The perfect storm: What is the impact of Covid-19 on the Scottish hospitality industry? August 2020 - Scottish Tourism Alliance
←
→
Page content transcription
If your browser does not render page correctly, please read the page content below
Executive summary One of the most significant victims of Covid-19 is the tourism and hospitality sector badly affected by travel restrictions and lockdown. There are concerns that many companies in this sector will not be able to recover. This report provides insights into expected default rates in the next twelve months for the Scottish tourism and hospitality. The analysis utilises Wiserfunding expertise in risk modelling and applies its models to a sample of the Scottish tourism and hospitality companies to estimate probability of default (PD) at the company level under three scenarios: baseline, mild downturn and severe downturn. The main findings of our research are: • The sample (and the Scottish tourism and hospitality particular, for large companies the proportion in sector) is dominated by small and relatively young the highest PD band (over 30%) increases from 4.55% firms. The latest financial statements from 2019 show under the baseline scenario to 27.27% in mild down- a relatively healthy risk profile with a good profitability turn and to 68.18% under severe stress. For small (ROA above 5% in 63% of the sample) and a generally companies that are riskier in normal circumstances, low level of debt (debt to equity ratio lower than 1 in the PD levels also increase but with the magnitude 70% of the sample). which is less pronounced (the corresponding increase is from 14.33% to 31.49% and 60.16%). This can be • However, given the deterioration in the economy attributed to the adaptability of smaller companies and the impact of the lockdown on this sector, that enjoy leaner structure and lower amount of the average PD of these firms has more than tangible assets and fixed costs. As such, they can doubled from Dec 2019 to June 2020 and reaches adjust faster to the challenging conditions. 15% even in the baseline scenario. • As for the company age, younger businesses • After stressing the financial inputs and macroeconomic are more vulnerable compared to the more variables to reflect the expectations of the next established ones. The response to the Covid-19 months, we forecast the average level of default shock is also determined by business fundamentals. varying between 28% (mild stress) and 43% More profitable companies are less likely to (severe stress). experience default, and the same applies to • Firms of all sizes are seriously affected. Yet contrary the companies with moderate levels of debt. to our expectations, medium and large companies The highest risk levels are exhibited by young seem to be more sensitive to the shock caused companies with no profit and high levels of debt. by Covid-19 as compared to small businesses. In The analysis in this report has not addressed explicitly the effect of the government support, this will be the subject of further research. However, given the high expected default rates, it confirms that the current government efforts to support the sector (e.g. VAT discount) are going in the right direction. However, we would recommend the support programs to be tailored on the company size to maximise their impact. Business fundamentals should be taken into account too. Firms that show the highest level of adaptability should be rewarded and offered additional support to overcome the crisis, in order to increase the chances of success in the deployment of public funds. Finally, the withdrawal of the current borrowing schemes should be carefully planned in order not to create additional shocks to the companies with high leverage. 2
Contents Executive Summary 2 Model Data Inputs and Scenarios 4 Results 6 • Risk Metrics Results Definitions 6 • Credit Risk Benchmarks by Region and Sector 7 • Overall Sample Results Distribution Comparisons across 3 Scenarios 9 • Sample Distribution Comparisons by Company Size 11 • Sample Distribution Comparisons by Age of Company 15 • Sample Distribution Comparisons by Company’s Profitability 19 • Sample Distribution Comparisons by Company’s Leverage 23 The speed of the economic recovery 27 • Most optimistic: The Z 28 • Still very optimistic: The V 28 • Somewhat pessimistic, and probably more likely: The U 29 • Another possible estimation: The W 29 • Most pessimistic: The L 30 Conclusions 31 Overview of risk modelling methodology 32 Appendix 34 Figures & tables 1. Historic data for turnover: 2001-2020, values are In THSD £ 5 2. SME Z-Score Risk Zone Mapping 6 3. UK Credit Risk Benchmark by region 2019-2020 7 4. Scotland Credit Risk Benchmark by sector 2019-2020 8 5. SME Z-Score Distributions under three scenarios (%) 9 6. Probability of Default Distributions under three scenarios 9 7. Bond Rating Equivalents (BRE) Distributions under three scenarios 10 8. Company Size Overview 11 9. SME Z-Score Distribution by company size – Baseline 11 10. SME Z-Score Distribution by company size – Mild Stress 12 11. SME Z-Score Distribution by company size – Severe Stress 12 12. Probability of Default Distribution by company size 13 13. BRE Distribution by company size 14 14. Company Age Overview 15 15. SME Z-Score Distribution by company age under 3 scenarios 16 16. PD Distribution by company age 17 17. BRE Distribution by company age 18 18. Company Profitability Overview 19 19. SME Z-Score Distribution by profitability (ROA) under 3 scenarios 20 20. PD Distribution by profitability level 21 21. BRE Distribution by profitability level 22 22. Company's Leverage Overview 23 23. SME Z-Score Distribution by leverage under 3 scenarios 24 24. PD Distribution by leverage level 25 25. BRE Distribution by leverage level 26 26. Z-Shaped Recovery 28 27. V-Shaped Recovery 28 28. U-Shaped Recovery 29 29. W-Shaped Recovery 29 30. L-Shaped Recovery 30 31. SME Z-Score Components 33 Table 1. Proposed Stress Factors for downturn scenarios 4 Table 2. Financial Ratios definitions 35 Table 3. Bond Rating Equivalent (BRE) Tier definition 36 3
