Guest post by Eoin O’Leary, Emeritus Professor of Economics, Cork University Business School, UCC.
There has been vast media coverage on the effects of the COVID 19 pandemic on the Irish economy, most of which has been speculative as the struggles of workers, businesses and government has been ongoing. Yet the recent publication of the monthly Retail Sales Index (https://www.cso.ie/en/statistics/services/retailsalesindex/) and the Monthly Services Index (https://www.cso.ie/en/statistics/services/monthlyservicesindex/) for October provides a fascinating insight into how sales in customer-facing businesses in 21 retail and market services sectors have been affected by the pandemic since the beginning of 2020.
These sectors, which employed 1.1 million workers in the Republic of Ireland in 2019, have been directly affected by the sudden onset of the first lockdown of many shops and services enterprises in March, followed by graduated re-openings with social distancing measures and increasing reliance on remote access for customers.
The period January to October 2020 is ideal for investigating the 1st wave of the pandemic, as it covers the first lockdown in March and the subsequent re-opening during the summer before the 2nd lockdown on 19th October 2020. The sectors being investigated are major contributors to the Irish economy. Together they employed over 1.1 million workers in 2019, which represents 49% of total employment in the state.
The first key date in the timing of the response to the pandemic was the government-imposed lockdown starting on 27th March. Apart from retail enterprises involved in food, fuel and pharmacy (ie Food, Beverages and Tobacco in Supermarkets; Food, Beverages & Tobacco in Specialized Stores; Fuel and Pharmaceuticals, Medical & Cosmetic Articles) which remained open throughout 2020, all other retail sectors closed from this date with some providing remote access for customers. This was followed by a graduated re-opening (with social distancing measures continued) from June to mid-October. There was variation, with for example Bars (not selling food) being closed until September 21st, while mainstream non-food retailing (ie the remaining 8 retail sectors) re-opened from June 8th.
In market services, food and accommodation services (ie Accommodation and Food Services Activities) were closed from 27th March and re-opened on June 28th. There were differences within the remaining sectors. For Wholesale Trade those businesses involved in food, fuel and pharmaceutical distribution would not have been significantly affected during the 1st lockdown, while non-food retailers would have experienced a downturn. For Transport and Storage, passenger transport was severely curtailed while, anecdotally, there is evidence that courier services prospered. The other 4 sectors (ie Information and Communication; Professional Scientific and Technical Activities; Administrative and Support Services Activities and Other Business Services) were influenced by having to serve customers almost exclusively via remote access.
The Data and Method
Table 1 shows detailed descriptions of the sectors covered and their 2019 employment levels. For the Retail Sales Index, sales from a sample of 1,700 enterprises of varying sizes are collected each month across 13 sectors and indices generated. In addition to sales, in relevant sectors, data are collected on the % of sales from online sales of enterprises with a presence in Ireland. This source provides very detailed and comprehensive coverage of businesses of all sizes in retailing, accounting for 253 thousand persons engaged in 2019 (22% of the 1.1 million covered in the paper). One exclusion is that the index does not cover the retail trade of food and non-food via stalls and markets.
the Monthly Services Index, sales from a sample of 2,250 enterprises
with more than €20 million sales and more than 100 persons engaged is
collected. In addition to the index only
covering these larger enterprises, the 8 sectors are more broadly defined than
retail (see Table 1). These sectors
covered 876 thousand persons engaged or 78% of the total in 2019. Notable exclusions from this series are
financial and insurance activities; public administration; education; human
health; creative arts, music and entertainment activities. Together, both indices provide interesting
insights into a substantial portion of the Irish economy that has been most
affected by the pandemic.
order to estimate the effect of the pandemic on each of these 21 sectors from
January to October 2020, the actual monthly value index is compared to what the
index would have been if no pandemic occurred.
The latter is estimated using a simple model which first estimates the
trend of the value index for each sector, calculated over the previous 5 years,
for the 9 months starting in February 2020.
