Impacts of Euro adoption on SME’s distress rate

(1)

Amsterdam Business School

MSc Business Economics, track Finance

MASTER THESIS

Impacts of Euro adoption on SME’s distress rate.

Author:

Bc. SimonaTothova

Supervisor:

Dr. Stefan Arping

(2)

i

Acknowledgements

Hereby I would like to thank my supervisor, Dr. Stefan Arping for his valuable recommendations and insightful advice throughout the process of writing this thesis. I would also like to thank my family and friends for the moral support over the course of writing.

(3)

Abstract

This thesis is an empirical investigation of the impacts of change in exchange rate regime, e.g. due to entering the Eurozone, on probability of distress in SME segment. These effects are examined by two multi-period logistic models –indirect and direct in which we use binary variable Euro in connection with volatility of exchange rates. The relationships are tested on a sample from period 2003-2010 considering two generally very close countries – Slovakia and the Czech Republic from which Slovakia is Eurozone member since January 2009 as opposed to the Czech Republic. This sample is analyzed in detail with special emphasis on our dependent variable – distress rate. Both models point out that Euro has a positive (decreasing) effect on probability of SME distress but reveal its high dependence on macroeconomic indicators and unstable results during the financial crisis.

Keywords Exchange rate risk, Credit risk, Eurozone,

Distress rate

Author’s e-mail tothova.simona@gmail.com

(4)

iii

List of Tables

Table 1: SME Definition ... 4

Table 2: SMEs Overview ... 4

Table 3: Distress Rate Analysis – Total Sample ... 26

Table 4: Distress Rate Analysis – Restricted Sample... 27

Table 5: Distress Rate Analysis – Size Comparison Source: Author’s calculations ... 29

Table 6: Accounting Ratios Analysis ... 30

Table 7: Indirect Models – Full Sample ... 35

Table 8: Indirect Models – Sub-samples ... 38

Table 9: Direct Model 1... 40

(7)

List of Figures

(8)

Introduction

1

Introduction

European Monetary Union (EMU) and euro currency reflect more than 40 years of economic and monetary integration which was completed with the introduction of Euro currency in January 1999. The most of European Union members adopted Euro in 1999 and hence fixed their currencies. The impacts of this step on economy and overall business sector are often subject of discussions with still open conclusions. Especially, there is a considerable lack of evidence about the impacts of this step on the most important segment of every economy – segment of Small and Medium Enterprises (SMEs). One of the impacts which have not been sufficiently discussed is the effect of adoption of Euro currency, and therefore fixing the exchange rate, on SMEs credit risk.

Generally there is a noticeable shortage of research about SMEs especially concerning the credit risk modeling. It has been proven that firm-specific indicators (accounting as well as non-financial) are crucial for modeling probability of distress in SMEs (Altman et al. (2010)). However, there is still lack of studies concerning macroeconomic indicators and their impacts. One of these macroeconomic indicators which have been proven to be statistically significant (Michela et al. (2014)) is exchange rate volatility which may be a threat for SMEs in the form of exchange rate risk. Movements in exchange rate have a large impact on exporters and importers and those competing with imported goods, causing a threat to cash flows of SME companies. Moreover, exchange rate changes also affect companies with debt denominated in foreign currency since the debt value changes with the exchange rate (Vasary (2008)). However, this threat could be eliminated by fixing the exchange rate what can be done by the EU members particularly by entering the EMU and adopting the Euro currency.

This thesis focuses on exchange rate risk and currency impacts on probability of distress. For this purpose we chose Slovakia and the Czech Republic and their SMEs. These two countries have common history and share similar cultural, economic and financial characteristics. Moreover, they both entered the European Union in 2004, but, in contrast to the Czech Republic, Slovakia adopted Euro currency in January 2009. Michala et al. et al. (2014) imply that entering the Eurozone should eliminate certain risks for the country, one of which is the exchange rate risk. Therefore, as their results suggest, stability

(9)

2

of Czech crown plays an essential role in the solvency of Czech SME segment1. We investigate the impact of Euro adoption on probability of distress on Slovak and Czech SME segment.

Above all, this research is also one of the first comparative credit risk studies of two countries with focus on macroeconomic indicators and currency impacts, contributing to academic literature concerning credit risk modeling of SME segment. To underline more practical contributions, SMEs are usually dependent on external funding – mostly bank loans – therefore predicting financial distress is critical for banks (for internal risk management since they face asymmetric information issues) as well as for regulators (to determine capital requirements).

The thesis is structured as follows. First chapter introduces SME segment, gives overview of main financial risks faced by companies and discusses Eurozone impact on Slovakia and the Czech Republic. Chapter 2 presents literature review. Chapter 3 describes our research design. In chapter 4 we analyze our data sample and explain the variable selection process. Chapter 5 presents results and discussion. The final chapter concludes.

(10)

Introduction to SMEs

3

1. Introduction to SMEs

“Micro, small and medium-sized enterprises (SMEs) are the engine of the European economy. They are an essential source of jobs, create entrepreneurial spirit and innovation in the EU and are thus crucial for fostering competitiveness and employment”

Günter Verheugen, Member of the European Commission responsible for Enterprise and Industry

Statement above shows the importance of SMEs in the European economy. They are not only a key employer and gross value added generator, but also the main innovator and entrepreneurial skills and creativity contributor to global society. This section provides European Commission’s definition of SMEs which will be used in this thesis and the main Czech and Slovak SMEs business statistics and recent trends. In the second part we focus on the main financial risks faced by SMEs and conclude by discussing European Union and its impact on SMEs segment. (EC (2003), EC (2013))

1.1. European Commission’s Definition of SMEs

To help SMEs to improve their consistency and effectiveness and to overcome difficulties in obtaining capital or credit, European Commission introduced revised common European SME definition that entered into force on 1.1. 2005 and will be used in this thesis. The main purpose of this revised definition was to promote the micro enterprises, improve SME’s access to capital, promote innovation and improve access to Research & Development and to take into account different relationships between enterprises. According to this definition the main factors determining whether a company is an SME are number of employees and either turnover or balance sheet total. (EC (2003))

(11)

4

Table 1: SME Definition

Source: European Commission (2003)

In this thesis we will consider all three categories – Micro, Small and Medium-sized enterprise, i.e. companies with less than 250 employees and with maximum turnover of €50 million.

