombie-lending framework of Caballero, Hoshi, and Kashyap (2008) by measuring underreporting and linking it to input misallocation. Schivardi, Sette, and Tabellini (2022) report no TFP effects of zombie lending in Italy; this study estimates allocative-efficiency losses in Portugal, so the outcomes and settings differ.

Core results

Magnitudes and significance are as reported; \*\*/\*\*\* = 5%/1%. Locators point into the source PDF.

#	Result	Locator	Magnitude
R1	Exposed banks increase credit to underreported firms, cut credit to all others: triple-interaction coefficient positive and significant during EBA, negative and significant for baseline group	Figure 3 Panel A, p. 1615; Table A2, online appendix (point estimates)	+2 pp quarterly credit growth for underreported firms at exposed banks; -2 pp for all other relationships at exposed banks (each ~4% of 1 SD of credit growth)
R2	Firm-level credit reallocation is real and not undone by substitution: total credit rises for underreported firms with high exposure to EBA banks, falls for all others	Figure 5 Panel A, p. 1620; text p. 1619	+16% cumulative credit for underreported firms borrowing entirely from exposed banks; -14% for all other firms relative to base quarter 2011:III
R3	Credit shock transmits to labor: IV elasticity of labor w.r.t. credit supply is large and significant	Table 5 Panel B col. 2, p. 1625	Elasticity = 0.52 (SE 0.094); first-stage F = 111.2
R4	Credit shock transmits to capital: IV elasticity of capital w.r.t. credit supply is significant	Table 5 Panel B col. 4, p. 1625	Elasticity = 0.14 (SE 0.046); first-stage F = 111.2
R5	Total EBA intervention caused large allocative-efficiency loss: aggregating all firm-level wedge changes explains majority of 2012 AE decline	Table 6 Panel A col. 1, p. 1628	Total estimated AE effect: -6.59% = 54% of actual -12.24% within-sector AE decline in 2012
R6	Credit reallocation to underreported firms alone accounts for ~22% of the AE decline: reallocation component isolated via simulation	Table 6 Panel B col. 1, p. 1628	Mean -2.71% AE (range -0.89% to -4.43%); mean = 22% of actual AE decline
R7	Bunching validity: underreporting is statistically higher in overdue buckets immediately before a jump in the regulatory deduction rate, confirming strategic behavior	Table B1 Panel A, p. 1636	Coefficients 0.014-0.451 across collateral types and increment sizes (all positive and significant); placebo using the other collateral type’s rate increment yields negative coefficients
R8	Underreported firms have lower marginal products than performing firms: correctly reported nonperforming firms are closer to underreported firms	Table 4 Panel A, p. 1622	MRPL: 37.30 vs. 48.24 (thousand euros); MRPK: 47.44 vs. 63.16 (percent); correctly reported nonperforming firms: 42.39 and 55.51
R9	Firm-level wedges are persistent	Table 4 Panel B, p. 1622	Lagged-wedge coefficient: 0.786 (SE 0.002) for labor and 0.594 (SE 0.002) for capital
R10	Exposed banks increase underreporting during the EBA intervention, then roll it back	Figure 4, p. 1617	Underreported losses scaled by 2010 bank capital rise after the announcement and recede at the EBA deadline; the paper reports the pattern graphically
R11	Risk-shifting evidence is absent: exposed banks’ new lending falls more for higher-risk new clients, while lending to existing clients does not increase	Table 3, p. 1618	New-client EBA × exposed × risk coefficients: -3.375 (SE 1.274) for predicted default risk and -0.157 (SE 0.053) for sales cyclicality; existing-client approval: 0.002 (SE 0.004); new-loan estimates: -0.024 (SE 0.008) and -0.022 (SE 0.009) in the two specifications
R12	Credit has no detectable effect on firm-level TFP	Table 5 Panel B col. 5, p. 1625	IV coefficient 0.001 (SE 0.005); first-stage F = 111.2
R13	Input-wedge responses vary with prior financial constraints	Table 5 Panel A, p. 1625	Share EBA × 2011 capital wedge: labor-wedge coefficient 0.012 (SE 0.008); capital-wedge coefficient 0.028 (SE 0.013)
R14	Credit reallocation changes firm input use: underreported firms expand inputs while other groups contract	Text p. 1624	Underreported firms: labor +6%, capital +9%; non-distressed firms: labor -6%, capital -8%; correctly reported distressed firms: labor -4%, capital -6%
R15	Credit results show no differential pre-trends for baseline and underreported firm-bank relationships	Figure 3 and text p. 1616	No differential credit allocation in the two pre-intervention periods; no coefficients reported in the article text
R16	Firm liquidity measures show no pre-trends before the intervention	Text p. 1621 (Figure A7, online appendix cited)	No pre-trends in current ratio or cash/assets; numerical estimates not reported in the article text
R17	Pre-intervention placebo estimates show no significant effects	Text p. 1625 (Table A5, online appendix cited)	No significant effects in the pre-intervention years; numerical estimates not reported in the article text
R18	The credit reallocation is specific to underreported relationships among overdue borrowers	Figure 3 Panel B and text p. 1615	Exposed banks reduce credit to overdue but non-underreported relationships, while credit increases only for underreported relationships; point estimates are in online Appendix Table A2