This month is taken as the beginning of the pandemic because there is
evidence, especially in food retailing, that customer behaviour shifted during
that month when it became clear that a lockdown was imminent.
trend is computed using the average annual percentage change in the seasonally
adjusted value index from January 2015 to January 2020 which is applied to the
February to October 2020 indices. It
then adjusts the resultant series for each sector by monthly seasonal
factors. These are derived by dividing
the seasonally adjusted index by the unadjusted value index for February 2015
to October 2019 and calculating the average seasonal factor for each of the 9
months over these 5 years (see https://www.cso.ie/en/statistics/services/). The resultant seasonal factors
are then applied to the estimated trend from February to October 2020.
effect of the pandemic is assumed to be the difference between the estimated
trend with seasonal variation and the actual unadjusted value index for each of
the 9 months being analysed. It is
assumed that the pandemic is the chief irregular component effecting the sales
indices in these 21 sectors. There can
be little doubt that it was by far the most significant abnormality that
prevailed during 2020.
Overall, of the 21 sectors investigated, actual monthly sales are
estimated to have been down in 16 sectors, with 5 either unchanged or doing better than they would have
if there was no pandemic.
By far the most negatively affected were Accommodation Services, which relates to
hotels and other kinds of accommodation, down a massive 72%, and Bars
Other very large declines were in Transport and
Storage (-37%); Books, Newspapers and Stationery (-32%); Clothing
Footwear and Textiles (-30%); Fuel (-25%); Food Services, which
includes restaurants and cafes (-24%) and Department Stores (-23%).
7 sectors were down between 3 and 18%, namely: Furniture
and Lighting (-18%); Motor Trades (-12%); Administrative and
Support Services (-11%); Professional, Scientific and Technical Services
(-10%); Wholesale Trade (-8%); Other Retail (-7%) and Electrical
2 sectors were unaffected, namely Information
and Communication and Pharmacies.
Notably, 3 sectors experienced higher sales than
they would have experienced without the pandemic. These were predictably Specialized Food
Stores (+11%), and Supermarkets (+7%) and, perhaps less
predictably, Hardware, Paints & Glass stores (+9%).
Figures 1–21 present detailed graphs of the actual value index for each sector from January to September 2020 and the estimated trend and seasonal variation, using the simple model described above. The base is January 2020 = 100 to facilitate comparison between sectors. The average monthly difference in percentage terms between the estimated and the actual is presented under the heading of each graph, to signify the magnitude of the effect of the pandemic. This magnitude is used to determine the order in which sectors are presented, beginning with those whose sales are down the most due to the pandemic and ending with 3 sectors whose sales improved in the pandemic. The commentaries beside each Figure bring out the highlights with suggestions of possible explanations. Where available the trends in the % of sales from online sales are included in the commentaries for retail sectors.
These findings provide fascinating insights on the effect of the COVID 19 pandemic on the sales performance of Irish customer-facing businesses in 21 sectors during the 1st lockdown and subsequent
re-opening in 2020. Three points are
worth making in conclusion:
The enormous variations evident here point to the huge complexity of the effects of the pandemic on Irish businesses providing livelihoods for very large numbers of owners and workers. It suggests that the government’s on-going response to this unprecedented public-health crisis should be informed by analysis such as this, as it comes under increasing pressure to withdraw supports and subsidies due to limits on its borrowing.
email: firstname.lastname@example.org. The author would like to thank
Eleanor Doyle and Owen O’Brien for
discussions that inspired this work.
A guest post by Niall McGeever (Central Bank of Ireland) on new company registrations and corporate insolvency in Ireland during the pandemic. [Disclaimer: This blog represents the author’s views and not those of the Central Bank of Ireland]
The severity of the COVID-19 shock
and the modest liquid asset holdings of many Irish firms (Financial Stability Review 2020 I; McGeever et al., 2020) raises the
question of how the pandemic is affecting business dynamism and failure rates. A
marked reduction in new firm formation or a spike in insolvencies could lower
the productive capacity of the economy and negatively affect output and
There’s lots of economic research
showing the importance of new firms for productivity and employment growth. Lawless (2013),
for example, shows that young firms contribute disproportionately to employment
growth in Ireland.