Model Data Inputs and Scenarios A sample of 5000 Scottish companies was selected The models estimate the Probability of Default (PD) from tourism and hospitality industry sectors using financial ratios (see Table 1), non-financial (SIC2007 codes = 55, 56, 79). This sample is used to variables and macroeconomic indicators. More generate outcomes under three scenarios: information about risk modelling is given in the next section. 1. Baseline scenario Baseline scenario uses the values from the latest 2. Mild downturn available year of financial statements submitted to 3. Severe downturn the Companies House (2018-2019) and corresponding macroeconomic inputs. To model the two downturn scenarios, the values for the financial ratios should be ‘stressed’, i.e. adjusted to reflect the negative effect of the pandemic. Table 1 / Proposed stress factors for downturn scenarios Stress table Numeric Examples £ Variables Stress Factors Baseline Mild Severe (average of 5000 Mild Severe companies in the sample) Financials Total Shareholder -60% -80% £801,695 £320,678 £160,339 Equity Total Assets -30% -65% £2,308,293 £1,615,805 £807,9 03 Turnover -50% -91% £9 67,911 £483,956 £87,112 Short Term +50% +80% £554,102 £831,153 £997,384 Debt Long Term Debt +40% +50% £810,429 £1,134,601 £1,215,644 Cash -40% -89% £137,131 £82,279 £15,084 Working Capital -35% -68% £9,158 £5,953 £2,931 Tangible Assets -20% -60% £1,535,132 £1,228,106 £614,053 Intangible -10% -25% £24,758 £22,282 £18,569 Assets EBITDA -70% -96% £290,087 £87,026 £11,603 Retained -90% -99% £680,718 £68,072 £6,807 Earnings Interest Expense +130% +260% £26,863 £61,785 £96,707 Macro GDP -6% -13% Unemployment -7% -12% Notes: The % reflects the change from the baseline scenario, e.g. if the baseline Total Shareholder Equity is £100, then in mild scenario it drops by 60% or becomes £40, and in severe scenario it drops by 80% or becomes £20. For Financial Ratios definitions, see Table 2 in Appendix. 4
Model Data Inputs and Scenarios As for the mild downturn scenario, we have exam- variable between the peak and through points in the ined the last 20 years of financial accounts for Scottish 2008 crisis as the initial estimates of mild downturn for tourism and hospitality sectors. The worst changes were our current event. observed during the Global Financial Crisis (GFC) Period We then made further conservative adjustments following (2007-2008), and post-GFC recessions followed by the the feedback from the tourist industry experts, and European debt crisis (2009-2011). The tourism industry these are the values in Table 1. also suffered the effects of swine flu in 2009-2010. We took the observed percentage change of each financial Figure 1 / Historic data for Turnover: 2001-2020, values are in THSD £ 7000 6000 5000 4000 3000 2000 1000 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 * Further graphs of other financial indicators in 2001-2020 can be found in the Appendix. For the severe downturn scenario, we initially used an The estimated drop for Accommodation & Food industry additional 50% adjustment as compared to the 2008 in April 2020 is 85%, and this was used as the severe crisis, as it is commonly acknowledged that the current stress factor for P&L inputs: Turnover, Cash, EBITDA, COVID-19 event would bring about a more prolonged Retained earnings. and severe impact on the economy and tourism industry. Similar to the mild downturn scenario, these initial We applied this adjustment to balance sheet financials. estimates were adjusted following the expert advice, As for Profit & Loss (P&L) variables, we took the recent and the final estimates are given in Table 1. values from the Scottish government estimates of GDP (https://www.gov.scot/collections/economy-statistics/ #gdpmonthlyestimates). 5
Results Risk Metrics Results Definitions Our results are presented as SME Z-Score, Bond Rating Equivalent (BRE) and Probability of Default (PD). The SME Z-Score is a risk metric derived from Wiserfunding’s proprietary risk models (which are explained in the last section). The score goes between 0 and 1000 where the higher the score, the better the risk profiles of the companies. Risk zones are provided to help interpreting the number. These are presented in Figure 2: Figure 2 / SME Z-score Risk Zone Mapping 0 – 100 101 – 250 251 – 450 451 – 700 701 – 1000 Distress High Risk Medium Risk Low Risk Lowest Risk Bond Rating Equivalent (BRE) is the transformation of Z-Score in line with the metrics used by credit rating agencies. It represents the credit worthiness of the company with the following risk grades (from the best to the worst): AAA, AA, A, BBB, BB, B, CCC, CC. For more detailed definition, please refer to Appendix Table 3). One of the most popular risk measures that is also obtained from Z-Score is Probability of Default (PD). It is a financial term describing the likelihood of default over a particular time horizon. It provides an estimate of the likelihood that a borrower will be unable to meet its debt obligations. It ranges between 0 and 1 with higher values corresponding to the higher risk of default. We use this measure in reporting the results in Executive Summary and Conclusions, whilst the results in this section present all three measures for completeness. 6
Results Credit Risk Benchmarks by Region and Sector To better understand the relative impact of this pandemic on the financial health of Scottish hospitality industry, we compare the change of credit risk from December 2019 to June 2020 across the UK regions and industries, respectively, as shown in Figures 3-4. The following graphs (Figure 3) provide the comparison of credit risk ratings for different UK regions. The overall estimated Probability of Default (PD) for Scottish economy is going up from 5.30% in December 2019 to 8.07% in June 2020. Figure 3 / UK Credit Risk Benchmark by Region 2019-2020 UK Credit Risk Benchmark by Region (Dec 2019) SME Z-Score PD 320 8% 309 7% 310 6.28% 5.80% 6% 300 5.30% 290 5% SME Z-Score 4.01% 281 280 4% PD 272 270 3% 264 260 2% 250 1% 240 0% England Scotland Wales Northern Irelands Region UK Credit Risk Benchmark by Region (June 2020) SME Z-Score PD 300 10% 285 9.28% 8.74% 250 239 8.07% 231 225 8% 200 SME Z-Score 6% 5.09% 150 PD 4% 100 2% 50 0 0% England Scotland Wales Northern Irelands Region 7
Results Credit Risk Benchmarks by Region and Sector The following graphs (Figure 4) provide the comparison of credit risk rating for different industries in Scotland. The overall estimated Probability of Default (PD) for Scottish ‘Leisure’ sector almost doubled from 7.37% in December 2019 to 15.15% in June 2020, given the deterioration in the economy and the impact of the lockdown on this sector. Figure 4 / Scotland Credit Risk Benchmark by Sector 2019-2020 Scotland Credit Risk Benchmark by Sector (Dec 2019) SME Z-Score PD 350 10% 9.37% 289 301 300 7.37% 261 8% 248 250 224 6.48% SME Z-Score 6% 200 4.89% PD 4.34% 150 4% 100 2% 50 0 0% Services CRE/Construction Leisure Retail/Wholesale Manufacturing Sector Scotland Credit Risk Benchmark by Sector (June 2020) SME Z-Score PD 300 20% 9.37% 301 18% 289 261 250 7.37% 16% 14% 200 248 SME Z-Score 12% 224 150 10% PD 4.89% 6.48% 8% 100 4.34% 6% 4% 50 2% 0 0% Services CRE/Construction Leisure Retail/Wholesale Manufacturing Sector 8
Results Overall Sample Results Distribution Comparisons across 3 Scenarios Figure 5 shows the SME Z-Score distributions under the three scenarios, where the average scores are 263, 165, 93 under baseline, mild stress and severe stress scenarios, respectively. To bring these estimates into perspective, they can be compared to results reported in Figures 3-4, that show the risk distributions for different UK regions and for different sectors in Scotland. Figure 5 / SME Z-Score Distributions under three scenarios (%) Base Mild Severe % of companies SME Z-Score Figure 6 indicates the changes of the companies’ PD and distribution in the assumed downturn conditions. In baseline condition, most companies in the hospitality industry show an average 15% PD, increasing to 25% and 43% after applying mild and severe stresses. Figure 6 / Probability of Default Distributions under three scenarios Base Mild Severe % of companies PD 9