1.2. Overview of recent SMEs business environment in

Slovakia and the Czech Republic

As

Table 2 shows, European SME sector is dominated by micro enterprises and employs more than 65% of employees. Both the Czech Republic and Slovakia are above EU average considering the number of micro enterprises and the share of employees in micro and small-medium size enterprises. However, value added by Czech and Slovak SMEs is under the EU average. The same holds also for micro enterprises in the Czech Republic. They are above EU average in terms of number of enterprises and employees however their value added is smaller compared to Slovakia or European Union average (EU27). (EC (2013))

Table 2: SMEs Overview

Source: European Commission (2013)

Company category Employees Turnover Balance Sheet Total Medium-sized < 250 ≤ € 50 m ≤ € 43 m

Small < 50 ≤ € 10 m ≤ € 10 m

Micro < 10 ≤ € 2 m ≤ € 2 m

Segment Czech

Republic Slovakia EU27

Czech

Micro 95.50% 95.60% 92.10% 30.10% 33.40% 28.70% 19.90% 22.70% 21.10% Small 3.70% 3.60% 6.60% 18.90% 16.60% 20.40% 15.20% 15.30% 18.30% Medium 0.70% 0.60% 1.10% 19.50% 17.30% 17.30% 19.90% 17.90% 18.30% SMEs 99.80% 99.90% 99.80% 68.50% 67.30% 66.50% 54.90% 55.90% 57.60% Large 0.20% 0.10% 0.20% 31.50% 32.70% 33.50% 45.10% 44.10% 42.40% Total 100.00% 100.00% 100.00% 100.00% 100.00% 100.00% 100.00% 100.00% 100.00%

(12)

5

Generally, the SME sector in Slovakia is characterized by high concentration of small and medium enterprises compared to EU 27. Manufacturing sector is considered to be the backbone of Slovak economy since it commands significantly higher share than in the EU both in terms of employment and value added. It also receives substantive foreign direct investments. Important for this thesis are exports and imports since they are directly exposed to exchange rate risk. In this area SMEs contributed to approximately 40% (50%) of all Slovak exports (imports) to the EU and about 19% (16%) outside of the EU in 2010 (EC (2013)).

According to 2013 Small Business Act report, Czech Republic struggles with competitiveness and performance. Moreover, economic situation seems to be unfavorable to SMEs. Alarming is that 50% of the enterprises closed in 2012 were over 15 years’ old and Czech intra-EU exports and imports showed decreasing trend. Considering sectorial distribution, same as in Slovakia, manufacturing is the backbone of economy with employment and value added above EU average and also attracting considerable foreign investment volumes especially in motor vehicle sub-sector. (EC (2013))

1.3. Financial Risks faced by SMEs

Now we look more closely to the risks faced by SME companies. The main goal of every firm is to maximize its profit. However, they face risks from variety of different sources and have to find the right model by which they can reach this objective under condition of risk. Firms deal on daily basis with both financial as well as non-financial risks. Non-financial risks with which firms have to deal are for instance changes in demand for their products, changes in costumers’ expectations, changes in the cost of material, legal aspects, tax systems, entry of new competitors, new regulations, policies or changes in political system (Mejstrik, Pecena and Teply (2008)). However, in this thesis we focus mainly on financial risks, specifically the exchange rate risk in connection with credit risk. The following section provides theoretical introduction and definitions of main financial risk categories: Market risk, Credit risk and Liquidity risk.

1.3.1. Market Risk

Market risk or so called external financial risk is an uncertainty of future earnings resulting from changes in market conditions. In other words it is a risk resulting from changes in prices, exchange rates and interest (Berg and DeMarzo (1962)). These risks arise for every

(13)

6

firm naturally as part of their business operations. They are usually more pronounced in SME sector which very often struggles with proper planning and forecasting. Market risks can be mitigated only partly, usually through different types of hedging strategies. Main external financial risks are defined as follows:

 Commodity Price Risk is the threat that changes in the prices of the raw materials company uses and the goods they produce will negatively affect firm’s profitability. (Mejstrik et al. (2008), Berg and DeMarzo (1962))

 Exchange Rate Risk is defined as the variability of a company’s value due to uncertain fluctuations in the exchange rate. It usually occurs when company is involved in transactions with international customers or suppliers are exposed to exchange rate fluctuations what leads to changes in amount of payables or receivables. Exchange rate risk is hard to measure but usually is expressed through the exposure – the degree to which firm is affected by fluctuations in the rate of exchange. (Mejstrik et al. (2008), Berg and DeMarzo (1962))

 Interest Rate Risk is a concern for companies if interest rates are volatile since an increase of interest rates raises borrowing costs paid by firm and may reduce its profitability. Moreover in the case when company has fixed long-term future liabilities (e.g. capital leases, pension fund liabilities), a decrease in interest rates increases the present value of these liabilities and consequently may lower overall value of the firm. (Mejstrik et al. (2008), Berg and DeMarzo (1962))

1.3.2. Credit Risk

Credit risk (sometimes called “default risk”) is a potential loss arising from the inability of counterparty to meet its obligations. Good credit risk management is crucial for SMEs since they are usually dependent on only a small number of major customers due to limited resources. Therefore a default or a slowdown of a customer can lead to actual credit (or solvency) problems for the company. In this thesis we look on firms as subjects facing they own credit (solvency) problems caused not only by customer default but also by decrease in sales or an increase in costs leading to consumption of equity and consequent loss of solvency. (Mejstrik et al. (2008), Berg and DeMarzo (1962))

(14)

7

1.3.3. Liquidity Risk

Liquidity risk is tightly connected to credit and solvency risks mentioned above and it is mostly the result of other risks. Liquidity is firm’s ability to meet its financial obligations when they are due and cover its expenses. Financial analysts typically use financial ratios to assess whether a firm is sufficiently liquid. Keeping liquidity solid is one of the most challenging tasks for the companies since short term liquidity problems may very easily transfer to already mentioned more severe solvency issues. (Mejstrik et al. (2008), Berg and DeMarzo (1962))

1.4. Eurozone impacts on Slovak and Czech SMEs

The main goal of this thesis is to investigate whether changes in exchange rate regime (entering the Eurozone) impact SME’s probability of distress. For this purpose we chose Slovakia and the Czech Republic as countries of interest. Before proceeding to literature review which will discuss recent empirical findings in this area we firstly look closely to European Monetary Union and process of Euro adoption in the Slovak Republic. Secondly, we discuss its primary impact on the whole economy as well as on the SME sector.

