Editors' note: This is the second in a two-column series. The first column introduced a new methodology to classify zombie firms which suggests they may be less prevalent than previously thought.
Zombie firms are poorly performing firms that are unable to service their debt from cash flows over a prolonged period. In a perfect market, these firms would exit, yet they often linger due to market inefficiencies, sometimes sustained by bank lending. Such zombie lending can have a large negative economic impact. It contributed, for instance, to the ‘lost decade’ of Japan’s economy (Caballero et al. 2008). Zombie firms have become more prevalent in many European countries over the last decades (Banerjee and Hofmann 2022), and this rise in zombie firms has been put forward as an explanation for Europe’s low productivity (McGowan et al. 2018), low innovation and patenting activity (Schmidt et al. 2023), and lack of inflation (Acharya et al. 2024) during the recovery from the financial and sovereign debt crisis.
In this column, based on De Jonghe et al. (2024a), we investigate a novel, information-based mechanism that might induce banks to reduce credit supply to zombie firms, namely bank specialisation.
The mechanism has two ingredients. First, zombie firms can create negative congestion externalities on healthy borrowers. Such externalities arise when zombie firms are not forced to downsize, thereby taking too much market share in the input and output market (Caballero et al. 2008), and hence put pressure on healthy borrowers by keeping wages high and mark-ups low (Acharya et al. 2024). If so, keeping zombie firms alive will deteriorate the bank’s lending portfolio, not only directly through having zombie firms in the portfolio, but also indirectly because the performance of the bank’s healthy borrowers might deteriorate if they must compete with zombie firms.
Second, to understand whether congestion externalities occur, banks need to have industry-specific knowledge. This type of knowledge cannot be inferred through firm-specific hard or soft information but can be learned by having many interactions with borrowers from the same industry, i.e. by specialising (Berger et al. 2017, De Jonghe et al. 2020). Banks thus have an incentive to internalise congestion externalities created by zombie firms but are only likely to do so if they are aware of these externalities to healthy borrowers and can identify the zombie firms.
Data
To investigate this mechanism, we combine three data sources of the National Bank of Belgium (Belgium’s central bank). We draw firm-bank level credit information for the universe of banks and firms in Belgium from the Corporate Credit Register and augment these data with firms’ and banks’ balance sheets and income statements. The dataset runs from 2004 to 2018 and covers 614,320 firm-bank-year observations for 54 banks and more than 50,000 firms. The credit register allows us to capture both the extensive margin of lending for existing borrowers and the intensive margin of lending.
As banks are most informed about industries where they grant most loans, banks’ industry specialisation is often proxied by an industry’s share in the bank’s lending portfolio (De Jonghe et al. 2020, Blickle et al. 2024). We follow this approach and measure the specialisation of a bank in each industry as the ratio of the total lending volume of a bank to an industry to the total lending volume of the bank to all industries. Specialisation captures how important a given industry is for a bank. Note that this is very different from market share, which captures how important a bank is for a given industry.
Many studies show various approaches to identifying zombie firms. We identify zombie firms using a modification of the often-used OECD zombie definition. We explain the exact conditions and motivation for this in detail in an accompanying VoxEU column “Counting the undead: a new metric for identifying zombie firms” (De Jonghe et al. 2024b); but in short, thanks to the adjustments, we avoid the misclassification of firms with high levels of depreciation (e.g. from high investments) or high recurring financial revenues (e.g. from treasury bills or intracompany loans).
Baseline results
Our baseline results at the extensive margin show that, holding credit demand constant, zombie firms are less likely to receive a new loan from a bank that is more specialised in lending to the zombie’s industry than from a bank that is less specialised. Results at the intensive margin confirm this. For a given level of demand, a zombie firm sees its credit supply significantly more reduced from the bank that is relatively more specialised in lending to the zombie’s industry.
Figure 1 Effect of a one standard deviation increase in bank specialisation on credit growth for healthy firms (blue bar) and zombie firms (red bar)
Our results can also be interpreted as follows:
banks reduce their credit supply significantly more to zombie firms in industries in which they are specialised than to zombie firms in industries in which they are not specialised. Both interpretations, however, are in line with the idea that banks are more aware and concerned about congestion externalities of zombie lending in industries where they are specialised.
Characteristics of the specialisation channel
If specialisation provides banks with superior information about zombie congestion externalities, banks’ incentives to reduce credit to zombie firms should be a function of how severe these externalities are for healthy borrowers.