Overall (paper’s conclusion). Ratio-based capital requirements create distorted lending incentives when banks are already in distress: exposed banks intensify loss underreporting and roll over credit to underreported borrowers to avoid booking additional losses. This credit misallocation prevents inputs from being reallocated to their highest-value uses, widening the dispersion of capital and labor wedges and contributing meaningfully to aggregate productivity decline.

Theory / model

The paper has no formal equilibrium model. Its tested mechanism is that ratio-based capital shortfalls make already-distressed banks reluctant to recognize losses: they preserve lending to borrowers whose losses were underreported, while cutting other credit. The paper tests this against risk-shifting and demand explanations, then traces credit reallocation into factor use and allocative efficiency.

EBA capital exercise and identification. The October 2011 EBA exercise required affected banks to meet a 9 percent Core Tier 1 threshold, net of the sovereign-debt buffer (text p. 1610):

$$\frac{\text{Core Tier 1} - \text{sovereign debt buffer}}{\text{RWA}} \geq 0.09.$$

Exposure combines EBA eligibility with above-median sovereign holdings among eligible Portuguese banks. The comparison is eligible banks with below-median holdings plus other Portuguese commercial banks. The identifying assumption is no unobserved difference between exposed and comparison banks that independently drives credit allocation during the intervention. Evidence includes pre-period balance, parallel credit and liquidity trends, firm-quarter fixed effects in the main firm-bank design, and null risky-lending evidence (pp. 1611-1621).

Input wedges. For sector $s$, labor and capital distortions are defined by equations (4)-(5), p. 1622, where $Y_{it}$ is value added, $L_{it}$ employment, $K_{it}$ capital, and the wedges measure gaps between marginal products and factor user costs:

$$\alpha_s \frac{Y_{it}}{K_{it}} = (r_t + \Delta_t)(1 + \tau_{it}^{K}). \tag{4}$$ $$\beta_s \frac{Y_{it}}{L_{it}} = w_t(1 + \tau_{it}^{L}). \tag{5}$$

Underreported firms have lower mean marginal products than performing firms (Table 4 Panel A, p. 1622). The authors’ causal claim is that directing credit toward these firms prevents resources from moving to higher marginal-product uses, increasing within-sector wedge dispersion.

