AdobeStock_393712680.jpeg
VoxEU Column COVID-19 Microeconomic regulation

How businesses are surviving Covid-19: The resilience of firms and the role of government support

As the economy locked down in March 2020, businesses across the UK struggled to operate. And yet, fewer firms declared bankruptcy during the pandemic than in preceding years. This column introduces a model designed to examine the economic impact of Covid-19. It determines that government assistance rescued previously profitable firms that might not have survived lockdowns, but also propped up weaker firms that would have failed in normal times. The difficulties in effectively targeting aid justifies the expansive support distributed during the crisis.

As the economy locked down in March 2020, businesses in the UK suddenly found themselves unable to open normally, with retail premises, offices, factories, and construction sites closed. Over the following months, new waves of infections and restrictions continued to disrupt many activities (Bloom et al. 2020). This led to fears that a ‘domino effect’ would ripple through the business sector, leading to widespread job losses and bankruptcies.

The Corporate Sector Agent-Based (CAB) Model (Hillman et al. 2021) is new, large-scale, calibrated, agent-based model designed to examine the economic impact of Covid-19. Agent-based models are simulations in which aggregate outcomes are built up by simulating individual agents interacting across a network. This approach can generate emergent aggregate behaviour and non-linearities that could become highly relevant with exceptional shocks on the scale of the Covid-19 (OECD 2017).

The model takes into account many features of the real-world business sector that depart from standard assumptions: the heterogeneity in production structures and skewed distribution of financial strength across firms; customer-supplier networks where many firms are heavily dependent on a small number of customers or suppliers; rule-of-thumb behaviour by firms as they adjust to incoming orders; and bankruptcy constraints. These features amplify effect shocks and generate substantial persistence and overshooting, as well as displaying a number of non-linearities. The CAB model is calibrated by combining ORBIS firm-level data, the OECD Input-Output tables, and available evidence on the network connections between firms using a new algorithm.

In the absence of any policy response, the Covid-19 shock would have to led to a peak fall in output of 21%, larger than the observed fall (Barnes et al. 2021). Over 15% of firms would have failed in the first two years, 50% more than in normal times and a similar relative increase to that experienced in the Global Crisis. Firm failures and job losses would have led to a substantial shortfall in output, even as restrictions eased. SMEs and firms in sectors highly exposed to Covid-19 would have fared worse: almost a quarter of firms would have failed in the Accommodation and Food sector. 

While this outcome would have been dramatic, the economy may have been more resilient than many expected. Many firms carry cash reserves of at least three months’ sales, which would have been enough to see them through the most severe period of restrictions. Firms’ ability to shed labour and costs increases their individual survivability, although it imposes a cost on their workers and suppliers. A large share of output is accounted for by larger firms that tend to be financially stronger and have more diversified and robust customer and supplier networks. A bigger shock or a less resilient corporate sector could have triggered a more severe outcome (Sharma et al. 2021). Indeed, the analysis from the CAB model is broadly in line with a range of other approaches (Demmou et al. 2021).

While some firms can be brought down by the failure of key customers or suppliers, ‘domino effects’ rarely appear because the conditions needed – a group of closely related firms that are all financially vulnerable to the point of failure – seldom arise. Rather, the reasons for individual firm failure are multi-factored and typically reflect the accumulation of shocks and unfavourable cash flow dynamics, and can depend critically on small differences.

In reality, the massive policy response has been highly effective in supporting output and avoiding firm failure. The model is used to simulate the impact of the UK Coronavirus Job Retention Scheme (CJRS) ‘furlough’ program and a credit guarantee.

The rate of firm failures is predicted to be lower than in normal times, consistent with the observed bankruptcy statistics to date from the UK Insolvency Service and as noted by Djankov and Zhang (2021). This counterintuitive outcome reflects the fact that policy support has rescued not only the vast majority of previously profitable firms that would have failed due to Covid-19, but also more than half of weaker firms that would have failed in normal times.

Figure 1 With massive policy support, the depth and persistence of the Covid-19 loss of output has been reduced

 

Source: Authors’ calculations based on CAB model.

Looking ahead, one risk is that ‘zombie firms’ could drag on aggregate productivity in the years ahead (Cros et al. 2021). First, some lower productivity firms that would otherwise have failed may remain in business longer as a result of the policy supports, although their failure may just have been delayed. Second, permanent changes in demand for specific activities – such as business travel – could mean that previously profitable firms no longer have a viable business model. Many were in good shape financially before Covid-19 and they have been helped by government supports, so there is a risk that these firms remain in business but do not contribute to the dynamicism of the recovery. 

Table 1 Policy supports have protected firms hard hit by Covid-19 and also supported firms that were already vulnerable (% share of firms)

 

Could policy have been better targeted? An approach, for example, of only helping previously profitable firms or those in severely affected sectors would have led to fewer inherently weak firms surviving, but at the cost of lowering incomes for workers and slowing the recovery. Given the many challenges of targeting, the use of broad supports during the Covid-19 crisis appears justified.

References

Barnes, S, R Hillman, G Wharf and D McDonald et al. (2021), “The impact of Covid-19 on Corporate Fragility in the United Kingdom: Insights from a new calibrated firm-level Corporate Sector Agent-Based (CAB) Model”, OECD Economics Department Working Papers, No. 1674, Paris, OECD Publishing.

Bloom, N, P Bunn, S Chen, P Mizen, G Thwaites and P Smietanka (2020), “Coronavirus expected to reduce UK firms’ sales by over 40% in Q2”, VoxEU.org, 20 May. https://voxeu.org/article/coronavirus-expected-reduce-uk-firms-sales-ove...

Cros, M, A Epaulard and P Martin (2021), “Will Schumpeter Catch Covid-19?”, CEPR Discussion Paper 15834.

Del Rio Chanona, R-M, P Mealy, A Pichler, F Lafond and J D Farmer (2020), “Supply and demand shocks in the COVID-19 pandemic: an industry and occupation perspective”, Oxford Review of Economic Policy.

Demmou, L, S Calligaris, G Franco, D Dlugosch, M Adalet McGowan and S Sakha (2021), “Insolvency and Debt Overhang Following the Covid-19 Outbreak: Assessment of Risks and Policy Response”, OECD Economics Working Paper, No. 1651, Paris, OECD Publishing.

Djankov, S and E Zhang (2021), “As COVID rages, bankruptcy cases fall”, VoxEU.org, 04 February.

Gourinchas, P O, S Kalemli-Ozcan, V Penciakova and N Sander (2020), “COVID-19 and SME Failures”, NBER Working Paper, 27877.

Hillman, R, S Barnes, G Wharf and D McDonald et al. (2021), “A new firm-level model of corporate sector interactions and fragility: the Corporate Agent-Based (CAB) Model”, OECD Economics Department Working Papers, No. 1675.

OECD (2017), “Debate the Issues: Complexity and policy making, OECD Insights”, OECD Publishing, Paris, http://dx.doi.org/10.1787/9789264271531-en 

Sharma, D, J-P Bouchaud, S Gualdi, M Tarzia and F Zamponi (2021), “V-, U-,L-or W-shaped economic recovery after Covid-19: Insights from an Agent Based Model”, PLoS ONE, Public Library of Science, 16 (3).

1,680 Reads