While a certain level of insolvency over time is inevitable
and even desirable to ensure resource re-allocation to productive firms, the
failure of otherwise viable firms due to the pandemic could reduce output and
productivity growth. See Lambert
et al. (2020) for more discussion on this point.
The chart below, Figure 2 from the Letter, shows the new company registration rate between January 2001 and September 2020. The rate averages around 9.5 per cent per annum and is broadly pro-cyclical.
The initial Covid-19 shock coincided with a sharp decline in new
company registrations, with the rate falling to 5.3 per cent in April and 6.1
per cent in May. The Companies Registration Office tell us that over 90
per cent of applications to register a new company are made online, so the
decline in April and May cannot be explained by procedural delay due to the
pandemic. Instead, it likely reflects a temporary decline in both new
enterprise formation and stalled investment decision-making by pre-existing
The largest declines during this period were, perhaps
unsurprisingly, in Accommodation and Food and in Arts, Entertainment and
Recreation. New registrations in these sectors were down 50 per cent on the
same period in 2019.
Whilst the number of registrations in the first nine months of
2020 were down around 12 per cent on the same period of 2019, new company
registrations rebounded quite strongly over the summer and had returned to
roughly pre-pandemic levels by September. An emerging trend in the
Wholesale and Retail trade category is the consistent increase in new
registrations in “retail sales via mail order houses or via internet” and in
“other retail sales not in stores, stalls or markets” between June and
September relative to the same period in 2019. This trend is also reflected
Census Bureau data, for example, shows higher new business applications by
non-store (e.g., internet sales) retailers during 2020.
We next look at insolvent liquidations. The next chart (Figure 4 in the Letter) shows the insolvent liquidation rate from January 2001 to September 2020. The rate generally tracks macroeconomic conditions very closely and it is worth noting that it rose notably rose with the unemployment rate in early 2008.
immediate impact of Covid-19 shock was to sharply reduce insolvent liquidations. The annualised rate was
exceptionally low at 0.07 per cent in April 2020 and only a touch higher at
0.10 per cent in May. This is due principally to the inability of company
directors to safely convene creditors’ meetings. Prior to the pandemic, it was
a requirement to hold a physical meeting with creditors to initiate a
creditors’ voluntary liquidation. This became impractical during the acute
phase of public health restrictions and so the main channel for insolvent
liquidations was blocked. This procedural issue was quickly resolved and the
Oireachtas passed a company law amendment to facilitate creditors’ meetings by
liquidation rate reverted to pre-pandemic levels in June and showed no signs of
a marked increase up to September. At a sectoral level, Accommodation and Food
and Wholesale and Retail Trade show signs of higher liquidations both during
the pandemic and relative to 2019. These patterns are aligned with the negative
labour market shocks in both sectors.. To a lesser extent, we also see the Arts
and health sectors recording higher numbers.
clear evidence of financial distress facing many firms, there is no evidence yet
of a marked increase in corporate insolvencies. The striking contrast between
the insolvent liquidation rate and current labour market conditions is unusual
and points to the significant role of government supports, loan payment breaks,
and forbearance from other creditors in helping firms to stay cash-flow
Today the Bank published its fourth and final Quarterly Bulletin for 2020. The report contains a detailed overview of developments in the economy since the publication of last Bulletin in July as well as our latest macroeconomic forecasts out to 2022.