Method

Loan-loss underreporting measure. The paper uses monthly firm-bank overdue balances by regulatory bucket. Inflows into a bucket and outflows from the preceding bucket adjust the observed change; equation (1), p. 1607, defines excess mass:

$$E(t;k) = [B(t;k) - IN(t;k)] - [B(t-1;k-1) - OUT(t;k-1)]. \tag{1}$$

The measure is Markovian: it flags a discrepancy relative to the prior month, not the true age of a loan. For multi-month reporting buckets, Appendix A’s flow-adjusted monthly-bucket expression is equation (A3), p. 1632:

$$E(t;k) = [C(t;c) - IN(t;c)] - [C(t-1;c-1) - OUT(t;c-1)]. \tag{A3}$$

For the validity test, the share of excess mass in overdue loans is regressed on deduction-rate jumps; equation B1 as printed includes bank, firm, and month fixed effects and firm-bank clustered standard errors (pp. 1635-1636). The accompanying text and Table B1 notes instead describe firm-bank fixed effects, an internal specification inconsistency in the PDF; the collateral-type samples are estimated separately:

$$\frac{\text{excess mass}_{ibkct}}{\text{overdue loans}_{ibkct}} = \sum_{j=1}^{5} \beta_j \Delta \text{deduction rate}_j + \phi_b + \theta_i + \mu_t + \epsilon_{ibkct}. \tag{B1}$$

Productivity decomposition. The accounting decomposition following Osotimehin (2019) separates aggregate productivity change into technical efficiency and within- and between-sector allocative efficiency (equation 7, p. 1625):

$$\Delta \ln TFP \simeq \Delta TE + \Delta AE_{\text{within}} + \Delta AE_{\text{between}}. \tag{7}$$

The sector-weighting formula for technical efficiency and the within-sector component are equations (8)-(9), p. 1626. Here $s_t^x$ is sector $s$‘s share of input $x$, $s_t^Y$ its value-added share, $\gamma_s$ returns to scale, $\rho$ the elasticity of substitution across sectors, and $\epsilon_t^x$ the elasticity defined in Appendix C:

$$\Delta TE = \sum_{s=1}^{S} \frac{1}{1-\gamma_s\rho}\left[s_t^Y - \rho \sum_x \epsilon_t^x s_t^x\right]\Delta TE_s. \tag{8}$$ $$\Delta AE_{\text{within}} = \sum_{s=1}^{S} \frac{1}{1-\gamma_s\rho}\left[s_t^Y - \rho \sum_x \epsilon_t^x s_t^x\right]\Delta AE_{\text{within},s}. \tag{9}$$

Equation (10), p. 1626, gives the within-sector change as a weighted sum of firms’ lagged input shares and proportional changes in capital and labor wedges:

$$\begin{aligned} \Delta AE_{\text{within},s} ={}& \frac{\alpha_s}{1-\gamma_s\theta_s}\sum_i\left[(1-\beta_s\theta_s)s^K_{i,t-1}+\beta_s\theta_s s^L_{i,t-1}-s^Y_{i,t-1}\right]\frac{\Delta\tau^K_{it}}{1+\tau^K_{i,t-1}} \\ &+ \frac{\beta_s}{1-\gamma_s\theta_s}\sum_i\left[\alpha_s\theta_s s^K_{i,t-1}+(1-\alpha_s\theta_s)s^L_{i,t-1}-s^Y_{i,t-1}\right]\frac{\Delta\tau^L_{it}}{1+\tau^L_{i,t-1}}. \end{aligned} \tag{10}$$

The between-sector component is equation (11), p. 1626:

$$\begin{aligned} \Delta AE_{\text{between}} ={}& \sum_{s=1}^{S}\frac{1}{1-\gamma_s\rho}\left[\epsilon_t^K(1-\beta_s\rho)s_t^K+\epsilon_t^L\alpha_s\rho s_t^L-\alpha_s s_t^Y\right]\frac{\Delta\mathcal{T}_{st}^K}{1+\mathcal{T}_{s,t-1}^K} \\ &+\sum_{s=1}^{S}\frac{1}{1-\gamma_s\rho}\left[\epsilon_t^K\beta_s\rho s_t^K+\epsilon_t^L(1-\alpha_s\rho)s_t^L-\beta_s s_t^Y\right]\frac{\Delta\mathcal{T}_{st}^L}{1+\mathcal{T}_{s,t-1}^L}. \end{aligned} \tag{11}$$

