VoxEU Column COVID-19 Financial Markets

Quantifying the impact of the COVID-19 pandemic on firms' default probability in Japan

COVID-19 has brought about severe adverse effects on the economy around the globe, and Japan is no exception. This column introduces a model that maps cash shortages to firm's default probability, employing the balance sheet data of about 730,000 SMEs. It uses the model to assesses how a decline in sales due to Covid-19 increases the default probability of firms and how much the government's financial support mitigates a rise in that probability.

Covid-19 has brought about severe adverse effects in economies around the world, and Japan is no exception. People have suspended social interaction at schools, work places, and restaurants and shops, and this has caused an unprecedented level of decline in sales among firms. The government and the Bank of Japan have implemented various measures, the scale of which is also unprecedented, to counter the adverse effects of the pandemic by providing financial support for economic activities.

In our latest Financial System Report (Bank of Japan 2020), we assess the implications of the Covid-19 pandemic for financial stability in Japan, taking into the account the impact of the financial support measures undertaken by the government. We study how the liquidity position and default probability of small- and medium-sized enterprises (SMEs) have been affected, using firm-level granular balance sheet data and the default/non-default figures for 730,000 companies.  We focus on a variable that we refer to as “short-term cash shortages” of individual firms, and show how the pandemic and the financial support aiming to tackle the impact of the pandemic affect the default probability of firms through a change in this variable.

Our work is in line with a growing number of studies that explore the implications of the pandemic for the liquidity position of firms and/or the implications of government financial support for troubled firms. This research includes Banerjee et al. (2020), Bircan et al. (2020), Schivardi and Romano (2020), De Vito and Gomez (2020), as well as analyses in the IMF’s GFSR (October 2020) and Bank of England’s FSR (May and August 2020). Our work supplements these works by highlighting the results found within the granular data set of Japanese SMEs, and by explicitly addressing the mapping from cash shortages to a firm’s default probability.

COVID-19 and short-term cash shortages

Figure 1 illustrates our process of measuring the magnitude of firms’ potential liquidity shortage due to the pandemic and the impact of the corporate financing support provided by the government and the Bank of Japan. We estimated initial cash reserves at the beginning of 2020, as well as the operating cash flows that occurred in 2020. When a firm's operating cash outflows exceed its initial cash reserves, this metric – which we call the “net cash surplus” – takes on a negative value and the firm faces a potential cash shortage. Firms with cash shortages are assumed to receive the full amount of their eligible support to weather this liquidity stress.

Figure 1 Impact of COVID-19 on firms' operating liquidity


Our setting of the variables for the 730,000 SMEs is similar to that sound in previous works such as Banerjee et al. (2020). Operating cash flow is the difference between what firms earn from selling their products/services and costs paid for their production inputs including employees, facilities, and raw materials. Covid-19 is considered to affect the size not only of sales, but also of costs, because firms may downsize their production endogenously when facing a decline in sales. Our estimate of the decline in sales uses industry-level data from the Bank of Japan's Tankan business survey. Based on our estimates, the pandemic reduced SME sales by 20% on average for the 2020 financial year. This effect was larger for some industries such as food, accommodation, and consumer services. Regarding costs, we estimate an elasticity of hiring costs, rent, and payment for raw materials, to a change in sales by industry. Based on our estimates, assuming that the costs and sales of a firm are initially both ¥1,000, the firm reduces costs by ¥80 when its sales decline by ¥100. In considering the amount of financial support, we focus on cash payments and financial measures. Cash payments are the subsidies mainly for SMEs that see a large decline in sales. The aggregate size of the payments amounts to ¥8.4 trillion. ‘Financial measures’ are loans whose interest rates are covered by the government, consisting of loans extended by private financial institutions (¥53 trillion) and those extended by government financial institutions (¥47 trillion). These interest-free loans are secured by the public guarantee programmes whose fees are covered by the government.

Figure 2 shows the total amount and distribution of cash shortages among SMEs. Due to Covid-19, cash shortages are expected to rise from about 1.1 to ¥3.3 trillion, reflecting the fact that firms are unable to downsize their production scale after a decline in sales. With cash payments and interest-free loans, these shortages shrink by 1.4 and ¥1 trillion, respectively. In terms of distribution, while Covid-19 increases the cash shortages of firms, most of them see the shortages less than 10¥ million – far below the maximum amount of interest-free loans. Cash payments – whose amount received for each SME can be greater than ¥5 million – restore the bulk of the shift in distribution.   

Figure 2 Impact of support measures on cash shortages (SMEs)

a) Total amount of cash shortages


b) Distribution of cash shortages


Source: BOJ (2020).

From cash shortages to defaults

We next explain how the short-term cash shortages are translated into a firm's default probability. Theoretically, there is a difference between short-term liquidity problems and becoming insolvent. Firms have various ways to get around a default when they are short of cash. We therefore need to examine the data. 