The forecast for GDP growth has been revised upwards to -0.4 per cent in 2020 reflecting more positive developments in consumption, strong export performance and an enhanced level of fiscal support arising from the July stimulus package. Growth prospects for next year and 2022 are more subdued compared to the previous Bulletin due to the implications of a WTO Brexit. As outlined Box A, a disruptive transition to a WTO trading relationship would frontload associated output and employment losses. In this baseline scenario, the growth rate of the Irish economy is 2 percentage points lower in 2021 relative to a Free Trade Agreement due to the introduction of tariff and non-tariff barriers. The ILO unemployment rate is projected to average 5.3 per cent for this year, rising to 8 per cent in 2021 following the closure of income-support schemes at the end of the first quarter (Box D in the Bulletin discusses the challenges that arise for measuring unemployment in the time of COVID-19).
Since re-opening from a period of lockdown, the recovery of the
Irish economy has been uneven as levels of domestically focussed economic
activity remain well below pre-pandemic levels. In particular, consumer-facing services
sectors, such as tourism, hospitality and retail services, which are also more
labour-intensive, have been slower to recover contributing to a projected
decline in underlying domestic demand of 7.1 per cent this year. The strong
performance of exports, which are expected to decline by just 0.3 percent in
2020, is the main factor driving an upward revision in the baseline projection for
GDP. Box C details the
relative resilience of high-value exports such as computer services and
pharmaceuticals during a period of declining trade-weighted world demand.
The Central Bank’s Business Cycle Indicator (BCI), a monthly
summary indicator of overall economic conditions estimated from a larger dataset
of high-frequency releases, fell sharply during the months of March and April
reaching a historical low (Figure 1). The latest estimates show that economic
conditions continued to improve into July and August, but the rate of recovery has
slowed down. Despite the improvement over the four months to August, the
overall level of the BCI remains substantially below that observed prior to the
emergence of the COVID-19 crisis.
Figure 1: Business Cycle Indicator (BCI) for Ireland’s Economy
The outlook remains highly uncertain, depending not only on the economic
consequences of COVID-19 and its containment, but also on the nature of the
trading relationship between the EU and the UK. Recognising this uncertainty, Box E analyses the
impact of a ‘severe’ COVID-19 scenario as an alternative to the baseline
forecasts in which there is a strong resurgence of the pandemic, leading to the
restoration of widespread and stringent containment measures for a more
prolonged period. Underlying domestic demand is projected to fall by 8.5 per
cent in 2020 in this case with a continued contraction of -1.3 per cent into
2021. While the economy does not begin to recover until 2022, underlying
domestic demand remains 6 percentage points below 2019 levels. In the ‘severe’
scenario, the unemployment rate rises to 12.5 per cent in 2021 before
moderating to 10.1 per cent the following year.
A guest post by Fergal McCann (Central Bank of Ireland) on SME finances and firm supports during the pandemic.
The Central Bank has published a Financial Stability Note, written by Derek Lambert, Fergal McCann, John McQuinn, Samantha Myers and Fang Yao, entitled “SME finances, the pandemic, and the design of enterprise support policies”. The Note estimates the likely losses that are being experienced in the SME sector over from March 2020 to year end, introduces a model of SME financial distress which can be used to evaluate the effect of announced SME support policies, and discusses policy implementation issues in the current climate.
Aggregate revenue shortfalls We begin by updating an aggregate model of SME revenue shortfalls. McGeever, McQuinn and Myers (‘MMM’, 2020) used a first variant of this model to estimate three-month initial liquidity needs for SMEs of €2.4bn to €5.7bn. We update this model to account for reported reductions in wage and non-wage costs, using both PUP/TWSS take-up rates and CSO surveys on the business impact of COVID-19. Firstly, we show that, using observed data, our estimates for Q2 are at the very high end of the MMM estimates, suggesting the effect of the COVID-19 shock is about as severe as we were willing to project back in March/April when the initial MMM work was being carried out. Revenue shortfalls for 2020 are estimated between €10.3bn and €11.7bn across the SME sector, which are of course subject to significant uncertainty, both due to the uncertain outlook and the use of firms’ survey responses during the pandemic.
