Discussion of “The Asymmetric Impact of COVID-19: A Novel Approach to Quantifying Financial Distress across Industries”

Abstract

The COVID-19 pandemic crisis and the associated lockdown measures have exerted significantly adverse effects on corporate sectors globally. Archanskaia et al. (2023) provide a novel empirical strategy to timely assess corporate financial distress in the EU. The contribution is two-fold. First, this paper’s notion of financial distress considers both the equity position and corporate indebtedness. Second, the methodology proposed in this paper allows the authors to estimate corporate financial distress in the EU at a highly granular level and link micro-level simulations to sectoral macroeconomic outcomes. The methodology employed by Archanskaia et al. (2023) consists of three steps. First, the authors apply a nowcasting model to acquire monthly industrial turnover data. Second, they feed the obtained monthly industrial turnover into a profit-generating process via an accounting identity to estimate monthly firm profits at the firm level. Third, the authors use the estimated firm profits with a snapshot of information on pre-existing liquid assets to deduce the firm-level liquidity needs and the depletion of equity through the focus period during COVID-19. These estimated results on firm equity position and indebtedness enable the authors to quantity corporate financial distress in the EU via various angles (e.g., country-level heterogeneity, industry heterogeneity, and the targeting of COVID support policies). The primary advantage of this approach is that it deals with large datasets at the granular level and produces firm-level results almost in real-time. Therefore, it can help policymaking track the effects of crises over time. However, one can quickly critique this three-step approach for its susceptibility to the usual Lucas critique. That said, since the objective here is to estimate firm-level financial distress, a large structural model being more or less aggregate in nature, though able to mitigate the Lucas critique concern, will encounter significant challenges in estimating firm-level results with the requisite level of granularity offered by the available data. Therefore, I broadly concur with the authors’ position that ‘the specific contribution of this paper consists in striking a better balance between the need to carry out a multi-country evaluation of the pandemic’s effects on industrial activity in a strongly integrated region like the EU and the difficulty of capturing time, industry, and country variation in turnover with sufficient granularity.’

The COVID-19 pandemic crisis and the associated lockdown measures have exerted significantly adverse effects on corporate sectors globally. Archanskaia et al. (2023) provide a novel empirical strategy to timely assess corporate financial distress in the EU. The contribution is two-fold. First, this paper’s notion of financial distress considers both the equity position and corporate indebtedness. Second, the methodology proposed in this paper allows the authors to estimate corporate financial distress in the EU at a highly granular level and link micro-level simulations to sectoral macroeconomic outcomes.

The methodology employed by Archanskaia et al. (2023) consists of three steps. First, the authors apply a nowcasting model to acquire monthly industrial turnover data. Second, they feed the obtained monthly industrial turnover into a profit-generating process via an accounting identity to estimate monthly firm profits at the firm level. Third, the authors use the estimated firm profits with a snapshot of information on pre-existing liquid assets to deduce the firm-level liquidity needs and the depletion of equity through the focus period during COVID-19. These estimated results on firm equity position and indebtedness enable the authors to quantity corporate financial distress in the EU via various angles (e.g., country-level heterogeneity, industry heterogeneity, and the targeting of COVID support policies).

The primary advantage of this approach is that it deals with large datasets at the granular level and produces firm-level results almost in real-time. Therefore, it can help policymaking track the effects of crises over time. However, one can quickly critique this three-step approach for its susceptibility to the usual Lucas critique. That said, since the objective here is to estimate firm-level financial distress, a large structural model being more or less aggregate in nature, though able to mitigate the Lucas critique concern, will encounter significant challenges in estimating firm-level results with the requisite level of granularity offered by the available data. Therefore, I broadly concur with the authors’ position that ‘the specific contribution of this paper consists in striking a better balance between the need to carry out a multi-country evaluation of the pandemic’s effects on industrial activity in a strongly integrated region like the EU and the difficulty of capturing time, industry, and country variation in turnover with sufficient granularity’.