Results Overall Sample Results Distribution Comparisons across 3 Scenarios The below graph illustrates the impact on companies’ Bond Rating Equivalents (BRE) under the stress scenarios. As shown in Figure 7, it indicates the rating deterioration from the baseline scenario to the assumed downturn conditions. In the baseline condition, companies’ BRE is fairly well diversified varying from investment grades to non-investment grades. With a greater magnitude of economic stress applied, the sample’s BRE worsen compared to the baseline with a concentration in the lowest rating grades in the severe stress scenario. Figure 7 / Bond Rating Equivalents (BRE) Distributions under three scenarios BRE CC- CC CC+ CCC- CCC CCC+ B- B B+ BB- BB BB+ BBB- BBB BBB+ 36.50% 40% % of companies within certain scenario 35% 30% 23.00% 22.00% 25% 17.50% 16.00% 20% 13.90% 13.50% 12.60% 12.20% 11.50% 10.60% 15% 10.30% 10.10% 9.90% 9.50% 9.50% 9.20% 8.10% 8.10% 6.90% 10% 6.50% 5.50% 3.60% 3.30% 2.70% 2.40% 1.90% 5% 1.10% 0.90% 0.70% 0.40% 0.10% 0% Baseline Scenario Mild Stress Scenario Severe Stress Scenario 10
Results Sample Distribution Comparisons by Company Size Figure 8 – Figure 13 illustrate the results distribution by Figure 8 / Company Size Overview company size under the three scenarios. To begin with, we categorized companies into three groups: “Small”, Small Large “Medium”, and “Large”, by the level of total assets. Medium 2.20% Medium “Small” presents the companies with total assets less 9.20% Large than £2million, while the ones with total assets between £2 million and £15 million are regarded as “Medium”. Those with greater than £15 million of total assets are regarded as “Large”. Figure 8 illustrates company size distribution within the sample. Small company, medium- size and large company account for 88.6%, 9.2%, and 2.2% of the entire sample, respectively. SME Z-Score Distribution by Company Size under different scenarios are shown in Figure 9 – Figure 11. Small SME Z-Score, which higher value correspond to lower 88.60% risk, could provide us the overview on the sink of the creditability of tourism companies. In the baseline situation as shown in Figure 9, the small and medium companies are seemed to be more agile to respond Figure 9 / SME Z-Score Distribution the current situation, with more proportion located well by Company Size – Baseline above the investment score, compared to the large Baseline Scenario: SME Z-score Distribution companies. However, larger companies have less Small Medium Large proportion that rated down below 100, which is the score range indicates the high possibility of default of 0-50 12% 1% 5% the rated companies, compared to small companies. 50-100 2% 1% 0% All companies have higher presence among the 100-150 8% 7% 9% medium rating level between 150 and 300. 150-200 16% 18% 23% 200-250 15% 26% 41% 250-300 11% 9% 14% Small Medium Large 300-350 10% 7% 5% 45% 350-400 6% 7% 0% 400-450 6% 12% 5% 40% 450-500 6% 5% 0% 35% % of companies within category 500-550 5% 5% 0% 550-600 3% 2% 0% 30% Total 100% 100% 100% 25% 20% 15% 10% 5% 0% 0-50 50-100 100-150 150-200 200-250 250-300 300-350 350-400 400-450 450-500 500-550 550-600 SME Z-Score 11
Results Sample Distribution Comparisons by Company Size Figure 10 / SME Z-Score Distribution by Company Size – Mild Stress Under the mild stress, the companies move into Mild Stress Scenario: SME Z-score Distribution higher risk areas, and large companies show the most Small Medium Large pronounced shift, as noted before. However, there is 0-50 18% 7% 5% also a relatively high spike in the highest risk band from 50-100 11% 9% 14% small companies. 100-150 8% 11% 18% Small Medium Large 150-200 19% 25% 41% 200-250 24% 33% 18% 45% 250-300 14% 12% 5% 40% 300-350 7% 4% 0% Total 100% 100% 100% 35% % of companies within category 30% 25% 20% 15% 10% 5% 0% 0-50 50-100 100-150 150-200 200-250 250-300 300-350 350-400 400-450 450-500 500-550 550-600 SME Z-Score Figure 11 / SME Z-Score Distribution by Company Size – Severe Stress Under severe scenario, the companies of all sizes Severe Stress Scenario: SME Z-score Distribution move even further into the highest risk area, with Small Medium Large larger companies appear to be hit hardest, and 0-50 36% 32% 41% smaller companies’ ratings also dramatically plunge. 50-100 17% 20% 23% 100-150 22% 26% 18% Small Medium Large 150-200 22% 23% 18% 45% 200-250 2% 0% 0% 40% 250-300 0% 0% 0% Total 100% 100% 100% % of companies within category 35% 30% 25% 20% 15% 10% 5% 0% 0-50 50-100 100-150 150-200 200-250 250-300 300-350 350-400 400-450 450-500 500-550 550-600 SME Z-Score 12
Results Sample Distribution Comparisons by Company Size Looking at the companies’ PD distributions under stressed scenarios, large companies seem to be more sensitive to the shock caused by Covid-19. As shown in Figure 12, under baseline scenario, companies of all sizes are predominantly in the low risk segment, with PD under 10%. In mild-stress scenario, the companies shift into higher risk segments. Large companies show the most pronounced shift into 10% -20% risk segment, whilst small and medium companies are more resilient and are spread across the risk levels. Under severe stress all companies migrate to the highest risk segment with the PD above 30%, with large companies demonstrating the highest presence in this band (68.18%). No companies remain in the lowest risk segment. Figure 12 / Probability of Default Distribution by Company Size Small Medium Large PD: 0% – 10% PD: 10% – 20% PD: 20% – 30% PD > 30% 70% 65.22% % of companies within category 60% 57.45% 59.09% Baseline: PD Distribution by Company Size 50% 40% 30% 26.09% 27.27% 20.77% 20% 14.33% 9.09% 10% 7.45% 6.52% 4.55% 2.17% 0% PD: 0% – 10% PD: 10% – 20% PD: 20% – 30% PD > 30% 70% % of companies within category 60% Mild Stress: PD Distribution by Company Size 50.00% 50% 40.22% 40% 33.86% 32.61% 31.49% 30% 28.56% 27.27% 20% 16.30% 13.64% 10.87% 9.09% 10% 6.09% 0% PD: 0% – 10% PD: 10% – 20% PD: 20% – 30% PD > 30% 70% 68.18% 60.16% % of companies within category 60% Severe Stress: PD Distribution by Company Size 58.70% 50% 40% 30% 21.78% 20.65% 20.65% 20% 18.18% 17.61% 13.64% 10% 0.45% 0% 13