We highlight in the paper that such aggregate revenue shortfall estimates are not necessarily estimates of the size of required government support. These shortfalls already account for wage cost reductions through the TWSS. In aggregate, these shortfalls can be met by a combination of utilisation of pre-existing cash reserves, draw-down of existing credit commitments, new borrowing, additional cost reductions or loss-sharing, or if necessary governmental non-wage grants, reliefs and guaranteed loans. In this vein, we also model that across the board non-personnel cost reductions of 30 per cent could have large effects in reducing the overall shortfalls, suggestive of the importance of burden-sharing and cost efficiencies, along with fiscal support, in addressing the crisis.
Protection or liquidation We outline considerations for designing policy responses to
SME financial distress. Some enterprises entered the COVID-19 shock with
unsustainable business models and during a typical downturn the closure of such
companies can be seen as part of the overall process of economic restructuring
and dynamism. However, identification of such firms is difficult given the
nature of the COVID-19 shock.
There is a risk that, if traditional financial signals were being used, widespread liquidation could arise in the short run. We also point out that there are employee, supplier and customer relationships tied up in all firms, with relationship-specific capital on the line as firms are liquidated. Blanchard, Philippon and Pisani-Ferry (2020) summarises much of our thinking in this area:
“In normal times, policies should help the reallocation process, letting some firms fail and others expand, and helping the reallocation of workers across sectors. These are not normal times, however: many firms may fail because they are insolvent even if they are viable. Given the very high uncertainty, banks may be reluctant to advance credit. Unemployment is extremely high, making it difficult for laid off workers to find other jobs. For these reasons we think that protection (of workers) and preservation (of firms) should be given a higher priority than in normal times.”
Despite the above, we of course acknowledge that those designing SME supports must do whatever feasible to use taxpayer funds as efficiently as possible, given the nature of deficit dynamics during the pandemic. Targeting is difficult, especially in the current climate, but it is not fiscally or economically sustainable for supports to be provided without regard to viability, particularly as the likely duration of the pandemic period elongates.
Firm supports: loans, grants or equity? Outside of the over €5bn that will be used to support wages
through the TWSS and EWSS, there is a 60/40 split between debt and grants in
the announced SME support packages. If tax warehousing is accounted for as a
debt, this rises to 70/30. We point out that the amounts announced are committed funds – in the case of
debt-based supports in particular, final take-up rates are unknown.
We highlight the risk that debt-based supports may have weak
demand from firms wary of borrowing, may lead to debt overhang issues over the
medium term and face implementation issues when channelled through lenders. However,
debt-based supports benefit from lenders’ access to information and incentives
to screen credit risk, in cases where banks retain appropriate levels of risk
(such as the 80-20 split embedded in the Irish Credit Guarantee Scheme).
Direct fiscal supports, such as grants or tax or rate waivers, provide liquidity and support the economy but raise issues regarding costs, targeting and moral hazard. Relative to a guaranteed loan, where costs only arise as defaults occur, grant funding is far more expensive up-front for the taxpayer. One potential option to address this drawback is for the State is to provide equity-like or “conditional grant” injections to SMEs which involve an element of clawback or potential return, lowering the cost of intervention relative to a direct grant. In this light, the proposal of Boot et al. (2020) is worthy of further consideration, where higher future tax rates are agreed in exchange for up-front aid.
A model of financial distress Finally we present key results from a forthcoming model of SME financial distress (McCann and Yao, 2020). The financial distress (FD) indicator is based on SMEs’ capacity to meet losses through cash holdings, or to service interest expenses during the shock. We calibrate 2018-2019 data on SME balance sheets to the revenue and cost reductions reported by SMEs in 2020, and use the model to assess the role of various policy support options.
