Zombie Banks and Companies
Table of Contents
Zombie banks and companies raise a question that ordinary debt ratios cannot answer: is continued lending giving a viable business time to recover, or postponing recognition of a loss? A company can report weak interest coverage without being beyond repair. A bank can have inadequate capital without being insolvent. And a loan extension can improve repayment prospects rather than conceal failure.
To assess a zombie-lending claim, separate the research definition, the borrower’s business prospects and the lender’s incentives. This guide explains those distinctions using historical Japan and U.S. evidence, followed by hypothetical examples. It does not estimate how many zombies exist today or diagnose a named institution.
What Do Zombie Banks and Companies Mean?
In economic discussion, a zombie company usually means a business that cannot sustain itself on a viable basis but continues operating with repeated financing support. Researchers need observable rules to study that idea, so they use measures of weak earnings, debt burdens or financing subsidies. Those measures identify candidates; they cannot directly observe every firm’s future.
A firm passing a zombie screen, a borrower lacking a credible path to repayment and a bank unable to absorb its losses are different findings. Evidence for one does not establish the other two.
| Question | Evidence to examine | What the label cannot establish |
|---|---|---|
| Does a company pass a research screen? | The study’s exact ratios, history, sample and thresholds. | A legal insolvency finding or certain future failure. |
| Is a borrower economically viable? | Credible operating cash flows, financing needs and restructuring alternatives. | That every loss-making business should close. |
| Is a bank financially impaired? | Asset recoverability, loss-absorbing capital, funding and available support. | Its condition from a nonfinancial company’s interest-coverage test. |
Zombie bank is an informal description often used for an institution whose continued operation depends on support despite serious financial impairment. Specify whether the concern is capital weakness, losses exceeding its economic equity or a funding shortage. These require different evidence. Rosengren’s historical analysis explicitly distinguishes troubled but solvent banks from clearly insolvent ones. [5] The broader run and contagion mechanisms belong to financial crises explained.
How Researchers Identify Zombie Companies
There is no universal empirical definition. Two frequently cited studies illustrate why the qualifying conditions matter. All conditions within a row below apply together.
| Study and sample | Screening conditions | Important boundary |
|---|---|---|
| BIS, September 2018: listed nonfinancial firms in 14 advanced economies | Broad: interest coverage below 1 for at least three consecutive years; firm at least ten years old. | Narrow version additionally requires Tobin’s q below the sector-year median. |
| Federal Reserve staff, July 2021: U.S. public firms and private-firm borrowers in supervisory data | Leverage above the annual sample median; interest coverage below 1; negative real sales growth over the preceding three years. | No age filter or three-consecutive-year interest-coverage condition is imposed. |
The BIS adds a market-based prospects measure: Tobin’s q compares assets’ market value with replacement cost. The Fed authors instead use real sales growth and warn that missing age information can cause young firms to be misclassified. These are historical research designs, not interchangeable population counts. [1] [2]
When someone cites a zombie percentage, ask which definition, observation period and denominator produced it. A share of listed firms is not a share of all businesses, employment or credit. The Fed’s private-firm data exclude small banks and very small businesses. [2]
What Interest Coverage Tells You
The interest coverage ratio, or ICR, compares earnings before interest and taxes with interest expense for the same period. For a positive interest expense:
An ICR below 1 means the specified earnings measure is smaller than interest expense. It does not mean the company has already missed a payment. Cash reserves, asset sales or new equity could fund a shortfall. Conversely, coverage above 1 does not show that principal repayments, taxes, investment needs and other cash obligations are manageable. EBIT is an accounting earnings measure, not cash available for every payment. The Fed note uses this EBIT-based definition. [2]
Worked Example: One Company, Different Classifications
Consider fictional Lakeview Components, which is twelve years old at the end of Year 3. The following hypothetical annual amounts are in USD millions. Interest expense stays at $10 million while EBIT changes.
| Year | EBIT (USD millions) | Interest (USD millions) | ICR calculation |
|---|---|---|---|
| Year 1 | 6 | 10 | 6 / 10 = 0.60 |
| Year 2 | 8 | 10 | 8 / 10 = 0.80 |
| Year 3 | 9 | 10 | 9 / 10 = 0.90 |
At the end of Year 3, also assume leverage is 60% against an annual sample median of 45%. Its real sales index falls from 100 at Year 0 to 90 at Year 3: (90 / 100 − 1) × 100 = −10% over the preceding three years. Its Tobin’s q is 1.2, above the sector-year median of 1.1.