The paper focuses on within-sector allocative efficiency because the between-sector contribution is small (Figure 6, p. 1627). The estimates of firm-level TFP effects are imprecise and near zero, so the aggregation sets $\Delta TE$ to zero (Table 5 Panel B and text p. 1628). For the reallocation-only estimate, the authors add underreported firms to a randomly selected set of non-underreported firms with an equal credit decline, repeating the draw 10,000 times (Table 6, p. 1628-1629).

Empirical specifications

Firm-bank dynamic difference-in-differences. Equation (2), pp. 1613-1614, estimates quarterly firm-bank credit-growth responses across event-time windows $\tau=-2,\ldots,5$:

$$\begin{aligned} g^{\text{credit}}_{ibt} ={}& \sum_{\tau=-2}^{5}\beta^{\text{treat}}_{\tau}(\text{period}_{\tau}\times\text{exposed}_b) + \sum_{\tau=-2}^{5}\beta^{\text{period}}_{\tau}(\text{period}_{\tau}\times\text{underreported}_{ib}) \\ &+ \sum_{\tau=-2}^{5}\beta^{\text{treatgroup}}_{\tau}(\text{period}_{\tau}\times\text{underreported}_{ib}\times\text{exposed}_b) + \theta_{it} + \phi_b \\ &+ \beta^{\text{base}}_1(\text{underreported}_{ib}\times\text{exposed}_b) + \beta^{\text{base}}_2\text{underreported}_{ib} + \alpha_2 X_{ibt}+\epsilon_{ibt}. \end{aligned} \tag{2}$$

Here $g^{\text{credit}}{ibt}=\text{credit}{ibt}/\text{credit}_{ib,t-1}-1$. The sample is continuing firm-bank relationships over 2009:I-2014:IV, $N=1{,}981{,}219$. It includes firm-by-quarter and bank fixed effects and relationship controls (lending share, relationship length, main-lender indicator, and firm share in the bank portfolio); standard errors are two-way clustered by firm and bank. The separate overdue-loan subsample has $N=426{,}127$ (Figure 3, p. 1615).

Firm-level credit response. Equation (3), pp. 1618-1619, tests whether firms offset the firm-bank response by borrowing elsewhere:

$$\Delta\log\text{credit}_{it} = \sum_{t=-5}^{10}\Delta^{\text{treatgroup}}_t(\text{quarter}_t\times\text{treatment}_i\times\text{underreported}_i) + \sum_{t=-5}^{10}\Delta^{\text{treatment}}_t(\text{quarter}_t\times\text{treatment}_i) + \text{controls} + \alpha_1 X_{it} + \theta_i + \epsilon_{it}. \tag{3}$$

Treatment is the standardized pre-EBA borrowing share from exposed banks. The specification includes firm and industry-by-quarter effects, firm controls averaged over 2008-2010 interacted with quarter, and firm fixed effects; standard errors are clustered by firm. The quarterly 2009-2015 sample has $N=1{,}346{,}771$ firm-quarter observations. The estimated cumulative treatment effects are +16% for underreported firms and -14% for all others over the EBA and bailout period (text p. 1619; Figure 5, p. 1620).

