The Cheap-Goods Paradox: When Zombie Firms Export Disinflation
This series examines zombie firms in ten Asian emerging market economies and the relationships that keep them operating: weak or exposed banks, healthy competitors, global value chains, advanced-economy importers, and insolvency institutions. The opening question is simple but consequential: how can firms that are less productive and financially distressed contribute to lower prices beyond their own borders? Answering it gives readers a way to distinguish a beneficial productivity shock from a price decline created by excess capacity and delayed exit.
The evidence follows zombie firms from their measurement and persistence to their domestic effects and cross-border spillovers. It also asks why trade appears to carry the distortion internationally while foreign banks’ consolidated balance sheets remain relatively insulated. The aim is not to treat every distressed firm as a policy mistake. It is to understand when financial forbearance preserves viable activity, when it traps unproductive capacity, and what an effective response must observe.
What counts as a zombie firm?
“Zombie firm” is not a legal verdict. In the central study, it is an operational financial classification. A firm is counted as a zombie when its interest coverage ratio—the ratio of EBIT to interest expense—stays below 1 for at least three consecutive years. In practical terms, its operating earnings do not cover the interest due on its debt for a sustained period. The study smooths the classification: a one-year recovery does not immediately remove the label, while three consecutive years with an interest coverage ratio of at least 1 do.
That rule is useful because it avoids treating a single shock as a permanent condition. It is also limited. A low interest coverage ratio does not establish that a court would declare the firm insolvent, nor does it prove that every such firm has no economically valuable assets or possible restructuring path. The study reports similar results under a stricter definition that also requires below-industry-median sales growth, but the operational measure remains a measure of financial distress, not a universal definition of nonviability.
The study applies this definition to China, Chinese Taipei, Hong Kong SAR, India, Indonesia, Korea, Malaysia, the Philippines, Singapore, and Thailand. Its evidence combines firm financial statements and firm–bank links, bank balance-sheet information, macroeconomic data, merchandise trade, and bilateral international banking claims. That combination matters: the subject is not just a list of troubled companies. It is a system in which a firm’s financial condition can influence who receives credit, how much capacity remains in a sector, and how goods move through production networks.
The scale of the linked evidence also explains why the study can ask questions that a country-level insolvency count cannot. The firm data contain about 717,000 firms and 3.4 million firm-year observations with a computable interest coverage ratio over 2008–2020. The bank match covers 326 banks represented by 622 bank–country units. The macro and cross-border analyses use their own observation windows and units, so a percentage in a firm regression should not be read as a national forecast. What the design makes possible is a chain of comparisons: distressed firms against healthier firms, banks with different asset quality, sectors with different zombie shares, and advanced economies with different trade exposure.
The map is deliberately wider than a bank-risk diagram. Banks can help a distressed firm survive, but the consequences may appear in the performance of healthy firms, in domestic price formation, and in the prices paid by importers. Insolvency systems sit at a different interface: they can coordinate restructuring or exit, but only if their legal rules and institutions can distinguish viable businesses from firms whose capacity is merely being prolonged.
A stock that refuses to shrink
The scale and persistence of the measured problem make the distinction important. Across the ten economies, the share of firms classified as zombies rose from about 11% in 2008 to 17% in 2021. Measured by assets, the share rose from 6% to 13%, with a spike near 16% in 2015–2016. Measured by debt, it roughly doubled from 9% to 20%. The debt measure is especially revealing for a problem tied to credit: the firms may be fewer than the headline count suggests, but they can command a substantial share of the liabilities and productive resources in a market.
Exit is much less common than continued status. From 2011 onward, roughly 93–95% of zombie firms remained zombies in the following year, while only about 3% of healthy firms entered the category in a given year. These are descriptive transition estimates for the study’s harmonized sample, not proof of a single cause. They do, however, change the question. A rising zombie share cannot be understood only as a wave of new distress. It also reflects a low rate of recovery or exit once a firm enters the state.
That persistence is the first clue that firm weakness alone is not enough to explain the phenomenon. A company can use cash, sell assets, or draw on other liabilities for a time. It cannot normally sustain a multiyear inability to cover interest without some form of accommodation. The next question is therefore institutional: who continues to provide the room, and what do they gain by doing so?
