Overview

This chapter compares foreign-bank and trade channels for zombie-firm risk. In an 18-advanced-economy panel, bank-weighted exposure has negative but statistically insignificant CPI and GDP coefficients, whereas GVC exposure in the broader panel is associated with lower inflation and growth. Foreign banks can remain connected without comparable consolidated impairment, while trade distributes effects across prices, production, and investment. The chapter then treats insolvency as a coordination mechanism for viable restructuring or orderly exit and argues for surveillance combining firm, bank, court, trade, price, and investment data.

Learning outcomes

In the 18-economy bank panel, foreign-bank exposure has negative but statistically insignificant CPI and GDP-growth coefficients.
Trade exposure is decentralized across firms and products, so stable foreign-bank balance sheets do not rule out real-economy spillovers.
Insolvency systems can interrupt persistence only when they make viable restructuring and nonviable exit workable in practice.

Keywords

Zombie firmsForeign banksTrade exposureSurveillanceInsolvencyRestructuring

The Missing Transmission Route: Why Trade Matters More Than Foreign Banks

The evidence now permits a comparison between two ways zombie-firm risk might cross a border: through internationally active banks or through the goods and production networks those firms serve. This chapter makes that comparison on four dimensions—exposure, balance-sheet effects, macroeconomic outcomes, and controllability—then applies it to insolvency and surveillance. The practical conclusion is not that banks are irrelevant. It is that a sound foreign-bank balance sheet can coexist with a real-economy shock that no single lender can manage away.

The two channels behave differently

The bank-lending channel is measured with a spillover index based on predetermined bilateral foreign claims of banks headquartered in advanced economies on the non-bank private sector in the ten Asian emerging economies. In a panel of 18 advanced economies, a one-standard-deviation increase in this bank-weighted exposure is associated with a −0.027 percentage-point CPI coefficient and a −0.232 percentage-point GDP-growth coefficient. Neither is statistically significant; the p-values are 0.40 and 0.17.

The trade-weighted channel has a different pattern. In the broader panel, a one-standard-deviation increase in GVC exposure is associated with about 0.135 percentage points lower CPI inflation and 0.405 percentage points lower GDP growth. The comparison is not just a contest between a significant and an insignificant coefficient. The channels also differ in where risk sits. Bank exposure is concentrated in identifiable lending groups with capital, asset quality, and risk-management systems. Trade exposure is distributed across many firms, contracts, products, and downstream investment decisions.

The bank result does not mean that foreign banks withdraw from Asian zombie firms. The evidence indicates that foreign banks maintain some relationships, with higher persistence in some comparisons. Their Asian exposure is small relative to their consolidated portfolios, and the study’s interpretation is that they monitor or selectively restructure the exposure without allowing it to impair group-level bank health. A bank can therefore remain connected to a zombie without transmitting the full macroeconomic effect to its home economy.

The contrast is sharper when bank headquarters and lending location are separated. A one-percentage-point increase in a bank’s zombie-loan share raises NPLs by 0.019 percentage points and lowers ROA by 0.002 percentage points in the all-bank results. The effect is driven by domestic Asian lending: the corresponding coefficients for Asian domestic lending are 0.065 for NPLs and −0.006 for ROA. Loans made as a foreign bank within the region do not show the same impairment, and non-Asian foreign banks show no significant NPL or ROA response whether their headquarters are in advanced economies or other EMEs.

This produces an important asymmetry. The banks closest to the firms and the local supervisory environment bear the measurable balance-sheet cost, while foreign banks with smaller consolidated weights can preserve the relationship without a comparable deterioration.

Published by DeepMox

FAQ

Does a null foreign-bank coefficient prove no lender can fail?

No. The bank data lack direct failure or exit flags, so the null result indicates no significant macro spillover in the studied sample, not universal lender safety.

What indicators should complement foreign-bank surveillance?

It should add firm viability in GVC sectors, zombie lending, trade composition, origin-specific import prices, manufacturing prices, and importer investment.

What determines whether insolvency reforms work?

Courts, insolvency practitioners, information systems, enforcement, and creditor coordination determine whether formal rights produce usable restructuring or orderly exit.

Previous8 minZombie Firms: Slowdown, Productivity, Demand, or China?

Next13 minZombie Firms: Policy for Viable Recovery and Orderly Exit

Zombie Firms: Why Trade Matters More Than Foreign Banks

The Missing Transmission Route: Why Trade Matters More Than Foreign Banks

The two channels behave differently

FAQ