- BIS broad conditions: satisfied, given its age and three consecutive ICR readings below 1.
- BIS narrow conditions: not satisfied, because q is above the relevant median.
- Fed conditions: satisfied, given high relative leverage, current ICR below 1 and the decline in real sales.
This applies the studies’ stated screening conditions to invented inputs; it does not replicate their datasets. The classification changes even though the company does not. A relatively high q does not prove recovery, and improving EBIT does not establish that Lakeview can refinance its debt when due.
Viable Refinancing vs. Destructive Evergreening
Evergreening, in the problematic sense used here, means continuing or replacing troubled credit to keep a nonviable borrower afloat and postpone confronting the lender’s loss. The extension can preserve the appearance of payment performance without improving the underlying repayment source. A lender’s desire to avoid recognizing a loss is distinct from a credible recovery plan.
Financing a Credible Recovery
- A temporary disruption explains the cash shortfall.
- Updated forecasts support repayment on realistic terms.
- New capital, operational changes or debt restructuring address the problem.
- The lender assesses and reports the remaining risk.
Postponing an Unresolved Loss
- Repeated extensions replace a credible repayment source.
- Repayment depends on further advances or implausible assumptions.
- Little changes in the business or debt burden.
- Avoiding a write-down drives the decision.
These are analytical contrasts, not a rule that a particular number of extensions proves misconduct. For example, a supplier awaiting payment on a sound contract may need bridge financing; that differs from a business whose continuing operations consume cash with no credible improvement. Both might ask to extend a maturity.
The distinction also appears in official practice. The U.S. agencies’ 2023 commercial real estate workout statement supports prudent accommodations based on a comprehensive borrower review, while retaining appropriate risk classification and loss recognition. It does not treat every modified loan as healthy, or every extension as improper. [6] For the wider monitoring toolkit, see bank credit risk management.
Why a Weak Bank Might Keep Lending
A bank nearing a capital constraint may prefer a rollover to a restructuring that makes an existing loss unmistakable. That incentive can persist even when recognizing the loss and repairing the balance sheet would be better for future lending. Historical Japan research examines this problem. [3] [4]
Consider a separate hypothetical bank. It reports $100 million of assets, including a $10 million loan, and owes $94 million in liabilities. Assume the loan’s economically recoverable present value falls to $6 million, all other assets and liabilities stay unchanged, and the illustrative book amounts have no allowance yet reflecting that shortfall. Compare book values with the assumed economic values at that same point in time.
| Item (USD millions) | Illustrative book value | Assumed economic value |
|---|---|---|
| Problem loan | 10 | 6 |
| Other assets | 90 | 90 |
| Total assets | 100 | 96 |
| Liabilities | 94 | 94 |
| Equity: assets less liabilities | 100 − 94 = 6 | 96 − 94 = 2 |
The loan shortfall is $10 million − $6 million = $4 million. Economic equity is therefore only $2 million. Rolling over the loan leaves that shortfall intact unless it changes the expected recovery. Recognizing the loss exposes damage that has already occurred; concealing it cannot restore the missing value.
This bank still has positive economic equity under the assumptions. The example establishes neither insolvency nor compliance with any regulatory capital minimum. It simplifies away taxes, other allowances and subsequent income. Actual reporting must follow applicable accounting and supervisory requirements; an extension does not permit required loss recognition to be delayed.
Likewise, interest paid entirely from a new advance is different from interest generated by the borrower’s business. Following the source of repayment helps distinguish cash arriving at the lender from an improvement in the borrower’s capacity.
