不良贷款激增时稳定信贷：资产管理公司的作用

Stabilizing credit when nonperformingloans surge: the role of assetmanagement companies Reiner Martin, Edward O’Brien,Udara Peiris, Dimitrios P. Tsomocos Abstract When default losses elevate borrowing costs, expanding credit cannot stabilize the econ-omy because default rates feed back to lending rates through bank balance sheets.Assetmanagement companies (AMCs) break this loop by purchasing nonperforming loans at theirlong-run recovery values, thereby fixing the effective default rate that banks face.Gov-ernment purchases of performing loans expand credit but leave this feedback intact.In amodel calibrated to the eurozone, the AMC reduces quarterly default rates by 0.8 percentagepoints, lowers lending rates by 1.6 percentage points, and raises welfare by 0.2%. Governmentpurchases crowd out bank deposits, contracting credit; default rates rise by 1.8 percentagepoints, lending rates increase by 1.2 percentage points, and welfare falls by 0.3%. Keywords:Nonperforming Loans, Asset Management Companies, Credit Stabilization,Bank Balance Sheets, Endogenous Default JEL Classification:E44, G01, G21, G28 Non-Technical Summary Between 2014 and 2018, nearly 10% of loans held by European banks were nonperforming. Su-pervisory evidence indicates that these elevated ratios constrained bank profitability and lendinggrowth, even after central bank operations had restored funding. The underlying problem wasnot liquidity but a feedback loop: high default rates forced banks to increase the cost of lending,higher rates increased debt burdens on borrowers, and heavier burdens led to more defaults.Several European governments responded by establishing Asset Management Companies.AnAMC purchases nonperforming loans from banks at prices reflecting long-run recovery values,removes these assets from bank balance sheets, and holds them until markets recover. Ireland’sNational Asset Management Agency accumulated assets equal to approximately 44% of GDP.Spain and Slovenia created AMCs managing 10% to 16% of GDP. South Korea’s AMC eventuallyrecovered about 60% of acquired assets. Despite their widespread use, little quantitative researchhas examined how AMCs affect the broader economy. AMCs perform many functions in practice, from managing workouts, to preventing fire sales,to catalyzing private markets for distressed debt. This paper focuses on one channel. In practice,AMCs purchase nonperforming loans above distressed market prices, compensating banks forlosses they would otherwise absorb.The model captures this by assuming the AMC fixes theeffective default rate that banks face when pricing new loans.Banks expecting stable returnscan offer lower interest rates because the AMC absorbs losses exceeding normal levels.Lowerrates reduce debt servicing burdens, fewer borrowers default, and the vicious cycle reverses.Government purchases of performing loans operate differently. For high-grade instruments suchas commercial paper and agency debt, such purchases work well because intermediation costsare modest and default risk is negligible.Risky corporate loans present a different problem.Financing these purchases diverts household resources from bank deposits, forcing banks tocontract lending by more than the government expands credit. Because banks continue to facefull default risk on remaining portfolios, the feedback between default and lending rates persists. A static model characterizes these mechanisms analytically. Firms borrow to finance workingcapital and can default at a cost depending on aggregate credit conditions. Banks price loans tobreak even given expected defaults. Two inefficiencies arise: individual firms do not internalizehow their default affects borrowing costs for all firms, and deteriorating conditions reduce themarginal cost of default, inducing more of it. Output sensitivity to shocks rises with the defaultrate, so economies with high default are less efficient and more volatile.An AMC that fixesthe effective default rate eliminates this amplification. Alternative policies cannot replicate thisoutcome: direct government lending displaces bank lending without altering default pricing, andequity transfers stabilize profits but not lending rates because default still varies with borrowing. The dynamic model embeds these mechanisms in a real business cycle framework calibratedto eurozone data. Banks face an agency friction where bankers can divert assets, so depositors impose an incentive constraint tying lending capacity to net worth.This creates a financialaccelerator: adverse shocks erode capital, tighten credit, and amplify disturbances. Firms financelabor costs through working capital loans and choose default optimally. Credit conditions dependon both firm equity and bank balance sheet stress, so each sector’s troubles feed back to the other. Monte Carlo simulations compare three regimes: no policy, an active AMC, and governmentpurchases of performing loans. When the AMC absorbs half of excess default l