European Monetary Union (EMU) and Euro currency reflect more than 40 years of economic and monetary integration. European Monetary Integration proposed by Maastricht Treaty signed in 1992 was completed with the introduction of Euro currency on January 1, 1999 (EC (2013). To enter EMU, the main condition is to fulfill the Maastricht (convergence) criteria (Visegrad.info (2010)):

 Inflation (HICP) cannot be higher by more than 1.5% than the average of inflation rate in three states with the lowest inflation

 The ratio of the annual government deficit to GDP must not exceed 3%

 The ratio of state debt to GDP must not exceed 60%

 Participation in the ERM II for two consecutive years, national currency cannot be devalued during this period

 Average yields on the 10-year government bonds in the past year must be no more than 2.0% higher, than the un-weighted average of the 10-year government bond yields in the 3 EU member states with the lowest HICP inflation

It is up to each country to choose its own path toward integration and to calculate economic and political consequences of this step. Entering the Eurozone brings both – high

(15)

8

benefits and high costs. One of the main benefits of adopting single currency is the elimination of exchange rate risk towards the EMU members which leads to noticeable increase in trade among the members. Secondly, it eliminates transaction costs arising from currency exchanges and increases competitiveness among countries since buyers notice differences in prices of products more easily (McKinnon (1963)). Moreover, it decreases price volatility and speculator’s power to influence prices (Grubel (1970)). By entering the EMU countries also eliminates possible problems with inflation and gain credibility. On the other hand, country loses its independence of monetary policies and possibility to use the exchange rate as a macroeconomic variable or policy instrument (e.g. option of devaluation of its currency if needed). (Barro and Gordon (1983))

There has been a lot of debate about which exchange rate regime is the most appropriate for developing countries. Nevertheless, Slovakia, unlike the Czech Republic, decided to change its floating exchange rate regime and in January 2009 entered the European Monetary Union. This step was preceded by joining the ERM-II system in December 2005 and meeting the Maastricht criteria in April 2008. Euro-mood in Slovakia was very pro-euro oriented in 2009 since Slovaks believed that entering the Eurozone will protect the country from financial crisis and currency market disturbances. This expectation was partly fulfilled, but Slovakia had to deal with “shopping-tourism”. Moreover, Greek crisis brought up new issues. On the one hand, Slovak exporters benefited from weak euro currency. On the other hand, Slovakia, as a member of the Eurozone, had to participate in very costly rescue packages during Greek crisis. (Visegrad.info (2010))

Considering the impacts of currency change on the business sector, the main direct advantages are again eliminating exchange rate risk and the transaction costs. But of course, extent of this benefit depends on type of the business and industry in which company operates. From this point of view the highest savings are usually in manufacturing industry since it accounts for major part of foreign trade. As we revealed in section 1.2., manufacturing is the backbone of Slovak SME sector, therefore this advantage is really noticeable for the economy. Moreover, it generally gives SMEs new opportunities to develop foreign trade and makes it easier and cheaper to enter new markets. That in turn enhances their future profitability and growth. Besides the increase in foreign trade, entering Eurozone significantly increases foreign direct investments. Furthermore, SMEs also benefit from indirect impacts, for instance decrease in real interest rates to which

(16)

9

introduction of Euro should lead. For SMEs dependent on external funding, lower interest rates may promote new investments and consequent increase in efficiency and productivity. The general disadvantages are costs connected to changing currency regime which every company has to undertake and higher competitiveness with possible shopping around since prices of goods between countries become more comparable. (Arendas (2006))

Situation in the Czech Republic is vastly different. Even though Czech leading economists realize the benefits of Euro for the business sector, the overall mood is mostly against entering the Eurozone. Moreover, Czech economy has experienced several turbulences in past years and had harder time to reach the Maastricht criteria. Even though entering the Eurozone is an often-discussed economic and political question in the Czech Republic. However economists broadly differ in even approximating the earliest horizon of Euro adoption. (Visegrad.info (2010))

This chapter has provided theoretical framework to main issues covered in this thesis. We introduced SMEs sector which is the essential target for this research and showed its importance for the economy. Secondly, we have analyzed main treats for SMEs – market, credit and liquidity risks. From these risks we focused on exchange rate risk which can be eliminated by changing the exchange rate regime from floating to fixed. That can be done for example by joining the Eurozone as the Slovak Republic did in 2009. The last part of this chapter presented main benefits and costs of this step to give a background for the research goal of this thesis. In the next chapter we discuss the ways of measuring these impacts on credit risk (measured by probability of company’s distress) and look more closely to effects of exchange rate exposure on firms.

(17)

10

2. Literature Review

As academic research emphasize, there is a lack of specific literature concerning SMEs credit risk, exchange rate volatility and currency impacts on SMEs. The most of the literature is focused on large corporations. They however differ from SMEs in their credit risk – e.g. SMEs are riskier, but have lower asset correlation (Dietsch and Petey, 2004). Moreover, classic credit risk models (CreditMetrics, Merton’s) used for large corporations are not suitable for SME credit risk modeling since they are based on market value of their assets and SMEs considered are non-listed companies. On the other hand, classifying SMEs as “retail” is also not applicable. That leaves researchers with challenging question on what are the most appropriate methodology and indicators for SME credit risk modeling. This crucial question is still not completely answered and researchers differ in their opinions. They are not consistent in which group of indicators and which research design is the most suitable for estimating probability of distress of SMEs.

In this chapter, we present relevant literature review on credit risk modeling with focus on SMEs as well as literature concerning exchange rate exposure and currency impacts on firms and their credit risk. At the same time, we provide a brief summary of developments in credit risk modeling during past decades which are crucial for our methodological part.

2.1. Developments of Default Prediction Models without

Macroeconomic Indicators

The last few decades produced major advancements in theoretical credit risk models concerning both large corporations as well as SMEs. Until the late 1960s mostly pure ratio analysis had been used to assess the probability of company failure. Beaver (1966) used univariate prediction model to indicate that financial ration analysis may be useful in the prediction of company failure at least 5 years before company failure. The pioneer of more extensive credit risk models is Altman (1968). He successfully connected traditional (univariate) financial ratio analysis with more rigorous statistical technique by applying multiple discriminant statistical methodology on investigating bankruptcy predictions. Altman (1968) argues that classical ratio analysis is incoherent to faulty interpretation. To overcome this issue he used multiple discriminant analysis (MDA) which allows analyzing

(18)

Literature Review

11

combinations of ratios simultaneously instead of sequentially and therefore remove potential ambiguities and misinterpretations. Altman (1968) classified companies as bankrupt or non-bankrupt according to financial ratios - working capital, earnings, sales and value of equity. Thereafter, using MDA Altman (1968) computed a so called “Z-score” – overall index indicating bankruptcy.