What the Japan Evidence Shows
Japan’s experience gives the mechanism historical substance. Peek and Rosengren’s 2005 study reports that additional credit to impaired firms was associated with banks’ loss-avoidance incentives, especially when reported capital approached required minimums. That is the published abstract’s finding, not a diagnosis of every Japanese bank. [4]
Caballero, Hoshi and Kashyap’s 2008 paper studies Tokyo-listed firms over 1981–2002. Its financing-subsidy measure compares actual interest payments with a constructed favorable minimum borrowing-cost benchmark. This is different from either ratio screen above. Their analysis links zombie prevalence to weaker investment and employment growth among healthier competitors, although some employment results weaken under additional controls. [3]
In a March 2, 2009 speech, Boston Fed president Eric Rosengren described the allocation problem: troubled banks could preserve support for problem borrowers while reducing loans to stronger borrowers that could be cut without realizing losses. This historical interpretation concerns which credit survives, not only how much aggregate lending contracts. [5] The wider expansion and contraction framework is covered in credit cycles and business cycles.
How Zombie Lending Can Affect Credit and Productivity
The proposed congestion effect is that sustained support keeps resources tied to weak uses. Competitors may face pressure on selling prices or compete for labor, premises and financing that could support more productive activity. Preserving existing jobs can have immediate benefits, while blocking investment and restructuring can impose later costs. Both sides belong in the assessment.
Evidence depends on the definition. The 2018 BIS study finds congestion effects for its narrow measure, but not its broad screen. A count of mature firms with weak interest coverage cannot simply inherit the narrow measure’s conclusions. [1]
Nor does the Japan subsidy pattern describe every setting. The 2021 Fed note finds that, in its U.S. bank-loan sample for 2015–19, screened zombies faced higher average spreads, more collateral and shorter maturities than other firms. That challenges a universal assumption that zombie borrowers always receive favorable terms. [2]
A relationship between weak firms and weak growth also needs a causal argument. A sector downturn can simultaneously damage borrowers, lenders and investment opportunities. Researchers use controls and comparisons to address that problem, but an alarming company count alone does not identify the growth lost because financing continued.
Do Low Interest Rates Create Zombies?
There are effects in different directions. Consider another hypothetical sensitivity calculation: hold annual EBIT at $9 million and assume actual annual interest expense falls from $10 million to $7.5 million. ICR rises from 9 / 10 = 0.90 to 9 / 7.5 = 1.20. Lower financing costs move the company above the coverage threshold without changing its operations.
That is not automatic pass-through from a central bank rate cut: contracts, refinancing dates and credit spreads determine actual expense. At the same time, cheap funding can reduce pressure on lenders to confront bad exposures. BIS discusses both the mechanical coverage improvement and this incentive channel, including reverse causality and common economic factors. Its analysis does not settle the overall policy tradeoff. [1]
It therefore does not follow that rate cuts always create zombies or that rate increases cure them. Tightening can also damage viable borrowers. Separate monetary support from the work of assessing losses and restructuring debt; see monetary policy limitations for the broader transmission issues.
Match the Response to the Problem
An analytical response distinguishes several tasks: identify realistic asset recoveries; recognize losses; restore the lending capacity of viable banks; restructure businesses with credible prospects; and resolve nonviable operations with attention to disruption and transition costs. A liquidity loan alone cannot fill an economic capital loss. A borrower workout alone cannot repair every lender.
Rosengren’s speech emphasizes confronting troubled assets, while the U.S. workout statement demonstrates why constructive restructuring and accurate risk recognition can coexist. [5] [6] Actual powers and procedures vary by jurisdiction; the broader toolkit is introduced in financial regulation and deposit insurance.
How to Evaluate a Zombie-Lending Claim
- Identify the claim. Is it about a statistical screen, borrower viability, a bank’s solvency or the purpose of a loan modification?
- Reproduce the stated definition. Check the earnings measure, sample median, age requirement and history. Keep missing information visible.
- Examine the repayment path. Separate operating cash generation from asset disposals, new equity and repeated borrowing. Review debt maturities and plausible downside assumptions.
- Look for genuine changes. Does the workout reduce debt, supply loss-absorbing capital or improve operations? Merely moving a payment date answers little.
- Assess lender incentives and evidence. Compare loan terms, recoverability assumptions and disclosed losses. A rollover is observable; an intention to conceal losses needs additional support.