We perform three sets of analyses, each providing important support for the channel:
- Healthy borrowers compete with zombie firms for resources and market share. This competition is likely only distorted if zombie firms make up a meaningful fraction of the industry. In line with this, we find that banks are significantly more restrictive in their lending to zombie firms in specialised industries where the share of labour employed at zombie firms is higher.
- Zombie congestion might force healthy firms to forego some investment opportunities. This is arguably more costly in (fast) growing industries, where firms are more likely to have good investment opportunities in the first place, and these opportunities tend to yield higher returns. Therefore, the cost of congestion externalities increases with the industry’s growth opportunities. Well-informed banks have incentives to prevent this. In line with this, we find that banks are significantly more restrictive in their lending to zombie firms in specialised industries with better growth prospects.
Figure 2 Effect of a one standard deviation increase in bank specialisation for healthy firms (blue bar) and zombie firms (red bar) in industries with low growth opportunities (left) and high (growth opportunities)
- A zombie firm’s default may pose contagion risk for the collateral value of healthy borrowers’ assets if the zombie’s assets are sold at fire-sale prices. This is more likely to occur in industries with higher asset specificity, i.e. where assets are more difficult to redeploy elsewhere without significant value loss. Thanks to the information advantage, a bank is more likely to be aware of the consequences of collateral liquidation on its other borrowers in industries where it specialises. In line with this, we find that the effect of bank specialisation on zombie lending is weaker in industries with higher asset specificity.
Conclusion and policy implications
Banks that develop in-depth, industry-specific knowledge through specialising in a given industry tend to be more adept at curtailing credit to zombie firms, and especially so when zombie firms are expected to have stronger negative effects on healthy borrowers.
These results imply that bank specialisation can benefit financial stability (Beck et al. 2017) and support competitiveness (Draghi report, 2024). Bank specialisation not only improves the quality of the bank’s lending portfolio by reducing the relative share of credit going to zombie firms, but also improves the business climate and resource availability for healthy borrowers who then compete with fewer or smaller zombie firms. Our study, however, does not claim that a lending specialisation of 100% should be preferred for banks as we only provide a partial analysis rather than a general equilibrium approach. Our results merely show that there are also potential downsides to full diversification of banks’ loan portfolios.
Our findings also have implications for when an economy faces turbulent times. Crises like the global crisis or the COVID-19 pandemic put a severe strain on the liquidity of many firms. In such circumstances, governments may decide to provide loan guarantees or set up a debt moratorium programme. However, it is important to differentiate between viable firms and zombie firms when providing such aid in order to support the economic recovery. As many firms struggle during such times, it may be beneficial to rely on specialised banks to make this differentiation.
References
Acharya, V, M Crosignani, T Eisert and C Eufinger (2024), “Zombie Credit and (Dis-)Inflation: Evidence from Europe”, The Journal of Finance 79: 1883–1929.
Banerjee, R and B Hofmann (2022), “Corporate Zombies: Life Cycle and Anatomy”, Economic Policy 37: 757–803.
Beck, T, O De Jonghe and K Mulier (2017), “Sectoral concentration and bank performance: new measures and new evidence”, VoxEU.org, 9 May.
Berger, P G, M Minnis and A Sutherland (2017), “Commercial Lending Concentration and Bank Expertise: Evidence from Borrower Financial Statements”, Journal of Accounting and Economics 64: 253–277.
Blickle, K S, C Parlatore and A Saunders (2024), “Specialization in Banking”, Journal of Finance.
Caballero, R J, T Hoshi and A K Kashyap (2008), “Zombie Lending and Depressed Restructuring in Japan”, American Economic Review 98: 1943–1977.
De Jonghe, O, H Dewachter, K Mulier, S Ongena and G Schepens (2020), “Some Borrowers Are More Equal than Others: Bank Funding Shocks and Credit Reallocation”, Review of Finance 24: 1–43.
De Jonghe, O, K Mulier and I Samarin (2024a), “Bank Specialization and Zombie Lending”, Management Science.
De Jonghe, O, K Mulier and I Samarin (2024b), “Counting the undead: a new metric for identifying zombie firms”, VoxEU.org, 30 October.
Draghi, M (2024), The future of European competitiveness, September.
Khwaja, A I and A Mian (2008), “Tracing the Impact of Bank Liquidity Shocks: Evidence from an Emerging Market”, American Economic Review 98: 1413–1442.
McGowan, A, D Andrews and V Millot (2018), “The Walking Dead? Zombie Firms and Productivity Performance in OECD Countries”, Economic Policy 33: 685–736.
Schmidt, C, Y Schneider, S Steffen and D Streitz (2023), “Does Zombie Lending Impair Innovation?”, Unpublished working paper.