Later studies between 1970s to late 1990s were with some departures mostly motivated by Altman (1968) using MDA technique (Deakin, (1972), Tafﬂer and Tisshaw (1977), Blum (1974)) and were conducted almost exclusively on corporate level data. However, as Eisenbeis (1977) pointed out MDA technique causes frequent statistical difficulties when applied on this type of data. The most common problems concern the distribution of the variables, equal versus unequal dispersions, interpretation of the significance of individual variable, dimension reduction or the definition of groups (Eisenbeis (1977)). Therefore researchers started to apply different, more suitable statistical models. Ohlson (1980) took into account the above mentioned issues and was the first who used conditional logit model to estimate default probability, showing promising results since logit model may also capture information availability (timing problem). Comparably Zmijewski (1984) was first to apply probit model to default prediction study. However, as pointed out by Altman and Sabato (2007), logit analyses give generally better results since they better capture the characteristics of the default prediction issues and were mostly used in the studies thereafter.

The first study which took into consideration SMEs was conducted by Edmister (1972). The study analyzed the usefulness of financial ratios for predicting SME failure with positive results. However, there was not much focus on SMEs until new Basel Accords (Basel II). Focus on SMEs was firstly directed to effects of Basel II on SMEs (Berger (2006)), financing SMEs (Berger and Udell (2004)) and SMEs scoring and credit availability (Frame et al. (2005)). Another question tackled by researchers concerns the impact of SMEs on banks. Kalori and Shin (2004) imply that specialization on SME lending has positive effect on bank’s ROA and increases bank’s profitability. On the other hand Altman and Sabato (2007) conclude that SME lending is riskier for bank comparing to large corporate lending.

Altman and Sabato (2007) also revealed substantial differences between SMEs and large corporations and imply that banks should develop credit risk models specifically for SMEs in order to minimize expected losses. To demonstrate this hypothesis Altman and

(19)

12

Sabato (2007) applied both MDA model (Z-score corporate model) and specific SME logit model (both using firm-specific financial data – mostly accounting ratios) on US SMEs. As expected, specific SME logit model by which they estimated one-year SMEs probability of default has 30% higher prediction accuracy than generic corporate model suggesting that bank’s SMEs profitability is higher when their credit risk is modeled separately from corporate segment credit risk. In later study, Altman et al. (2010) applied previous model on UK SMEs sample and confirmed previous findings. Moreover, they expanded the research and were the first who included also non-financial firm-specific indicators to the model whereby they managed to increase its default prediction power by 13%. Moreover, Altman et al. (2010) argue, that this qualitative information are even more valuable for SMEs comparing to large corporations since there is usually scarcity of financial data.

2.2. Developments of Default Prediction Models with

Macroeconomic Indicators

Another literature stream focuses on connection between macroeconomic indicators and credit risk models. Unfortunately, macroeconomic indicators were mostly used on corporate level data, not SMEs data. Altman (1968) included change in gross national product, S&P 500 returns and money supply M1 in his study. Rose et al. (1982) included large sample of macroeconomic indicators e.g. S&P 500 returns, 3-month T-bill rate, the prime interest rate etc. The significance of these indicators was mixed in the early studies. Hol (2007) was first who besides financial ratios included also macroeconomic variables representing business cycles in the model tested on a sample of Norwegian unlisted companies and provided evidence of the significance of most of the indicators.

Research reviewed above clearly demonstrates that firm-specific (accounting and non-financial) as well as macroeconomic indicators are crucial for modelling probability of distress within SMEs. Focusing on exchange rate volatility, Goudie and Meeks (1991) used macro-micro model of failure to assess the response of the potential failure rate to movements in the effective exchange rate. Their research showed that the impact is substantial. Similar results were concluded also by Nam et al. (2008) who used discrete-time duration model incorporating temporal and macroeconomic dependencies. In their model, the most crucial macroeconomic indicators are volatility of foreign exchange rate and change in interest rate.

(20)

Literature Review

13

Literature presented so far shows an overview of credit risk modelling developments during the past decades. These are taken into account in the most recent study conducted by Michala et al. et al. (2014). Authors forecast distress in European SME portfolios for period 2000-2009 using panel structure dataset, Shumway (2001) hazard model and three main groups of indicators – firm-specific, macroeconomic and industrial. Moreover, this study is unique since it is the first to forecast distress not only on one country but on the sample of SME portfolios from 9 EU countries. Working simultaneously with different groups of indicators and multiple countries, our thesis is thus significantly motivated by Michala et al. et al. (2014). Additionally, Michala et al. et al. (2014) tested regional subsamples where one group was composed of non-Euro zone members – the Czech Republic and Poland. When FX rate was included in the model, they found it has positive (increasing) and pronounced relation to distress what was also concluded by Nam et al. (2008). Consequently, currency and volatility of exchange rates seem to be important and significant indicators of SMEs distress rates.

2.3. Exchange rate volatility and currency impacts on

firms

Literature review above provided summary of recent credit risk research with focus on SMEs and among other things also showed the importance of exchange rate as an indicator of company distress. In this section we discuss impacts of exchange rate fluctuations (exchange rate exposure) on firm.

Volatility of exchange rate is one of the main sources of macroeconomic uncertainty and has large effects on firms in open economies. It affects company’s operating cash flow, discount rate, sales and therefore overall firm value (Choi and Prasad (1995). In the earlier studies, Shapiro (1975) followed already shown exchange rate effects on multinational corporations. However, he presented new definition of exposure – economic exposure defined as exchange rate gains and losses. He showed that under this definition (compared to traditional accounting definition of net current assets) country’s economy (inflation, devaluation) has far greater impacts on the firm’s value. Hung (1992) on the other hand focused his research on exchange rate’s impacts on U.S. manufacturing industry after introducing float exchange rates in 1973. The results showed that consequent dollar appreciation has significant direct negative effect on manufacturing profits. Choi and

(21)

14

Prasad (1995) also studied U.S. multinational companies in period after Breton-Wood system collapse. Firstly, they showed that firm value is significantly affected by real as well as nominal exchange rate. Secondly, they concluded that variations in sensitivity of exchange risk of firm are related to firm-specific operational variables.

Whereas most of the literature explores exchange rate impacts on U.S. markets, there are several studies focusing on smaller open economies. They are especially interesting since they usually have higher foreign trading activity (Muller and Verschoor (2005). For example Dominguez and Tesar (2006) researched exposure on listed companies from eight non-US industrialized and emerging countries. Their first main result is that exchange rate exposure really matters also for these countries. Moreover, Dominguez and Tesar (2006) conclude that size of the company matters. According to their results small companies are more likely to be affected by exchange rate changes than medium and large enterprises.