- Size the exposure correctly. Distinguish firm counts from credit amounts, employment and bank capital at risk. State the observation period and population covered.
Public statements may reveal weak coverage and refinancing needs while leaving private loan terms, supervisory judgments or realistic restructuring alternatives unknown. In that case, describe the observed weakness and the missing evidence rather than turning a screen into a conclusive accusation.
Common Mistakes
- Treating ICR below 1 as legal insolvency. The ratio addresses one earnings-to-interest comparison, not every legal test or payment source.
- Combining definitions. Adding the BIS age and persistence rules to the Fed screen creates a different measure.
- Calling every refinancing evergreening. Test whether the plan improves realistic repayment prospects.
- Confusing a recognized loss with a newly created loss. A write-down can reveal deterioration that already occurred.
- Using an old study as today’s count. Dates, samples and surviving firms change.
- Assuming all weak firms receive subsidized credit. Examine the actual pricing and risk evidence in the relevant setting.
Limitations of Zombie-Firm Analysis
Historical ratios cannot establish a firm’s future viability, a bank’s legal status or the motive behind a refinancing decision. Those conclusions need additional evidence and a clearly stated framework.
Accounting measures can be affected by unusual expenses and changing reporting choices. Sales history can miss a credible turnaround; market valuations can be optimistic or pessimistic. Restricting a sample to listed companies or large-bank borrowers leaves other businesses outside the evidence.
There are also real transition costs. Ending support can release resources but destroy jobs or disrupt suppliers before alternative activity develops. The Japan paper does not establish an optimal regulation or settle the net welfare balance, and the BIS leaves the monetary-stimulus/productivity tradeoff unresolved. [3] [1] Avoid a universal policy prescription from either study.
Frequently Asked Questions
Yes. A temporary earnings shortfall can coexist with adequate cash, new equity or a credible recovery plan. Coverage below 1 warrants investigation, especially when persistent, but does not by itself establish nonviability or legal insolvency. Check future cash needs and how any restructuring changes them.
No. Insufficient capital relative to a requirement or risk exposure is different from losses exhausting economic equity. Funding problems are another distinction. Explain the specific weakness and available support rather than relying on an informal label; a corporate ICR test does not answer the bank question.
No. An extension can bridge a temporary mismatch or form part of a prudent workout. The relevant questions are whether repayment assumptions are credible, what changes in the business or debt structure, and whether the lender properly assesses the remaining loss. Contract terms alone do not establish intent.
They can use different profitability, age, leverage, sales-growth or financing-subsidy conditions, as well as different years and samples. Compare those choices before comparing percentages. A historical share of listed firms cannot be presented as a current share of all companies.
Sources
- Banerjee and Hofmann, BIS, The rise of zombie firms: causes and consequences, September 2018. Definitions, causes, consequences, conclusions and footnotes.
- Favara, Minoiu and Perez-Orive, Federal Reserve Board, U.S. Zombie Firms: How Many and How Consequential?, July 30, 2021. Identification, Table 2 and footnotes 3–6.
- Caballero, Hoshi and Kashyap, Zombie Lending and Depressed Restructuring in Japan, American Economic Review 98(5), December 2008, pp. 1943–1977. Author-hosted published paper; identification, sample, robustness and concluding policy limits.
- Peek and Rosengren, Unnatural Selection: Perverse Incentives and the Misallocation of Credit in Japan, American Economic Review 95(4), September 2005. Published abstract.
- Eric Rosengren, Federal Reserve Bank of Boston, Lessons from Japan, March 2, 2009. Historical speech; bank-capital incentives and troubled versus insolvent banks.
- U.S. federal banking and credit-union agencies, Policy Statement on Prudent Commercial Real Estate Loan Accommodations and Workouts, issued June 29, 2023. Opening principles and sections II–IV.
Disclaimer
This article is for educational and informational purposes only and does not constitute investment, legal or accounting advice. Research definitions and historical findings are specific to their samples and periods. All worked numerical examples are hypothetical. Actual viability, reporting, insolvency and restructuring decisions require current evidence and qualified professional judgment.