Relative to a no-policy scenario, we implement firms’ lowering of wage costs, both through TWSS wage supports and the transition of employees to the Pandemic Unemployment Payment (PUP), as well as non-wage policies worth €7.5bn, capturing the role of the credit guarantee, other lending, tax warehousing and enterprise grant policies. When the full package of policies announced in 2020 are included in the model, distress rates fall from 18.7 to 15.8 per cent (or 25.9 to 14.9 per cent when weighting firms by their debt balances outstanding).
Why does a cohort of SMEs remain in financial distress (FD) after policy supports are modelled? There are a number of factors at play. Firstly, the schemes’ total availability of €7.5bn means there is an aggregate maximum on the number of SMEs that can access funds. Secondly, specific schemes have specific maximum amounts, which in the cases of those experiencing the most severe financial effects of the pandemic, may not suffice to alleviate FD. Thirdly, the schemes have reasonably wide eligibility criteria, often related to firms experiencing a fall in revenues of a certain amount. This means that many firms that were never at risk of entering FD by our definition are just as entitled to draw down funds as firms experiencing the greatest losses. This issue of widespread access to funding was a necessary feature of scheme designs across the globe in response to the pandemic.
The latter finding on debt-weighted distress suggests that support schemes will have more beneficial financial stability effects than are visible when looking at a simple share of enterprises falling into financial distress. The greater efficacy of policy in lowering debt-weighted distress relates to the tendency of larger SMEs to have larger debts, implying that these firms draw down larger amounts of total scheme funds available, as well as to the concentration of SME debt among affected sectors such as the accommodation, food, wholesale and retail sectors. The table below reports the impact of the sequential addition of specific policy supports, highlighting that the effects of lowering wage bills through PUP and TWSS are larger than the effect of other supports.
Financial Distress (%), by firm count
Financial Distress (%), by debt balance
+ Income and Wage Supports
+ Tax Warehouse
Finally we show that, relative to currently calibrated support policy, a hypothetical “viability-based” grant system that targets firms based directly on the size of their operating losses, supporting firms closest to viability first, would reduce distress rates to about half the levels modelled under currently-designed policy (comparing the middle and right hand side sections of Figure 1). Such a hypothetical system would prioritise solely the minimization of the financial distress rate, for a given fiscal outlay, and is therefore not intended as a specific recommendation but rather to illustrate the effect of current supports relative to a benchmark model. In practice of course, policy must take on board sector-specific, regional and longer-run considerations that go beyond solely the minimization of financial distress rates.
Source: Model-based estimates from McCann and Yao (2020). Notes: “Targeted Grants” replicate payroll supports modelling from the “Current Supports” scenario, but replace the grant, credit and tax components with a €7.5bn grant that provides support to firms in order of their viability (with firms closest to exiting financial distress receiving support first). By construction, the debt-weighted exercise relates only to firms with debt balances above zero in the 2018-19 data
Conclusions and wider policy issues The model suggests that existing policy supports are likely to have mitigated SME financial distress in some cases, but challenges will remain for a relatively large group of SMEs. From a policy perspective many of the firms modelled as being in financial distress may be viable over the medium term – the identification of FD does not imply enterprise liquidation. This points to the importance of a dual approach to policy for SMEs, where targeted and effective financial support is required in the first instance, but a focus is also placed on the system-wide capacity to restructure the liabilities of potentially-viable firms. This latter step will ensure that the set of firms with the greatest prospects of survival over the medium term are given a chance to trade through the current challenges posed by the pandemic. The Central Bank is a key stakeholder in this process, with oversight of lenders’ approach to loan restructuring as SMEs’ payment breaks begin to expire. The Central Bank’s approach to this process was outlined by Deputy Governor Ed Sibley on Monday September 28th and can be accessed here.
References McCann, Fergal and F. Yao (2020, forthcoming), Modelling financial distress in SME sectors during the Covid-19 pandemic – from liquidity to solvency, Central Bank of Ireland, Mimeo. McGeever, Niall, John McQuinn, and Samantha Myers (2020). SME liquidity needs during the COVID-19 shock. Central Bank of Ireland Financial Stability Note, 2020 No. 2