The findings presented by Archanskaia et al. (2023) suggest the need to refine the criteria used to assess firm viability due to the pandemic. Their proposed methodology can offer near real-time information for the targeting of policy support for corporate sectors during crises. The targeting of policy support during unprecedented crises such as COVID-19 is notably challenging because in times of severe crises, lending to non-financial firms is subject to adverse selection (see asymmetric information issue in COVID support policies in Vardoulakis, 2020, Li and Li, 2021, Wang and Wang, 2021, Goodhart et al., 2023). The government and even the lending banks may not observe the borrowing firms’ own funds and the actual cost firms incur due to the pandemic crisis, but borrowing firms themselves have better knowledge, if not full information, of these two variables. Let me draw on a sketch of the theoretical model developed by Wang and Wang (2021) to conceptualise the asymmetric information issue and demonstrate why some of the government support programmes implemented during the pandemic crisis lack effective targeting.

Let us suppose the entrepreneur managing the firm is characterised by type $(a,x)$, where $a$ is the entrepreneur’s own funds, which resembles the liquidity buffers in Archanskaia et al. (2023), and $x$ is the cost due to the pandemic crisis. The entrepreneur’s type is her private information. Further, suppose $V$ is the firm value, and $c$ is the entrepreneur’s effort cost if she works diligently. Then $a-x$ indicates the liquidity shortage in Archanskaia et al. (2023), and when $V-c-x$ is negative, it indicates insolvency in Archanskaia et al. (2023). As we shall shortly see, due to agency frictions, even when $V-c-x$ is positive, without government support, some firms may still fail to survive due to liquidity shortage, which captures the risk of insolvency.

In Fig. 1, line $L_{IC}$ is the equality form of the incentive compatibility constraint where $k_e$ is the pledgeable income, or the maximum an entrepreneur could borrow from a bank. The entrepreneurs that satisfy the incentive compatibility (IC) constraint have sufficient skin in the game to work diligently. They will have access to bank lending if they cannot survive with self-funds (to the right of the self-funding line $L_{SF}$). However, because bank lending is costly (gross rate $R>1$), only those that satisfy the individual rationality (IR) constraint $L_{IR}$ will choose to borrow from banks. In sum, only entrepreneurs that need external funds and satisfy both the IC and IR constraints, namely, those in the light green area, will and are able to borrow from banks. Those in the dark red area, even though they have positive NPVs, are not able to obtain bank lending. For them, the liquidity shortage causes insolvency if there is no government support.

Fig. 1.

The types in the light green area survive by borrowing from banks. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Wang and Wang (2021).

How should government support target those in the dark red area? The difficulty lies in that firm’s funding gap or liquidity shortage cannot be well observed. In this example, the difference between the actual cost due to the crisis $x$ and the borrowing firm’s own funds $a$ is its funding gap ($x-a$). The higher the funding gap is, the more the firm is in financial distress. It is unlikely that the government has sufficient funding to support all the firms in financial distress without incurring significant fiscal costs; therefore, it is important to target specific types of firms that government wishes to save. Many of the policies implemented during COVID have the objective of protecting employment. If this is the aim, then for similar-sized firms, the government should target those whose funding gap is small while disqualifying those whose funding gap is large, so that given a limited public budget, the number of firms that can be saved is large, and hence, more jobs are protected.

With government support funds, Wang and Wang (2021) show that if the objective is to protect jobs, then the optimal policy should target entrepreneurs in the light green area in the right-side graph of Fig. 2. This is because these types have positive NPVs and have smaller funding gaps ($x-a$) than the other positive NPV firms which cannot survive without public support. This allows a given supply of public goods to save a larger number of firms and protect more jobs. In practice, however, many existing COVID-19 support programmes deviate from this optimal policy. Those programmes typically place a cap on the size of loans that are subsidised. Besides the subsidised loans, firms are free to borrow any additional amount from banks through unsubsidised loans. The effect, as Wang and Wang (2021) show, is equivalent to moving both the $L_{IC}$ and $L_{IR}$, as well as the self-funding line $L_{SF}$ to the left (see the left graph of Fig. 2). This strategy saves those in blue and green areas, but those in the area $ABD$ have smaller funding shortfalls and have positive NPVs are excluded. In net, the number of firms and jobs saved in the left graph is smaller than that saved in the right graph, which suggests an inefficient deployment of public funds, and the macroeconomic implications could be capital misallocation for employment stabilisation or subsequent surge in inflation to put credit risks under the cover.