The different denominators also matter. A 13% asset share and a 20% debt share in 2021 do not describe the same risk as a 17% count share. They suggest that a portion of the firms classified as zombies is relatively large or highly leveraged, so their persistence can influence prices and credit allocation beyond the number of legal entities involved. Nor does the aggregate trend imply identical conditions across the ten economies. The figures establish a regional pattern; the mechanisms still have to be tested against firm, bank, sector, and country variation.
Why lower prices can signal weaker production
The central paradox is that zombie firms can put downward pressure on prices without being efficient producers. In the firm-level evidence, zombie status is associated with a 2.77 percentage-point lower return on assets, about 5 percentage points lower sales growth after controls, roughly a 1 percentage-point lower capital-expenditure ratio, and about 3 percentage points lower employment growth. A transition into zombie status is also associated with lower sales per employee and lower value added per employee.
At the same time, these firms can continue selling output. Their lower margins and markups are consistent with a volume-over-margin strategy: keep capacity running, generate enough cash flow to service part of the debt, and rely on a favorable borrowing cost to avoid an immediate reckoning. In the sample, zombies pay an average interest rate of about 5.4%, compared with 6.0% for non-zombies. The paper does not call this a fiscal subsidy. It describes the gap as the equilibrium price of bank forbearance when lenders prefer to avoid recognizing losses.
This distinction prevents a common mistake. If a more productive exporter becomes cheaper, an importing economy can receive a real-income gain. If a less productive firm is kept alive and sells from excess capacity at depressed prices, the immediate price effect can be accompanied by weaker investment and displaced production. The price decline is then a symptom of misallocation rather than proof of improved productive capability.
There is an important reason to compare conditional and unconditional performance. In the raw summary statistics, zombies have slightly higher average sales growth than non-zombies—about 0.39 versus 0.35. Once firm characteristics and fixed effects are taken into account, the comparison reverses: zombie status is associated with lower sales growth and lower asset turnover. The difference is a warning against reading headline volume as proof of health. A heavily leveraged firm can keep sales moving while earning too little on each unit of capacity. The social stakes are visible as well. Zombie firms accounted for roughly 6–9% of reported employment in the ten economies in 2015–2020, or about 2.5–3.2 million workers in 2016–2017, even though their asset share was larger. Exit and restructuring therefore affect workers and communities, not just bank ledgers.
The relationships that make the problem travel
The domestic and international questions are connected by several relationships that must not be collapsed into one vague idea of “financial fragility.” First, banks can sustain zombie firms through loan rollovers or temporary restructuring. Second, the resulting firms compete with healthy firms for credit, workers, and market share. Third, sectors with excess capacity can sell goods—especially goods used as inputs by other producers—at prices that affect downstream production. Fourth, advanced economies may import the price effect through global value chains even when their own banks do not suffer a visible balance-sheet shock.
The evidence therefore has to answer more than whether zombie firms exist. It must establish whether their persistence is linked to bank behavior; whether the firms and their healthy competitors perform worse; whether domestic inflation and growth respond; and whether the international pattern follows trade exposure, bank exposure, or both. It must also distinguish zombie-related excess supply from a broad Asian slowdown, genuine productivity gains, or weaker demand.
That last comparison is especially important for policy. A bank supervisor looking only at foreign-bank nonperforming loans could conclude that an Asian zombie problem has not crossed a border. A central bank looking only at lower import prices could conclude that foreign efficiency has improved. Both readings can miss the possibility that goods-market competition is lowering prices while displacing investment. The relevant exposure may be distributed across many firms and supply chains rather than concentrated in one institution’s balance sheet.
The final relationship is institutional rather than purely financial. Insolvency regimes determine how distress can be recognized, negotiated, reorganized, or closed. The World Bank’s review treats a sound regime as an economic infrastructure that can support entrepreneurship and credit, preserve viable firms and jobs, manage nonperforming loans, and move assets out of nonviable firms. That does not mean liquidation is always preferable. It means indefinite survival and orderly restructuring are not the same thing. It also means legal rules cannot be assessed apart from courts, practitioners, enforcement, and the data needed to see whether firms actually exit or recover.
The puzzle is now specific enough to investigate. A persistent low-coverage firm is not automatically a legally insolvent one, and a low export price is not automatically a productivity gain. The next step is to examine the bank incentive and creditor relationships that can turn financial distress into a durable operating state. That mechanism determines whether the cheap goods are the result of efficiency—or the outward expression of a problem that has not been allowed to end.