New questions in this area were brought up by the creation of European currency union. The introduction of Euro currency in 1999 provided unique experimental opportunity to investigate exchange rate exposure. As we have already mentioned the main advantage of entering the Eurozone is considered to be a reduction of exchange rate risk since country adopts fixed exchange rate regime. Few authors studied this phenomenon and its impacts on the firms in Eurozone. Bartram and Karolyi (2006) investigated whether exchange rate exposure, market risk and stock return volatility significantly changed around the Euro introduction. Their results suggest that firms in Euro countries experienced a decrease in foreign exchange rate exposure after the introduction of Euro. On the other hand, Hutson and O’Driscoll (2010) looked to exchange rate exposure of Eurozone companies 10 years after the introduction. Findings showed that non-Eurozone companies’ exposure increased more than Eurozone firms and that they also have higher systematic risk than their Eurozone counterparts.

In the last part of this chapter we provide review of studies focused on the firms in individual countries and the impacts Euro has on their exchange rate exposure. French corporations have been studied by Nguyen, Faff and Marshall (2007). According to them entering the Eurozone is in the case of France associated with a decrease in the number of companies that have significant exchange rate exposure as well as the magnitude of exposure. Moreover, Nguyen, Faff and Marshall (2007) found that this reduction leads to less intensive use of foreign currency derivatives (FCD). Since the majority of the

(22)

Literature Review

15

empirical studies researched leading world economies, Koutmos and Knif (2011) decided to focus on Finland as a representative of small open economies. Their results mostly confirm previous studies, however, Koutmos and Knif (2011) noticed that exposure is asymmetric with respect to depreciation and appreciations of currency.

(23)

16

3. Research Design

In the previous chapters we have introduced and discussed main topics of this thesis. Hereafter we present research design used to fulfill our main goals. The first part presents methodological concepts which are applied on our model presented in the second part of this chapter together with hypotheses. In the last part we introduce our dependent and independent variables.

3.1. The Methodology

3.1.1. Hazard Model

Literature review already provided summary of main credit risk modeling studies. As we concluded the most often used are different types of “static” (mostly logistic) models. However, this type of models are not able to capture any dynamics of firm’s financial structure and to incorporate time –varying macroeconomic variables. In order to overcome these problems Shumway (2001) suggested that hazard models are more applicable. Later on, higher effectiveness of hazard models was also proved by Nam et al. (2008) or Michala et al. (2014). To support this choice Shumway (2001) provides three reasons to prefer hazard models for prediction of default. Firstly, hazard model can control for each company’s period at risk. Secondly, it can incorporate explanatory variables that change with time. In other words, in contrast to static type models, we can add macroeconomic variables that are same for all companies at a given point in time to the model. The last proposed advantage is that hazard models consider each firm year observation separately therefore are more efficient in out-of sample forecasts. In line with above mentioned authors, we use discrete-time hazard (duration) model with time-varying variables and macro-economic dependences. As Shumway (2001) proved, this model is equivalent to multi-period logit model since they have the same likelihood function.

The methodology explained in this subsection follows the methodology presented in Michala et al. (2014), Shumway (2001) and Nam et al. (2008). The hazard model belongs to the group of survival models, in which variables are related to the time that passes before some specific event occurs – in this case it is company distress. The time to firm distress is the “survival time” denoted as t. This survival time t is a continues random variable and follows probability density function in the form , where stands

(24)

Research Design

17

for a vector of distress prediction variables (for company i = 1,2,…..N) and β is a vector of parameters. Moreover, it has some cumulative probability density function which is defined as

(1)

Cumulative probability density function is then used to define the probability that company survives until t. This probability is described by the survival function:

(2) The hazard function, which is incorporated in the hazard model, can be measured as the conditional probability of bankruptcy at time t given survival to that time:

(3)

The most often used is Cox’s (1972) semi-parametric proportional hazard model which is expressed as:

(4)

However, as we mentioned this model allows us to incorporate time-varying firm-specific variables, therefore we can re-write it to the form which accounts for time:

(5)

The first part of the equation stands for firm-specific variables where represents covariates composed of financial statements items of each firm i = 1,2,…,N. The second part of the equation is time dependent baseline hazard function. There is several ways how to specify hazard function. For example Shumway (2001) used the natural logarithm of the company’s age. Carling et al. (2007) proxies the baseline hazard by time dummies. Since we are specifically focusing on macroeconomic variables in this thesis, we adopt approach used by Campbell et al. (2008), Nam et al. (2008), Michala et al. (2014) and include macroeconomic variables in the baseline hazard function. However, as Michala et al. (2014) points out, we will have to account for correlation and multicolinearity effects in macroeconomic variables and the time horizon of the data has to be long enough to capture effects of business cycles on the probability of distress.

The parameter estimates of this model are obtained by maximizing following hazard model likelihood function:

(25)

18

(6)

Since Shumway (2001) proved that discrete-time hazard model likelihood function expressed above is equal to multi-period logit model, we can estimate hazard model by using logistic regression. Therefore following Shumway (2001), Nam et al. (2008) and Michala et al. (2014), we can express hazard rate over next year as a logistic distribution given in the form:

(7)

is a binary variable equal to 1 if company is distressed in year t, zero otherwise. The first part of the hazard function, , is a function of firm-specific variables represented by financial ratios and other qualitative indicators, which are known at t-1 (at the end of previous year). The second part is time-dependent baseline hazard function which in this case incorporates macro-economic variables and affects similarly all companies in the economy. Generally, higher value of entails higher distress probability.

3.2. Model and Hypotheses Development

The literature review discussed in previous chapter revealed that exchange rate volatility seems to be significant variable for modeling SME’s distress rate. Secondly, it was shown that entering Eurozone and changing the exchange rate regime to fixed should have positive impact on firms or at least it suggests that firms in Eurozone country should have smaller exchange rate exposure than non-Eurozone countries. However, there are still a lot of unanswered questions in this field. We decided to investigate impacts of Euro on distress rate of country’s SME segment. These effects seem to have considerable importance. However, we have not found sufficient empirical evidence on this issue. We chose the Slovak Republic and the Czech Republic as countries of interest for this research since not only they are comparable, but also because our results could help in their further development (their suitability for this research is more deeply discussed in the next chapter). In order to investigate above mentioned dependencies we developed 3 types of econometric models. The first one studies Euro impacts on probability of distress indirectly using multi-period logistic regression. The second and third model uses difference-in difference research design and investigates the effect directly. Their aim is to support the

(26)

Research Design

19

results from the indirect model and they give us the opportunity to compare the effects using slightly different research techniques.