Results Sample Distribution Comparisons by Company Size BRE distribution shows similar patterns in Figure 13: Figure 13 / BRE Distribution by Company Size BBB+ Baseline: Small BRE Baseline: Medium BRE Baseline: Large BRE BBB BBB+ CC- CC- BBB- CC- CC 1.09% BBB+ BBB 11.74% 7.34% CCC- 1.09% 4.35% 4.55% 4.55% B+ 10.87% 4.55% BB+ CC BBB CCC- 1.13% BBB 13.64 B 9.71% 15.22% 4.55% CC+ BB 2.14% BB- CCC- BBB- 13.77% 8.13% B+ CCC B- BB+ 17.39% BBB- 18.18% 0.90% 9.78% B BB CCC 2.93% BB+ 1.09% 18.18% B- BB- BB 2.6% B+ 1.09% CCC+ B+ CCC 3.27% 3.26% BB- 11.96% 1.09% B CCC+ B CCC 6.66% 14.13% 5.43% CCC- CCC+ B- B- 8.47% 9.26% 14.13% CCC+ 31.82% CC+ CC CC- Mild Stress: Small BRE Mild Stress: Medium BRE Mild Stress: Large BRE BB BB- 1.09%- CC- 0.34% B+ CC- B+ B- 3.72% 6.52% 4.55% 4.55% CC 3.26% B CC CC- B 18.96% 6.52% 5.43% 9.09% CCC+ 10.16% 13.64% CC+ 7.61% B- B- 15.22% 9.93% CC+ 13.64% CC 9.48% CCC- 11.96% CC+ CCC+ 5.08% 13.43% CCC+ CCC- 18.48% 13.64% CCC CCC- 7.56% 40.91% CCC CCC 23.91% 21.33% Severe Stress: Small BRE Severe Stress: Medium BRE Severe Stress: Large BRE B- 0.11% CCC+ CCC CCC 0.79% 6.52% 4.55% CCC 12.19% CC- CC- 32.61% 36.79% CCC- 27.27% CCC- 27.17% CC- CCC- 40.91% 22.46% CC+ 4.55% CC+ CC 17.39% CC+ CC 11.96% 16.30% CC 15.69% 22.73% 14
Results Sample Distribution Comparisons by Age of Company Company’s age distribution in the sample is shown in Figure 14 / Company Age Overview Figure 14, where we divided the companies into 4 main age groups: young – companies who are less than or 30 (“[20,30]”), and mature – those that have been active 10.50% over 30 years (“>30”). In terms of the company age, the young companies dominate the overall sample, with nearly half of the sample being in this group, they are followed by companies with ages falling into [11-20], and [21-30] age brackets, 30%). 15
Results Sample Distribution Comparisons by Age of Company Figure 15 / SME Z-Score Distribution by Company Age under 3 Scenarios 30 Baseline Scenario 30% 28.97% 27.88% % of companies within Age Group 24.76% 24.05% 25% 22.78% 20.00% 19.05% 20% 20.25% 15.76% 16.46% 16.46% 18.10% 16.46% 15% 14.33% 15.15% 10.59% 9.66% 10% 8.89% 9.52% 8.57% 4.44% 5% 3.80% 0% 100 200 300 400 500 600 Mild Stress Scenario 50% 46.84% % of companies within Age Group 40% 38.01% 34.95% 31.46% 33.74% 30% 23.05% 25.32% 20% 21.90% 13.92% 13.92% 10% 7.48% 4.04% 0% 100 200 300 400 Severe Stress Scenario 70% 38.01% 34.95% % of companies within Age Group 60% 38.01% 23.05% 50% 21.90% 33.74% 40% 13.92% 33.74% 30% 20% 10% 4.04% 0% 100 200 300 Notes: In SME Z-Score graphs in this analysis, x-axis represents score band. For example, 100 at x-axis means SME Z-score ranging from 0 to 100; 200 represents SME Z-score between 101 and 200. The y-axis represents the proportion of company within respective age group falling in the corresponding SME Z-score range. For instance, under baseline scenario, 3.8% on the dark orange line represents only 3.8% of “old companies” (greater than 30 years) assigned with “
Results Sample Distribution Comparisons by Age of Company Figure 16 / PD Distribution by Company Age 30 PD: 0% – 10% PD: 10% – 20% PD: 20% – 30% PD > 30% 80% 75.95% % of companies within age group 62.93% Baseline Scenario 61.90% 60% 51.52% 40% 22.83% 23.81% 19.63% 20% 16.46% 16.36% 9.29% 10.90% 10.48% 6.54% 3.81% 3.80% 3.80% 0% 30 30 30 30 PD: 0% – 10% PD: 10% – 20% PD: 20% – 30% PD > 30% 80% % of companies within age group Mild Stress Scenario 60% 46.84% 45.71% 40% 37.07% 36.97% 35.44% 32.40% 27.47% 27.47% 24.92% 24.76% 24.76% 20% 13.92% 8.08% 5.61% 4.76% 3.80% 0% 30 30 30 30 PD: 0% – 10% PD: 10% – 20% PD: 20% – 30% PD > 30% 80% % of companies within age group 53.58% Severe Stress Scenario 60% 55.24% 68.08% 44.30% 39.24% 40% 26.79% 23.81% 20.95% 19.31% 20% 15.56% 15.76% 16.46% 0.61% 0.31% 0% 30 30 30 30 17
Results Sample Distribution Comparisons by Age of Company Figure 17 / BRE Distribution by Company Age BBB+ Baseline: BRE with Age 30 CC- CCC+ CC+ 2.53% CC- 1.27% B+ 1.27% CCC- BBB+ 8.86% 10.13% 6.33% 17.72% CC 3.80% CC- CCC CC+ 21.52% 21.52% CCC 2.53% 12.66% B CCC- 16.46% 8.86% BBB 13.92% CC 11.39% CCC+ 12.66% B- 15.19% CCC CCC- B- BBB- 21.52% 29.11% 6.33% 16.46% CC+ B 15.19% 5.06% BB CCC+ 2.53% BB+ 12.66% B+ 1.27% 1.27% 18
Results Sample Distribution Comparisons by Company’s Profitability Profitability is one the major factor determining company’s Figure 18 / Company Profitability Overview creditworthiness. In this analysis, we use Return on Assets (ROA), that is, Net Income divided by Total Low Assets, as the profitability proxy. And the ROA ratios Profitability are categorized into 3 groups: (1) low profitability group, 23.80% having ROA < 0%, (2) medium profitability group, with ROA between 0 and 5%, (3) high profitability group, with more than 5% ROA. Figure 18 provides an overview of company profitability level under the baseline scenario. The low profitability group, medium profitability and high profitability group accounts for 23.8%, 12.9%, and 63.3% of the entire sample, respectively. Medium Profitability 23.80% High Profitability 63.30% Figure 19 (overleaf) shows SME Z-Score Distribution by Profitability (ROA) under 3 Scenarios. Even under the baseline scenario, over half of low profitable companies have less than 100 SME Z-score assigned, which signals the riskier credit profile. For the rest of two groups, the proportion of having less than 100 SME Z-score is around zero, and the high profitability group is likely to achieve higher score. As situation get worse, all ROA groups’ credit score suffers and slide to lower score levels, having greater portion of falling into score band that is less than 100. In the mild-stress situation, none of them obtains score greater than 400, and the maximum score further drops to 300 in the severe stress scenario. Regardless of scenarios, high profitability group generally has better score than other groups. 19
Results Sample Distribution Comparisons by Company’s Profitability Figure 19 / SME Z-Score Distribution by Profitability (ROA) under 3 Scenarios Low Medium High Baseline Scenario 70% % of companies within Profitability Group 61.24% 60% 51.26% 50% 40% 37.21% 28.57% 27.65% 30% 24.80% 19.43% 20% 19.33% 15.01% 12.64% 10% 0.47% 0.42% 1.55% 0.00% 0.00% 0% 100 200 300 400 500 600 Mild Stress Scenario 80% 73.95% % of companies within Profitability Group 56.67% 60% 40.19% 39.25% 40% 28.65% 20.56% 20% 18.07% 11.31% 7.98% 3.47% 0.00% 0% 100 200 300 400 Severe Stress Scenario % of companies within Profitability Group 90% 91.60% 78.53% 80% 70% 60% 63.97% 50% 34.31% 40% 30% 20% 18.93% 7.56% 10% 0.84% 0% 100 200 300 Notes: In SME Z-Score graphs in this analysis, x-axis represents score band. For example, 100 at x-axis means SME Z-score ranging from 0 to 100; 200 represents SME Z-score between 101 and 200. The y-axis represents the proportion of company within respective profitability group falling in the corresponding SME Z-score range. 20