In Figure 3 we arrange firms by their net cash surplus normalised by total assets and plot them in ascending order on the x-axis, with the corresponding default rate on the y-axis. Cash shortage does indeed go hand-in-hand with default. As the net cash surplus becomes smaller, the default probability of firms gradually rises. In particular, firms see a disproportionately high default rate when they fall into the bottom 20% of the distribution. 

Figure 3 Relationship between net cash surplus and default rate (SMEs)


Note: 1. Default is defined as being downgraded to the borrower classification "special attention" or below, overdue, or subrogated in the following 1-year period. The reference period is from fiscal 2002 to 2018. 2. Firms are grouped into 1-percentile bins based on the net cash surplus. The vertical axis shows the default rate for each bin. The vertical line indicates the percentile bin where the net cash surplus is zero.
Source: BOJ (2020), CRD Association.

Clearly, we need to control for the effects of other variables that can potentially affect firms’ default probability. Conceptually, for example, default rates may likely be explained by the degree of firms’ solvency represented by variables such as interest coverage ratio and leverage ratio. We therefore regress the default probability of a firm on cash shortages as well as on other candidate variables of the firm. We find that cash shortages do indeed predict a higher default probability after controlling for the effects of other candidate variables.

Combining the results shown in Figure 2 and the estimation results, we are now in a position to quantitatively assess the impact of the pandemic and of the financial support on a firm’s default probability. Figure 4 shows the change in a firm's default probability due to these two factors. Two observations are noteworthy. First, the impact of the pandemic is substantial, increasing default probability by one percentage point on average if no financial support measures are undertaken. In particular, default among food, accommodation, and consumer services increases by around three percentage points. This easily exceeds the rate of about two percentage points that was observed during the global crisis. Second, the impact of the financial support is also substantial. Cash payments seem to play a dominant role. In all industries, cash payments alone lower the default probability by around one percentage point. This almost fully offsets any upward pressure on the default probability that stems from sales declines. Indeed, this prediction is supported within the data. On average, the default rate of firms up to now has been around 1.1% in 2020, which remains close to the figure for 2019.

Figure 4 Impact of cash shortages and support measures on PD by industry (SMEs)


Source: BOJ (2020).


Our simulation results show that the magnitude of firms' potential liquidity shortage caused by the spread of Covid-19 was significant, especially in sectors such as food, accommodation and consumer services. The public support measures in response, however, almost fully offset any upward pressure on the default probability caused by liquidity stress.

It is also notable that Japanese firms have been building up cash and capital steadily over the last two decades, possibly due to a low natural rate of interest, precautionary motives, or both in the case of the former (as shown in Figure 5). At the beginning of the current crisis, the cash ratio had risen by about 30 percent points since the global crisis ended. This secular increase in cash and capital has also improved the resilience of non-financial corporates against the severe sales decline.      

Figure 5 Capital ratio and cash ratio of firms by region


Note: 1. "Capital ratio" and "Cash ratio (= cash and cash equivalents / current liabilities)" are the median of the financial indicators of major companies included in S&P Capital IQ. 2. "Europe, etc." includes the Middle East and Africa.
Source: BOJ (2020), S&P Global Market Intelligence.

Admittedly, as of writing, developments in Covid-19 and in safe and effective vaccines remain highly uncertain. The current analysis, however, provides a basis to assess the implications of these developments for financial stability.  

Authors’ note: The views expressed are those of the authors and do not necessarily represent those of the Bank of Japan.


Banerjee, R, A Illes, E Kharroubi and J-M Serena (2020), “COVID-19 and corporate sector liquidity”, BIS Bulletin 10, Bank for International Settlements.

Bank of England (2020a), Financial Stability Report, May 2020.

Bank of England (2020b), Financial Stability Report, August 2020.

Bank of Japan (2020), Financial System Report, October 2020.

Bircan, Ç, R De Haas, H Schweiger and A Stepanov (2020), “Coronavirus credit support: Don’t let liquidity lifelines become a golden noose”,, 03 June. 

De Vito, A and J-P Gomez (2020), “COVID-19: Preventing a corporate cash crunch among listed firms”,, 29 March.

International Monetary Fund (IMF) (2020), Global Financial Stability Report, October 2020.

Schivardi, F and G Romano (2020), “A simple method to estimate firms’ liquidity needs during the COVID-19 crisis with an application to Italy”, COVID Economics.


1 We have conducted a similar analysis for 2,400 firms listed in the Financial System Report. We show that those firms are robust to liquidity problems due to both high initial cash reserves and relatively moderate sales declines.

2 For example, among the financial support measures, “subsidies for sustaining businesses” are paid up to 2 million yen to all SMEs that meet pre-specified conditions such as facing a severe sales decline of 50% in a certain month in 2020.

1,575 Reads