3.2.1. Indirect Model

Indirect model is composed of two stages in order to separate the currency (Euro) impact on volatility of exchange rate from other impacts.

The First Stage: This stage of the model can be considered to be a subsidiary regression to

the main second stage. In this stage we use classical OLS regression to regress currency variable on volatility of exchange rate:

(8)

Where FX is a volatility of exchange rate changes and EURO is a binary variable equal 1 for the observations with Euro currency (Slovak companies after January 2009). The main purpose of this stage is to distinguish between part of exchange rate volatility that is determined by currency and part which is determined by other factors. By estimating equation (8) we obtain:

(9)

Using equation (8) and (9) we get predicted values of exchange rate changes, as well as difference between predicted and true values, . The two variables will be incorporated into second stage equation as explanatory variables.

The Second Stage: This part is the core of our research. In the second stage we model

probability of distress of the company in our sample using discrete-time parametric hazard model explained in section 3.1.. The data are estimated by logistic regression technique. The main estimated regression equation2 looks as follows:

(10)

Where PD is a probability of distress, binary variable equal to 1 if company is distressed in year t. and are variables obtained from first stage model indirectly

(27)

20

showing impacts of Euro currency on distress probability. and are respectively Firm-Specific and Macroeconomic control variables.

To produce statistically correct results the model needs two adjustments. Firstly, test statistics obtained from logistic regression are incorrect since they ignore the panel structure of the data sample (they assume that number of the firm-years is the number of observations). Following Shumway (2001) and Michala et al. (2014) to correct this problem we have to modify the sample to account for dependences among firm-year observations. To do so we have to correct (cluster) standard errors to account for number of companies in the data sample. Second issue arises when we incorporate variables from the first stage of the model to the second stage. The problem is that we are transferring estimated values which have their own standard errors to the second equation. However, logistic regression is taking them as a constant with variance zero what leads to skewed (incorrect) standard errors and test statistics in the second stage of the model. To correct this problem we can apply delta method which retrieves standard errors to the correct form.

Besides investigating Euro impacts on whole sample we also want to focus on periods before and after year 2009 when the Slovak Republic entered Eurozone and compare the results for Slovak and Czech Republic. To do so, we can either include binary variables “country” and “period” (and their interaction variables) into the regression. Another possibility is to create corresponding subsamples and compare estimated coefficients. Even though, there are eventual statistical differences, the both methods are presumed to yield comparable results. When estimating indirect model we will use corresponding subsamples. The first possibility – including binary variables for “country” and “period” will be used in direct models explained in the following subsection.

3.2.2. Direct Models

In order to examine results obtained by indirect model we extend this research with the direct model. The methodology presented in section 3.1. still holds, however, compared to the previous model we incorporate the so called difference-in-difference technique and test hypotheses by adding binary and interaction variables into the model. We estimate this model using logistic regression. The exact equations which will be separately tested are as follows:

(28)

Research Design

21

(12)

Where is the actual treatment and equal to one for observations which uses euro currency (Slovakia from 2009), is binary variable equal to one if observation is from Slovakia (treatment group), zero if from Czech Republic and is equal to one for observation from year 2009 onwards, zero otherwise.

Equation (11) estimates directly pure Euro impact on probability of distress (actual treatment). As can be noticed is actually equal to therefore it is the interaction between them. The variable is in this equations included between macroeconomic control variables. Equation (12) is slightly different. Here we focus on Euro impacts on probability of distress through exchange rate volatility, similarly as in our indirect model. However, in this case instead of using two-stage model, we use interaction variable and solely . In this set up we can see different impacts these two variables ( and ) have on probability of distress.

3.2.3. Hypotheses

In order to give the necessary framework to our research we develop hypotheses which will be tested by the model and methodology introduced in this chapter. The four main hypotheses of this research are:

1. General Hypothesis (Indirect Model): and are significant variables. Where has a positive (decreasing) impact on probability of distress and has negative (increasing) impact on probability of distress.

2. General Hypothesis (Direct Models): Interaction variable between Euro and

exchange rate volatility has positive (decreasing) impact on probability of distress. Exchange rate volatility has negative (increasing) impact on probability of distress and binary variable EURO has positive (decreasing) impact on probability of distress.

3. Country Subsample Hypothesis: SMEs in the Slovak Republic have lower

probability of distress than Czech SMEs.

4. Period Subsample Hypothesis: Probability of distress is higher in period before

(29)

22

These hypotheses will be tested on the data sample analyzed in the next chapter and results will be presented and discussed in Chapter 5.

3.3. Variable selection and definition

In order to estimate our model, we need to properly define and select dependent (indicator of financial distress) and independent variables (predictors). This subsection introduces variables considered for the regressions. The variable selection procedure will be based on the existing literature. We focus mainly on the studies in this area for example from Altman (), Nam et al. (2008), Michala et al. (2014) or Shumway (2001). We will inspect the correlations and choose the best performing and the most commonly used ratios. Secondly we will study significance of these ratios in the model. However, the detailed analysis and selection will be presented in the next chapters.

3.3.1. Indicator of Financial Distress

Defining distress in SME segment is more problematic since we have to distinguish between actual distress and closure of the company. For this purpose we adopt distress indicator defined by Michala et al. (2014). We construct indicator, which is equal to 1 if firm is considered “distressed” at given year and 0 if it is considered “healthy”. We consider firm to be distressed if it has negative equity (or is marked in database as defaulted/bankrupt/in liquidation) and in next year it leaves our data sample.

3.3.2. Exchange rate volatility

Furthermore, we have to define our first stage main variable – volatility of the exchange rate (FX). Following Michala et al. et al. (2014) and Nam et al. (2008) we use US Dollar as a comparing currency to which we will truck depreciations and appreciate of local currencies. We calculate this variable as volatility of the daily exchange rate change of the USD/CZK in case of Czech Republic, USD/SKK in case of Slovak Republic before 2009 and USD/EUR for Slovak Republic after 2009. The daily changes data are in the percentage form.