Results Sample Distribution Comparisons by Company’s Profitability Looking at PD bands (Figure 20) with different levels of profitability clusters, we can see that the higher profitability is associated with lower default risk. The low profit cluster is more likely to default in baseline case and becomes even more vulnerable in the stressed situations. The findings are consistent with BRE distribution by profitability (Figure 21): under the baseline scenario, companies more capable of generating higher return have better rating assigned (around 75% of profitable group achieving BRE rating of B- and above), while companies in the low profitability group have generally poor rating with 75% within the group assigned CCC and worse ratings. The divergences of BRE among 3 profitability clusters also exist in mild-stress and severe- stress scenarios. The rating deterioration in the higher profitability group is less severe than lower profit group. Figure 20 / PD Distribution by Profitability Level Low Medium High PD: 0% – 10% PD: 10% – 20% PD: 20% – 30% PD > 30% 90% 80.41% 80% % of companies within Profitability Group Baseline Scenario 70% 60% 52.94% 50% 45.74% 40% 26.09% 27.91% 30% 26.36% 19.33% 16.39% 17.22% 14.33% 20% 11.34% 10% 6.52% 1.74% 0.63% 0% 0.05 0.05 0.05 0.05 PD: 0% – 10% PD: 10% – 20% PD: 20% – 30% PD > 30% 90% 80% Mild Stress Scenario 76.47% % of companies within Profitability Group 70% 60% 55.66% 50% 43.93% 40% 35.98% 35.22% 30% 20% 10.08% 11.21% 8.88% 6.72% 10% 6.72% 4.74% 4.38% 0% 0.05 0.05 0.05 0.05 PD: 0% – 10% PD: 10% – 20% PD: 20% – 30% PD > 30% 100% 92.44% % of companies within Profitability Group 90% Severe Stress Scenario 80% 69.36% 70% 60% 50% 43.79% 40% 30% 27.97% 27.97% 14.33% 20% 13.24% 16.67% 9.09% 6.52% 7.45% 10% 4.62% 2.17% 4.55% 0.74% 0.28% 2.94% 0% 0.05 0.05 0.05 0.05 21
Results Sample Distribution Comparisons by Company’s Profitability Figure 21 / BRE Distribution by Profitability Level BBB+ Baseline BRE with High Profitability Baseline BRE with Medium Profitability Baseline BRE with Low Profitability BBB CC- 0.32% B+ BBB- CC+ B+ B BBB 0.42% B CCC- 0.32% BBB+ BB- 0.78% 1.55% 0.42% 6.00% 10.90% 0.78% B- 3.36% B- BB+ 8.53% 6.30% CCC 11.06% CCC+ BB 7.56% BBB BB- 15.80% CCC+ CCC+ B+ 8.37% 18.60% CC- CCC CCC- 43.70% 11.76% B 48.06% B- CCC+ B- 11.53% BBB- 12.80% CCC CCC- BB+ CCC 14.71% CCC- B BB 1.42% 21.71% 8.69% B+ BB- 4.27% CC CC+ 4.90% 3.63% 4.62% 7.14% CC+ CC Mild Stress BRE with High Profitability Mild Stress BRE with Medium Profitability Mild Stress with Low Profitability CC- CC- 2.01% CC BB- CC- B- B- CC+ 0.91% 0.73% B+ 14.95% 11.21% CCC+ 11.21% CCC+ 3.47% 6.57% 0.93% CCC- 13.71% CCC 6.39% 11.84% B 17.34% CCC- CC 7.79% 28.35% CCC CC+ 26.09% CCC 14.33% 26.48% CC 13.40% B- CC- 16.61% 31.15% CC+ 4.05% CCC+ CCC- 19.89% 6.54% Severe Stress BRE with High Profitability Severe Stress BRE with Medium Profitability Severe Stress with Low Profitability B- 4.76% CCC+ CCC CC- CCC+ CCC- 0.98% CCC 0.84% 5.88% 7.62% 8.57% CC+ CC 6.37% 2.10% 1.90% CCC CC 13.33% 6.30% CCC- 17.65% CC- 36.76% CC+ 0.95% CC+ 12.25% CC- CCC- 84.87% 11.43% CC 25.98% 22
Results Sample Distribution Comparisons by Company’s Leverage Leverage level is an important factor in predicting Figure 22 / Company’s Leverage Overview the resilience of companies to economic shocks. We categorized leverage ratios (Total Debt divided by High Shareholder Equity) into 3 clusters: (1) low leverage Leverage group, with leverage ratio less than 1; (2) medium 12.50% leverage group, with their leverage ratio higher than 1; (3) high leverage group with negative leverage ratio, which results from negative total shareholder’s equity in that sample. It means that liabilities exceed assets which Medium Leverage could happen when company has accumulated losses 18.10% over several periods, and company chooses to borrow money to cover accumulated losses. Figure 22 provides an overview of company’s leverage level under the baseline scenario. Low leverage group, Medium Low leverage and High leverage group accounts for 69.4%, Leverage 18.1%, and 12.5% of the entire sample, respectively. 69.40% Figure 23 (overleaf) shows SME Z-Score Distribution by Leverage under 3 scenarios. Starting from the baseline scenario, more than half of high leverage companies have less than 100 SME Z-score assigned. For the low- and medium leverage groups, the proportion of having less than 100 SME Z-score is relatively low. These suggest that the higher leverage level is associated with greater default risk. Comparing low leverage group and medium leverage one, the latter has generally lower portion of company that fall into distress zone (SME Z-score from 0 to 100, see risk zone mapping in Figure 2) in all 3 scenarios. Low leverage group achieve average higher score in the baseline scenario by having over 50% of companies assessed as medium- and low- risk entities, while medium leverage group turns to slightly better position in severe stress scenario, having higher SME Z-score on average. 23
Results Sample Distribution Comparisons by Company’s Leverage Figure 23 / SME Z-Score Distribution by Leverage under 3 Scenarios Low Medium High Baseline Scenario 60% % of companies within Leverage Group 52.49% 56.80% 50% 38.67% 40% 29.60% 30% 22.91% 21.33% 20% 17.44% 19.45% 12.00% 7.49% 11.38% 10% 6.08% 1.66% 0.80% 1.10% 0.80% 0.80% 0.00% 0% 100 200 300 400 500 600 Mild Stress Scenario 80% % of companies within Leverage Group 78.74% 60% 43.52% 42.86% 40% 39.02% 24.06% 23.89% 20% 16.54% 8.53% 13.94% 4.72% 4.18% 0.00% 0% 100 200 300 400 Severe Stress Scenario 100% % of companies within Leverage Group 96.12% 80% 58.81% 54.10% 60% 44.03% 40% 39.40% 20% 2.33% 1.55% 0% 100 200 300 Notes: In SME Z-Score graphs in this analysis, x-axis represents score band. For example, 100 at x-axis means SME Z-score ranging from 0 to 100; 200 represents SME Z-score between 101 and 200. The y-axis represents the proportion of company within respective leverage group falling in the corresponding SME Z-score range. 24
Results Sample Distribution Comparisons by Company’s Leverage Negative leverage ratio usually signals potential financial distress. According to PD distribution by different leverage levels (see Figure 24), over 60% of high leverage companies are more likely to suffer financial distress (PD >30% – high risk band) and thereby default, compared to other two leverage groups. This portion surges above 80%, and even 95%, in mild and severe downturn respectively. On the contrary, the majority of less leveraged companies have lower PD and stay in the “low risk zone” under the baseline scenario. Medium leverage company group is roughly spread around the “low-and-medium risk zone” according to their PD distribution. Figure 24 / PD Distribution by Leverage Level Low Medium High PD: 0% – 10% PD: 10% – 20% PD: 20% – 30% PD > 30% 90% 80% Baseline Scenario % of companies within Leverage Group 69.60% 70% 60.00% 60% 46.96% 50% 42.54% 40% 35.98% 30% 18.40% 20% 16.43% 11.20% 10.08% 9.39% 10.40% 8.88% 6.72% 7.64% 10% 6.34% 4.74% 4.38% 1.10% 0% Low Medium High Low Medium High Low Medium High Low Medium High PD: 0% – 10% PD: 10% – 20% PD: 20% – 30% PD > 30% 90% 81.89% 80% Mild Stress Scenario % of companies within Leverage Group 70% 60% 50% 48.43% 44.20% 40% 35.98% 30% 26.48% 24.57% 25.60% 20% 16.03% 8.88% 8.66% 9.06% 6.72% 10% 3.94% 5.63% 5.51% 4.38% 0% Low Medium High Low Medium High Low Medium High Low Medium High PD: 0% – 10% PD: 10% – 20% PD: 20% – 30% PD > 30% 100% 96.12% Severe Stress Scenario % of companies within Leverage Group 80% 59.33% 60% 47.76% 40% 25.6% 26.87% 25.37% 23.13% 16.03% 16.98% 20% 9.06% 5.63% 4.38% 0.56% 0.78% 1.55% 1.55% 0% Low Medium High Low Medium High Low Medium High Low Medium High 25