Input effects by instrumental variables. Equation (6), pp. 1623-1624, estimates the effect of credit growth on input use and wedge growth:

$$\Delta\log y_{is} = \gamma\Delta\log\text{credit}_{is} + \text{controls} + u_{is}. \tag{6}$$

The endogenous credit change is instrumented with the firm-level pre-EBA borrowing share from exposed banks, interacted with underreporting status. The first stage is written out from Section IIIB (p. 1624), as the paper describes it:

$$\Delta\log\text{credit}_{is} = \Delta^{\text{treatment}}\text{borrowing share}_{is} + \Delta^{\text{treatgroup}}\text{borrowing share}_{is}\times\text{underreported}_{is} + \text{controls} + \epsilon_{is}.$$

The annual 2012 regressions use firm-size and two-digit-industry fixed effects and controls averaged over 2008-2010 (log assets, interest/EBITDA, capital/assets, current ratio, cash/assets, sales growth). Standard errors are clustered by industry; $N=104{,}499$ for labor and capital and $N=104{,}492$ for TFP. The reported labor and capital credit elasticities are 0.522 (SE 0.094) and 0.140 (SE 0.046), respectively; the TFP coefficient is 0.001 (SE 0.005), with first-stage $F=111.2$ (Table 5 Panel B, p. 1625). The translated mean input responses are reported in the Core results (text p. 1624).

Wedge prediction and productivity aggregation. Equation (12), p. 1628, maps the estimated credit response and initial capital wedge into predicted wedge changes, for $X=K,L$:

$$\frac{\Delta\hat{\tau}^{X}_{it}}{1+\tau^{X}_{i,t-1}} = \left(\hat{\gamma}^{X}_1+\hat{\gamma}^{X}_2\times\text{capital wedge}_{i,t-1}\right)\times\widehat{\Delta^{\text{treatment}}\text{ borrowing share}_{is}}. \tag{12}$$

These firm-level wedge changes are aggregated with equations (9)-(10). The reported full-intervention within-sector effect is -6.59%, or 53.83% of the actual 2012 decline; the credit-reallocation-only simulation averages -2.71%, or 22.10% of the decline (Table 6, p. 1628). Both exercises are partial equilibrium.

Datasets used

Dataset	Role in paper	Wiki page
Portuguese Credit Register (Central de Responsabilidades de Credito)	Monthly firm-bank loan balances by overdue bucket; universe of lending relationships above EUR 50 (2009-2015); primary source for underreporting algorithm and DiD	No page yet
Simplified Corporate Information / IES (Informacao Empresarial Simplificada)	Annual mandatory firm census; value added, employment, capital, sales, balance sheet; used for wedge measurement and productivity decomposition (2009-2015)	No page yet
Banco de Portugal quarterly bank balance sheet data	Bank-level capital ratios, sovereign bond holdings, liquidity; determines EBA exposure definition (2010-2012)	No page yet
EBA capital exercise disclosures	Bank eligibility and sovereign debt buffer used to define the exposed/nonexposed distinction (October 2011)	No page yet

Sample: quarterly loan data 2009:I to 2014:IV; annual firm data 2009-2015. Firms cover 81% of Portuguese sales and 73% of assets. Underreporting measured on firm-finance loans (36% of banks’ corporate portfolio; 73% collateralized). 56% of firms have multiple lending relationships, required for the within-firm firm-bank DiD.

When to read the full paper

Use the original if you are: (i) replicating the underreporting algorithm or DiD design (appendices A-C give the full algorithm, validity checks, and productivity decomposition details); (ii) studying how ratio-based capital requirements distort credit composition in distress (the paper tests risk-shifting vs. loss-delay mechanisms in Table 3); (iii) interested in the link between credit misallocation and aggregate TFP measurement in a bank-dependent economy; or (iv) extending the Hsieh and Klenow (2009) or Restuccia and Rogerson (2008) wedge framework to a credit-supply channel.

Attribution and rights

Source: peer-reviewed, American Economic Review 113(6), June 2023. Replication data available at ICPSR E120003V1. This distillation was first extracted on 2026-06-24 and expanded and AI-verified on 2026-10-04; it is not human-verified or independently reproduced.

Blattner, Laura, Luisa Farinha, and Francisca Rebelo. “When Losses Turn into Loans: The Cost of Weak Banks.” American Economic Review 113, no. 6 (June 2023): 1600-1641. DOI: 10.1257/aer.20190149. Copyright 2023 American Economic Association. Paywalled; this page is an extract-only distillation.