3.3.3. Firm-Specific Variables

We split firm-specific variables into two groups – accounting and quantitative variables. Firm-specific financial group is composed of accounting ration from the categories: ratios for liquidity, profitability, leverage, solvency and activity as proposed by Altman and

(30)

Research Design

23

Sabato (2007). Given our definition of distressed company we omit ratios which involve equity. For firm-specific but non-financial (qualitative) indicators we mostly consider variables tested by Altman et al. (2010) and Michala et al. (2014) – size, legal type, industry sector and region. Full list of the variables is provided in Appendix A – and will be analyzed within next chapters.

3.3.4. Macroeconomic Variables

Macroeconomic variables are especially important in this study. Not only because of the main focus of this research but also to control for differences between Czech and Slovak Republic which could cause bias in our research if not properly treated. Moreover, especially in case of Macroeconomic variables we have to be careful about correlations among variables. The variables which will be considered are financial market, business cycles, lending conditions or variables measuring insolvency. We also inspect variables considering Maastricht criteria to control for differences between Czech and Slovak Republic since as we mentioned in the first chapter there seem to be disproportions. To be concrete, we presume to consider GDP growth, inflation, balance of payments, unemployment, surplus and deficit as percentage of GDP, government debt, industrial production index (for full list see Appendix A – ). However, they will be closely inspected and analyzed in the following chapter.

(31)

24

4. Data Analysis

To create our data sample we used data provided by Amadeus database which includes information about private (non-listed) companies. However, to include macroeconomic indicators we merged our data sample with data from Eurostat and OANDA databases. We adopted SME definition stated in section 1.1. and dropped financial firms from the sample since their structure and practices differ from non-financial companies. Moreover, we applied data cleaning techniques and winsorized accounting variables to prevent outliers. Altogether we have 739,345 observations from which 498,950 are for Czech Republic and 240,395 are for Slovakia for period 2003-2012 (Amadeus database did not provide more recent data).

In order to do our research we firstly have to be sure that chosen countries are comparable enough for the analysis. First part of this chapter therefore deals with macroeconomic comparisons of Czech and Slovak Republic. The next sections then analyze distress rate indicator and focus on selection procedure of firm-specific and macroeconomic variables.

4.1. Czech and Slovak Republic – Comparison Analysis

There are no two countries which are completely identical. However, for purpose of this study, we needed to find two which have similar patterns of development, similar macroeconomic indicators and preferably also similar business environment. Hence, Czech and Slovak Republic seem to be eligible candidates sufficiently satisfying above mentioned requirements.

Czech and Slovak Republic share a long lasting history since they both have been part of Soviet Union until 1989 and coexisted as one nation – Czechoslovakia – until 1993. This fact is also mirrored in their very similar economies and culture. However, after the split their transformation followed different paths. The Czech Republic was more industrial and developed more quickly and smoothly. On the other hand, more agriculturally based Slovakia firstly struggled under policies of left wing government. Nevertheless, after the right wing government was elected in 1998, Slovakia undertook inevitable reforms to liberalize the economy. Generally, Czech economic level was above Slovak after the split in 1993. However, as Figure 1 and Appendix B – shows, the Slovak Republic has been

(32)

Data Analysis

25

considerably fast in developing. Moreover, according to Figure 1 showing real GDP growth rate statistics over time, Slovakia grew mostly faster than the Czech Republic. (Singer (2013))

Nonetheless, Figure 1 and Appendix B present similarities in performance of main macroeconomic indicators of Czech and Slovak economies through period 2003-2012. We can see that these two countries have been developing very similarly over past decades and their macroeconomic indicators follow very similar paths.

Figure 1: Macroeconomic Comparison

Source: EUROSTAT database (2013)

Before the crisis Slovak Republic grew more markedly than Czech Republic. The main reason of that seems to be reforms implemented by right wing government before the crisis. (CNB) However, because Slovak economy is more open than Czech, it is also more vulnerable to external shocks what caused higher decline in 2009. Moreover, at the same time, Slovakia entered the Eurozone and was unable to adjust its exchange rate. On the other hand, the Czech Republic depreciated its currency what partway offset loss of export income (CNB). In the recovery period after 2009, Slovak economy recovered and grew faster than Czech. The main reasons are that Slovakia had a much bigger government borrowings than the Czech Republic and expansionary fiscal policy. (Singer (2013))

-6 -4 -2 0 2 4 6 8 10 12 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012

Real GDP growth rate - 1 year % change

-9 -8 -7 -6 -5 -4 -3 -2 -1 0 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012

General government deficit/surplus in % of GDP

Czech Republic Slovakia 0 20 40 60 80 100 120 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013

Exports of goods and services, % of GDP

0 2 4 6 8 10 12 14 16 18 20 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012

Unemployment rate - annual data , %

Czech Republic Slovakia

(33)

26

In summary, there have been differences in developments of Czech and Slovak economies, but as was said before we cannot find two identical countries. However as graphs and analysis suggest, the differences between these two countries are really small compared to others. Therefore, we can conclude than data samples of Czech and Slovak Republic are applicable for our research and bias accrued from country differences should be minimal.

4.2. Financial Distress Indicator

The most crucial (dependent) variable in our research is financial distress indicator. The indicator is a binary variable which gives zero when company is considered to have no problems in given year and one when company is considered to be distressed. Since we wanted to distinguish between closed (merged) firms and firms which are really distressed in given year we adopted distress definition provided by Michala et al. (2014). Company is marked distressed when it fulfills two conditions. Firstly, variable “legal status” doesn’t say that firm is “Active” or firm has negative equity in the given year. Secondly, the year in which first condition is met is the last year for which we have information about the company (it is not in our sample next year). However, problems arise for the last year for which we have overall data (year 2011 for Slovakia and year 2012 for the Czech Republic) since we cannot determine whether company leaves the sample next year.

Table 3: Distress Rate Analysis – Total Sample

Source: Author’s calculations

Therefore we cannot distinguish whether the company is really distressed. Table 3 shows this paradox. We see that distressed rate for Slovakia in 2011 increased from 3.8% to

Year Healthy Distressed % Distressed Healthy Distressed % Distressed Healthy Distressed % Distressed

2003 19,414 251 1.28% 2,593 3 0.12% 16,821 248 1.45% 2004 32,486 305 0.93% 4,917 30 0.61% 27,569 275 0.99% 2005 51,063 1,008 1.94% 15,428 18 0.12% 35,635 990 2.70% 2006 67,983 488 0.71% 24,376 43 0.18% 43,607 445 1.01% 2007 80,031 1,301 1.60% 27,025 79 0.29% 53,006 1,222 2.25% 2008 77,282 2,715 3.39% 29,802 774 2.53% 47,480 1,941 3.93% 2009 100,168 3,798 3.65% 35,392 709 1.96% 64,776 3,089 4.55% 2010 110,976 6,912 5.86% 45,996 1,841 3.85% 64,980 5,071 7.24% 2011 104,603 20,434 16.34% 37,346 14,023 27.30% 67,257 6,411 8.70% 2012 43,351 14,776 25.42% 43,351 14,776 25.42% Total sample 687,357 68766 9.09% 408,404 21,017 4.89% 818,356 47,749 5.51% Slovak Republic

(34)

Data Analysis

27

27.3% and for Czech Republic in 2012 from 8,7% to 25,4%. We consider these jumps to be a big threat for the accuracy of our research, however there is no more recent data available yet. Moreover, there are no data available for Slovakia in 2012. Therefore we decided that reduction of our sample to only observations between 2003-2010 is inevitable.