Results Sample Distribution Comparisons by Company’s Leverage These insights are also supported by BRE distributions segmented by the leverage level (Figure 25). Under the baseline scenario, companies bearing less debt are as- signed to better rating (around 60% of them achieving B- and above rating), while high leveraged companies have worse ratings with only around 30% of them obtaining B- and above rating. The portion of having B- and above rating drops to 5% in the medium leverage group. As more severe stress factors applied, the sample’s overall BRE turns worse regardless of leverage level, and the rating of high leverage company deteriorates the most. Figure 25 / BRE Distribution by Leverage Level BBB+ Baseline BRE with Low Leverage Baseline BRE with Medium Leverage Baseline BRE with High Leverage BBB BBB BBB- BB+ 1.10% 1.10% 0.55% BBB- BBB- CC+ 0.80% B CC CC- 0.55% CC BB BB- B+ BBB+ 6.34% BBB+ 0.55% 2.76% 1.10% 0.55% 0.80% B- CC+ 0.86% 9.80% 0.80% 4.80% BB+ CCC- CCC+ 0.58% 17.68% B 4.97% 5.60% CCC- BB 12.39% BBB CCC BB- 14.27% 9.60% B- B+ 18.78% CC- CCC 49.60% B 8.93% B- BBB- 11.24% CCC- 13.60% CCC+ CCC+ CCC 7.06% BB+ 28.73% CCC 1.15% BB B- BB- 3.17% CCC+ CC+ CCC- 8.50% B B+ 3.17% 21.55% 11.20% 4.61% CC CC+ 7.93% 3.20% CC CC- Mild Stress BRE with Low Leverage Mild Stress BRE with Medium Leverage Mild Stress with High Leverage BB- B- 0.68% B+ CC- B+ 1.57% CCC+ CC 3.14% 3.83% B 2.36% CCC CC- 4.27% 12.80% 8.01% 6.27% 4.72% B 13.14% B- CC+ 6.27% CCC- 8.71% 8.66% CC 10.24% CC+ 3.94% CCC+ CC+ 15.68% CC 4.27% B- CCC- 7.09% 14.16% 12.54% CCC- 5.80% CC- 71.65% CCC+ CCC 15.53% 19.11% CCC 35.54% Severe Stress BRE with Low Leverage Severe Stress BRE with Medium Leverage Severe Stress with High Leverage B- CCC+ 0.19% CCC+ 0.30% CCC+ CCC CCC- 0.93% CCC 0.78% 0.78% 1.55% CC+ CCC 1.55% CC 12.31% CC- 14.33% 21.49% 5.43% CC- 33.02% CCC- 24.44% CC 16.12% CCC- 28.96% CC CC+ 18.47% 10.63% CC+ CC- 18.81% 89.82% 26
The speed of the economic recovery Economic crises have often been defined as black Often, we hear reference to the new normal. As for any swans, rare and unpredictable events. In the case of the big change, most people respond with denial and tend current economic crisis, its unpredictability is somewhat to believe that this “new normal” will only be temporary. questionable considering that most of the developed Unfortunately, there is no certainty with regards to the world had been enjoying one of the longest benign length of this new normal stage and waiting for an end periods in the recent financial history. However, the can generate wrong behaviours. People and business trigger of this crisis was definitely unexpected. A global will need to accept this new status quo and adapt. pandemic is a fairly rare event in history and the last time This is the only way to ensure a faster recovery. that something similar was experienced, the world was Individuals can adapt faster than businesses. Many far less connected than it is today. people learnt how to cook or cut their hair, even enjoyed Our way of living gave the virus a great help to spread exercising outside rather than going to gym. Many quickly throughout the world. Governments responded businesses managed to respond quickly. Even during slowly and in a fragmented way. Lockdown measures lockdown, some businesses modified their business were introduced and lifted at different speed creating model to adapt to the new context. These businesses public confusion. These measures significantly altered not only managed to survive the lockdown, but some the habits of the vast majority of the population affecting increased their revenues. every aspect of our life. The speed of the recovery will depend strongly on At the beginning of the pandemic, few understood how these behaviours. The faster businesses will adapt, the long it would be before life returned to normal, and many shorter will be the downturn. SMEs will have an advantage analysts talked of V-shaped recoveries. Many analysts compared to larger, more complex organizations. Small now believe that, barring major improvements in COVID businesses have a leaner structure, lower fixed costs treatment (which would make the disease less dangerous), and faster decision times. These elements will play a only a vaccine can allow economic activity to return to major role in the next months and hopefully provide the pre-pandemic baseline. Even once the economy SMEs with a competitive advantage. starts to reopen, measures will likely be put in place that curtail economic activity to some degree – travel will be less common, businesses will have to space workers and customers further apart, restaurants will be serving fewer customers at a time, and sporting events, concerts, and other activities involving large crowds probably will remain off limits for a long time. And even if the rules allow, many people may be reluctant to return to life as it was before the pandemic. 27
The speed of the economic recovery Most optimistic: The Z The economy suffers a downturn during the pandemic, Figure 26 / Z-Shaped Recovery but then bounces back up above the level it would have been in a pre-pandemic baseline, as pent-up demand Pre-Corona baseline creates a temporary boom. In this scenario, a good part of the GDP foregone during lockdowns – the shopping we didn’t do, the restaurant meals we didn’t enjoy, trips we didn’t take – was simply delayed, and is made up once the risk from the pandemic passes. GDP Time Still very optimistic: The V The economy permanently loses the production that Figure 27 / V-Shaped Recovery would have occurred absent the pandemic, but very quickly returns to its pre-pandemic baseline once social Pre-Corona baseline distancing is lifted. Trips not taken, restaurant meals not purchased, and concerts not attended are forgone, rather than delayed, but once life returns to normal, everything is just as it would have been before. GDP Time 28
The speed of the economic recovery Somewhat pessimistic, and probably more likely: The U The effects of the pandemic on economic activity last Figure 28 / U-Shaped Recovery well beyond the end of the social distancing, and GDP recovers slowly. Even after the health risks recede, the Pre-Corona baseline economy still doesn’t quickly go back to where it would have been, though it does get there eventually. This basic story has many possible shapes. In the U-shape, the level of GDP stays low for a while (perhaps because social distancing norms last a long time), but then recovers back to baseline slowly. GDP Time Another possible estimation: The W If the response to the pandemic is a first round of Figure 29 / W-Shaped Recovery openings that is followed by a surge in COVID-19 cases and another round of closures in the fall, the recovery Pre-Corona baseline could be W-shaped. But then the question will be, what does the recovery from the second (or third, if we do that multiple times) downturn look like? GDP Time 29