Fig. 2.

Comparing existing programmes with optimal targeting. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Wang and Wang (2021).

Theory only points to the broad direction of optimal targeting of public support, and in my view, the approach developed in Archanskaia et al. (2023) is complementary to theory and has the potential to provide more accurate and timely information on $x$ and $a$ from an empirical perspective, ideally by generating firm-level empirical estimates to inform policymaking in real-time. For example, Graphs 3.6 to 3.9 display the share of EU firms suffering losses and/or becoming illiquid by industry and by countries. These exercises provide information on $a-x$, and $a$ and $x$ are the source of adverse selection. Nevertheless, it is not yet clear how one can get a clear sense of the intrinsic value of $x$ and $a$ separately, without the confounding factors stemming from policy. Perhaps, it is beyond the scope of this paper to give a concrete answer, but the approach developed by Archanskaia et al. (2023) allows real-time assessment of vulnerabilities in the non-financial corporate sectors, and it would be useful to adapt the approach to support the future design of targeted policy measures.

Furthermore, amongst the wide range of detailed results reported in Archanskaia et al. (2023), interesting to note is the significant country heterogeneity due to the asymmetric nature of the pandemic crisis (see, e.g., Graph 3.15). These results give rise to concerns of divergence in the Euro Area in particular and underscore the need for greater risk-sharing in a monetary union that lacks a fully-fledged fiscal union (see Sargent, 2012, Farhi and Werning, 2017). In fact, the country heterogeneity in terms of both financial profiles and the real economy has widened between countries in the Euro Area since the eurozone debt crisis (Fig. 3). One way to improve risk sharing in a currency union that lacks a fully-fledged fiscal union is to introduce additional means of discharging debt, which essentially introduces more state-contingencies to the monetary union. Accordingly, the European Commission’s Capital Markets Union (CMU) aims to facilitate cross-border risk sharing via capital markets, and the European Commission sets out legal directives on the cross-border insolvency law as a key legal foundation of CMU. The findings of Archanskaia et al. (2023) suggest that the COVID shock may further necessitate these policy and legal reforms.

Fig. 3.

NPLs and unemployment profiles in eurozone.

Statistical Data Warehouse, ECB, Eurostat database and OECD Economic Surveys: Euro Area 2018, calculated as unweighted average. GIIPS: Greece, Ireland, Italy, Portugal and Spain.

Finally, the notion of corporate financial distress in this paper considers both the equity position and firm indebtedness. The granular results on corporate leverage and liquidity, in my view, make a natural and relevant link to the monetary policy transmission mechanism, albeit it is not the focus nor the objective of this paper. The current tightening of monetary stance in advanced economies with a large volume of corporate legacy debt has attracted public comments and debate. Recent literature suggests that corporate indebtedness and the firm distance to default matter significantly for monetary policy transmission (see, e.g., Gomes et al., 2016, Ottonello and Winberry, 2020, Jungherr et al., 2022). For example, Ottonello and Winberry (2020) show the mechanism in which corporate default risks affect the marginal benefit and marginal cost of capital and find that the investment of low debt firms or those with a high distance to default is more responsive to expansionary monetary policy shock. Goodhart et al. (2021) demonstrate that higher levels of corporate indebtedness dampen the effect of monetary contractions on controlling inflation. With the intuition built upon this literature, the estimated results in Archanskaia et al. (2023) on corporate financial vulnerability and the associated country heterogeneity can shed a quick light on how ECB’s post-pandemic monetary tightening may affect output and inflation differently amongst countries in the Euro Area. In sum, Archanskaia et al. (2023) is a welcome contribution to our understanding of the heterogeneous impact of the COVID-19 pandemic on the financial distress of the EU non-financial corporate sector.