Table 4: Distress Rate Analysis – Restricted Sample

Source: Author’s calculations

Considering restricted sample we have 556,181 observations from which 3.02% (16,778

observations) are marked as distressed.

Table 4 shows numbers and percentages of distressed companies in restricted sample sorted by year and country.

We see an increase in distressed companies from 2008 to 2010. This increase could be partly assigned to world financial crisis in these years. Interestingly, we observe higher distress rates in the Czech Republic than in Slovakia over the whole period and especially

after 2009 when the Slovak Republic adopted Euro.

2003 19,414 251 1.28% 2,593 3 0.12% 16,821 248 1.45% 2004 32,486 305 0.93% 4,917 30 0.61% 27,569 275 0.99% 2005 51,063 1,008 1.94% 15,428 18 0.12% 35,635 990 2.70% 2006 67,983 488 0.71% 24,376 43 0.18% 43,607 445 1.01% 2007 80,031 1,301 1.60% 27,025 79 0.29% 53,006 1,222 2.25% 2008 77,282 2,715 3.39% 29,802 774 2.53% 47,480 1,941 3.93% 2009 100,168 3,798 3.65% 35,392 709 1.96% 64,776 3,089 4.55% 2010 110,976 6,912 5.86% 45,996 1,841 3.85% 64,980 5,071 7.24% Restricted sample 539,403 16,778 3.02% 185,529 3,497 1.85% 353,874 13,281 3.62% Slovak Republic

Full Sample Czech Republic

(35)

28

Table 4 shows that between 2008 and 2009, Slovak SME distressed rate decreased from 2.53% to 1.96% while Czech grew from 3.93% to 4.55%. In 2010 when the crisis impacts could be in our sample considered to be the most severe, we see that the distress rate in Slovakia is considerably smaller (although it increased comparing to 2009) than in the Czech Republic. Therefore even these preliminary analyses suggest that adopting euro and fixing exchange rate was an advantageous step for the Slovak Republic.

(36)

Data Analysis

29

Table 5: Distress Rate Analysis – Size Comparison

Source: Author’s calculations

Table 5 focuses on the size of the companies and periods before and after 2009. Size of the companies however, is in this case distinguished only by number of employees. It is noticeable that the highest distressed rate is among Micro SMEs in both countries. Micro SMEs are in Slovakia followed by Small and Medium enterprises. However in Czech Republic, Medium enterprises show higher distressed rate than Small ones in both periods. This occurrence is the most visible in period 2009-2010 when difference between Medium companies’ distressed rate in Slovak and Czech Republic is almost 2.8%. Moreover difference in distressed rates for period before Euro is smaller between Slovak and Czech Republic than for period after Slovakia adopted Euro currency. Therefore after preliminary analysis of dataset we can conclude that Slovakia presumably prevented their enterprises from distress by adopting Euro currency.

4.3. Firm-Specific Variables

There is a wide range of firm-specific variables that can be used to model financial distress. Firstly we focus on firm-specific accounting variables. As was mentioned in literature review section these proved to be significantly important for our model. However, question is which ones to choose. Following authors mentioned in previous sections we considered 18 accounting ratios. In second step we computed correlation

Period Segment Healthy Distressed % Distressed Healthy Distressed % Distressed Healthy Distressed % Distressed 2003-2008 Micro 230,479 5,013 2.13% 64,493 620 0.95% 165,986 4,393 2.58% Small 83,924 911 1.07% 29,996 273 0.90% 53,928 638 1.17% Medium 13,856 144 1.03% 9,652 54 0.56% 4,204 90 2.10% Total 328,259 6,068 1.81% 104,141 947 0.90% 224,118 5,121 2.23% 2009-2010 Micro 163,655 9,109 5.27% 63,117 2,161 3.31% 100,538 6,948 6.46% Small 41,669 1,421 3.30% 14,207 317 2.18% 27,462 1,104 3.86% Medium 5,820 180 3.00% 4,064 72 1.74% 1,756 108 5.79% Total 211,144 10,710 4.83% 81,388 2,550 3.04% 129,756 8,160 5.92% Czech Republic Slovak Republic Full Sample

Impacts of Euro adoption on SME’s distress rate

Amsterdam Business School

MSc Business Economics, track Finance

MASTER THESIS

Impacts of Euro adoption on SME’s distress rate.

Author:

Bc. SimonaTothova

Supervisor:

Dr. Stefan Arping

Acknowledgements

Abstract

Table of Contents

List of Tables

List of Figures

Introduction

1. Introduction to SMEs

1.1. European Commission’s Definition of SMEs

1.2. Overview of recent SMEs business environment in

Slovakia and the Czech Republic

1.3. Financial Risks faced by SMEs

1.3.1. Market Risk

1.3.2. Credit Risk

1.3.3. Liquidity Risk

1.4. Eurozone impacts on Slovak and Czech SMEs

2. Literature Review

2.1.

Developments of Default Prediction Models without

Macroeconomic Indicators

2.2.

Developments of Default Prediction Models with

Macroeconomic Indicators

2.3.

Exchange rate volatility and currency impacts on

firms

3. Research Design

3.1.

The Methodology

3.1.1. Hazard Model

3.2.

Model and Hypotheses Development

3.2.1. Indirect Model

3.2.2. Direct Models

3.2.3. Hypotheses

3.3.

Variable selection and definition

3.3.1. Indicator of Financial Distress

3.3.2. Exchange rate volatility

3.3.3. Firm-Specific Variables

3.3.4. Macroeconomic Variables

4. Data Analysis

4.1.

Czech and Slovak Republic – Comparison Analysis

4.2.

Financial Distress Indicator

4.3.

Firm-Specific Variables