The speed of the economic recovery Most pessimistic: The L The pandemic has a permanent effect on GDP. Lost Figure 30 / L-Shaped Recovery investment during the crisis, a rethinking of global value chains, a permanent change to fiscal policy, and a Pre-Corona baseline slowdown in productivity growth all have the potential to cause the trajectory of GDP to be lower than it otherwise would. This is basically what the recovery from the Great Recession looked like. GDP Time 30
Conclusions The report provided forecasts of expected default rates For both stress scenarios we then made further in the next twelve months for the Scottish tourism and conservative adjustments following the feedback from hospitality sector. the tourist industry experts. The analysis used the risk models developed by With stressed financial inputs and macroeconomic Wiserfunding to assess the chances of default/financial variables, forecasts for the average level of default varied distress with financial ratios and macroeconomic between 28% (mild stress) and 43% (severe stress). variables as inputs. The models were applied to a Contrary to our expectations, medium and large sample of the Scottish tourism and hospitality companies companies appeared to be more sensitive to the shock to estimate probability of default (PD) at the company caused by Covid-19. In particular, for large companies level under three scenarios: baseline, mild downturn and the proportion in the highest PD band (over 30%) severe downturn. increased from 4.55% under baseline scenario to The sample of 5000 companies was characterised by 27.27% in mild downturn and to 68.18% under severe predominantly small firms (87% of sample had total assets stress. For small companies that are riskier in normal less than £2m), relatively young (82% of the firms in the circumstances, the PD levels also increased but with sample was less than 20 years old). The latest financial the less pronounced magnitude. This can be attributed statements from 2019 showed a relatively healthy risk to the adaptability of smaller companies that enjoy profile with a good profitability (ROA above 5% in 63% leaner structure and lower amount of tangible assets of the sample) and a generally low level of debt (debt and fixed costs. As such, they can adjust faster to the to equity ratio lower than 1 in 70% of the sample). challenging conditions. The baseline scenario used 2019 financial ratios and As for the company age, younger businesses appeared the most recent macroeconomic variables. Given the to be more vulnerable compared to the more established deterioration in the economy and the impact of the ones. The response to the Covid-19 shock varied also lockdown on this sector, the average PD of these firms according to business fundamentals. More profitable increased to 15% even in the baseline scenario, almost companies were less likely to experience default, and the double of the one observed in other sectors in the UK. the same applied to the companies with moderate levels For the mild stress scenario, we examined the last of debt. The highest risk levels were exhibited by young 20 year of financial statements for the Scottish tourism companies with no profit and high levels of debt. and hospitality sectors. The worst distress was observed The analysis in this report did not address explicitly the during the Global Financial Crisis (GFC) Period (2007 – effect of the government support, this will be the subject 2008), and post-GFC recessions followed by the of further research. However, the high expected default European debt crisis (2009-2011). The tourism industry rates confirm that the current government efforts (e.g. also suffered the effects of swine flu in 2009-2010. We VAT discount) to support the sector are going in the right took the observed percentage change of each financial direction. Nevertheless, we would recommend support variable between the peak and through points in the programs to be tailored to the company size to maximise 2008 crisis as the initial estimates of mild downturn for their impact. Business fundamentals should be taken our current event. into account too. Firms that show the highest level of For the severe downturn scenario, for balance sheet adaptability should be rewarded and offered additional financials we used an additional 50% adjustment as support to overcome the crisis, in order to increase the compared to the 2008 crisis. As for Profit & Loss (P&L) chances of success in the deployment of public funds. variables, we took the recent values from the Scottish Finally, the withdrawal of the current borrowing schemes government estimates of GDP, that showed the estimated should be carefully planned in order not to create additional drop for Accommodation & Food industry in April shocks to the companies with high leverage. 2020 of 85%. 31
Overview of risk modelling methodology The results presented above are based on Wiserfunding The Z-score model remained simple, transparent and proprietary risk models, and this section provides a brief consistently accurate for many years, and this is a possible overview of them in the context of the history of credit reason for its popularity in academic and practitioner risk modelling for businesses. Wiserfunding is a fintech research in finance and accounting. This model opened start-up offering innovative solutions for risk assessment. to door to many alternative models and frameworks It provides consultancy services to business lenders, to predict bankruptcy or defaults. The success of the including the alternative/non-bank ones. The core of its original model inspired more research into default risk methodology is based on the famous Z-score model, assessment in many scientific disciplines. In addition to pioneered by Wiserfunding co-founder - Prof Altman finance and accounting scholars, it attracted statisticians (New York University) and one of the most popular and mathematicians, to experiment with better and models in finance. more efficient methods and techniques, especially with the arrival of more data. More on the history of the The original Z-score model was proposed by Altman Z-score can be found in Altman (2018). (1968) and relied on ratios from financial statements to predict business failures. It used a multiple discriminant For many years thereafter, MDA was the most popular analysis technique (MDA) to a matched sample containing statistical technique used in bankruptcy and default 66 manufacturing firms (33 failed and 33 non-failed). prediction studies (to mention a few – Altman et al 1977; From the 22 potentially relevant financial ratios, five were Blum 1974; Edmister 1972; Micha 1984; Lussier 1995; selected into the final Z-score model as achieving the Taffler 1982; Taffler and Tisshaw 1977). However, the best prediction of bankruptcy: Working Capital/Total problem which was often pointed out is that the two Assets, Retained Earnings/Total Assets, EBIT/Total assumptions of MDA do not hold in most prediction Assets, Market Value Equity/BV of Total Debt and Sales/ problems, i.e. 1) the independent variables should be Total Assets. multivariate normally distributed (and often they are not); 2) the variance-covariance matrices should be equal The name was inspired by statistical Z-measures and for the failing and the non-failing groups. This topic was also because it is the last letter in the English alphabet. extensively discussed by Barnes (1982), Karels and The financial ratios were used as inputs into multiple Prakash (1987) and McLeay and Omar (2000). linear discriminant analysis to estimate the bankrupt/ non-bankrupt group and the resulting discriminant Besides, MDA models are less intuitive as compared to Z-score was compared to cut-offs between “Safe,”, regression when it comes to interpretation. In regression “Grey” and “Distress” zones. If the firm had a Z-Score analysis the coefficients can be interpreted as the slopes below 1.8 (Distressed Zone), it was classified as and therefore, indicate the relative importance of the “bankrupt” and did, in fact, go bankrupt within one different independent variables. This is not the case with year. Firms with a Z-score above 2.99 keep trading, MDA. Logistic regression does not have these limitations at least until the end of the study period in 1966. For and was first applied to the bankruptcy prediction bankrupt and non-bankrupt groups, the accuracy by Ohlson (1980). Ohlson analysed a sample of 105 was 100%, whilst there were 3 errors in classification bankrupt firms and 2,058 non-bankrupt firms from the for firms in “Grey” zone. COMPUSTAT database between 1970-1976. Probit ⟶ 32
Overview of risk modelling methodology ⟶ analysis is another popular way to predict bankruptcy In 2016 a fintech start up (Wiserfunding Limited) was (pioneered by Zmijewski 1984) but logistic regression is founded that focuses on developing credit risk models more widely used in this field. for SMEs. It used the Z-Score as the starting point, leveraging on its strengths (i.e. simplicity, transparency The popularity of logistic regression is based on the fact and consistency), but building a modern version that that it fits well the demands of the default/bankruptcy would be specific to the SME market. This is how the prediction problem, where the dependant variable is SME Z-Score was born. binary (default/non-default). The output from logistic regression is a score between zero and one which The SME Z-Score models are segmented by country provides an estimate for the probability of default (PD). and industry to maximize their prediction power, but The estimated coefficients provide the information on their structure is always the same with 3 modules: a the statistical effect of each of the independent variables typical financial module using SME specific financial on the estimated PD. Most of the academic literature in ratios; a corporate governance and qualitative module, default/ bankruptcy prediction used logistic regression which includes important information about the company (to name several - Becchetti and Sierra 2002; Gentry et structure, ownership and managerial skills; and last, al 1985; Lin et al 2012; Mossman et al 1998; Orton et al a macroeconomic module, that allows to put all the 2015; Platt and Platt 1990; Zavgren 1983), and it is also information specific to the company in the economic used in this study. context in which it operates. There are models for all countries in Europe and the development of models for North America and Asia is in progress. Figure 31 / SME Z-score components Option 1 or 1 Option 2 2 AI SME Z-Score Models 3 Manual feed: eg. interim accounts, management Automated sourcing accounts, projections, stressed financials Source: Wiserfunding, Ltd. 33
Appendix 34
Appendix Table 2 / Financial Ratios definitions Financial Ratios Definition Total Shareholder The shareholders’ funds in a company’s balance sheet is the excess of the assets over the Equity liabilities. Alternatively, you could view it as the shareholders’ investment in the company – the share capital plus all the retained profits of the company. Total Shareholder Equity is the sum of all items under shareholder funds including Share capital, Share premiums and retained earnings (Net earnings available for reinvestment in the firm). Total Assets Total Assets is the sum of all current and non-current assets owned by the firm based on the purchase price. Turnover Total earnings of the business from its daily operations from sale of goods and services to customers also to known as Revenue/Sales. Short Term Debt Short-term debt, also called current liabilities, is a firm's financial obligations that are expected to be paid off within a year. Long Term Debt Long Term Debt (non-current liabilities) is the sum of all financial debt owed for a period exceeding 12 months from date of reporting (bank loan, debenture and mortgage loans). Cash ‘Cash’ stands for ‘cash and cash equivalents’ and refers to the item on the balance sheet that reports the value of a company's assets that are in cash or can be converted into cash immediately. Cash equivalents include bank accounts and marketable securities, which are debt securities with maturities of less than 90 days. Working Capital Capital of the business used in the daily operations calculated as the difference between the current assets and the current liabilities. Finance provided to support the short-term assets of the business (stocks and debtors) to the extent that these are not financed by short-term creditors. Tangible Assets A tangible asset is an asset that has a finite monetary value and usually a physical form. Tangible assets can typically be transacted for some monetary value though the liquidity of different markets will vary. Intangible Assets An intangible asset is an asset that is not physical in nature. Goodwill, brand recognition and intellectual property, such as patents, trademarks, and copyrights, are all intangible assets. EBITDA EBITDA, or earnings before interest, taxes, depreciation, and amortization, is a measure of a company's overall financial performance and is used as an alternative to simple earnings or net income in some circumstances. Retained Earnings Retained earnings is the amount of net income left over for the business after it has paid out dividends to its shareholders. It is the accumulated net income subsequent to any withdrawals as dividends at the reported date under the shareholder funds section of the balance sheet. Interest Expenses The interest expenses is the annual accrued amount of interest that the company paid (or sometimes will have to pay) to its creditors. 35
You can also read