The Impact of the Covid-19 Pandemic Situation on the Perception of

the Type of Risks Albanian SMEs Face

(Case Study – SME’s in the Southern Region of Albania)

LORENC KOCIU

Department of Economic Policy and Tourism,

“Eqrem Çabej” University,

Rruga “Studenti’, Lagjia “18 Shtatori”, Gjirokastra,

ALBANIA

ALBAN KORBI

Department of Finance,

Tirana University,

ALBANIA

Abstract: - Albanian SMEs faced unforeseen difficulties during the period of the COVID-19 pandemic. This

type of situation, unprecedented before, caused many problems in areas such as the management of activities,

liquidity, human capital, sales, marketing, etc. This paper highlights the fact that during this pandemic period,

Albanian SMEs felt the need to understand the most important risks that their activities face under these

circumstances. The empirical study of this paper was based on the data collected to a structured questionnaire,

to identify the perception of the type of risks Albanian SMEs face. Qualitative data collected by SMEs in the

southern region of Albania were used for the successful realization of this study. These data were processed

with the help of the statistical software SPSS v.21, using logistic regression. As a result, it was concluded that

Albanian SMEs have preferences in the perception of some special types of risk that their activities face.

Keyword: - COVID-19 pandemic, Albania, SME, risk, SPSS, logistic regression, perception, Southern region.

Received: October 26, 2022. Revised: November 12, 2022. Accepted: December 2, 2022. Available online: December 30, 2022.

1 Introduction

The COVID-19 pandemic situation that the whole

world faced was an unprecedented situation before.

Albania and all types of businesses that extend their

activity in Albania also faced this situation. This

study is focused only on the businesses classified as

Small and Medium Enterprises (SME). In the

Republic of Albania there is a different definition of

SMEs from that of the European Union. Albania is a

country that lies in the Balkan Peninsula, in its

south-western part and is on the land border with

countries such as Greece, the Republic of North

Macedonia, Kosovo, Montenegro and is separated

by a maritime border with Italy through the Adriatic

Sea.

According to EU recommendation 2003/361 the

factors influencing the definition of SMEs are staff

headcount, either turnover or total balance sheet,

according to table 1.

While in the Republic of Albania the definition

of SME is regulated by Law No. 8957, dated

17.10.2002 "On Small and Medium Enterprises", as

amended.

Table 1. Definition of SMEs in European Union

Company

category

Staff

headcount

Turnover

Total

balance

sheet

Medium-

size

50 to 249

<ALL 250m

<ALL

250m

Small

10 to 49

<ALL 50m

Micro

<10

< ALL 10m

Since 1990, the time when the political and

economic system changed from a communist

country with a centralized economy to a democratic

country with a decentralized market economy. This

change is accompanied by a prolonged period of

political and economic transition, which is also

reflected in the many problems faced by Albanian

SMEs.

During the COVID-19 pandemic situation,

Albanian SMEs faced numerous problems such as

changes in sales volume, sales price, input costs, the

general economic climate, unforeseen changes in

cash flow, budgeting, tax liability and taxes, debtor

and creditor relations with third parties, changing

the structure of human relations within the business,

acceptance of the organization's norms by the

employee, the need for the implementation,

management, maintenance of computer networks

and large-scale use of the Internet, various problems

of personnel recruitment, of matching the

requirements of the workplace with the skills and

abilities of the employee, the significant impacts of

legal and fiscal changes.

The Southern region of Albania is the smallest

region in Albania. It comprises about 3.78% of the

country's population, [1]. The SMEs of this region

make up about 3.5% of the businesses of the whole

country, [2]. As you can see, the southern region of

Albania has a low specific weight in the country's

population and economy, being considered as one of

the regions with low economic development. Due to

these factors, SMEs in this region faced greater

difficulties during the COVID-19 pandemic period

compared to those in the rest of the country.

These many problems that appeared during the

COVID-19 pandemic period brought the need for a

new approach to the perception of new risks faced

by Albanian SMEs. Albanian literature is still poor

in relation to the risks faced by Albanian SMEs and

the process of their identification and evaluation.

This work aims to fill this gap, but it is impossible

to exhaust all possible problems with this work.

As a result, it remains the task of future

researchers to conduct in-depth studies in this

direction to analytically and scientifically identify

the risks faced by Albanian SMEs and the factors

that affect these risks.

It must be admitted that this work, because it

provided data for the period of the COVID-19

pandemic (March 2020 - May 2021 in Albania), had

some limitations, the absence of which would make

possible a more accurate presentation of the results.

These limitations bring an obstacle in the provision

of raw data and their correct processing. Some of

the main limitations are listed below;

 Difficulty in physical proximity between the

interviewer and the interviewee. Due to the

psychological impact caused by the COVID-19

pandemic, fearing people from being infected with

this virus, it was difficult for the interviewer to get

physically close to the interviewee. As a result,

many businesses were reluctant to accept the

interviewer in their premises.

 Lack of regular documentation on the part of

SMEs. A small part of SMEs, from the beginning of

the pandemic until the period when the interview

was carried out, had not kept regular economic

documentation, for various reasons such as the

infection of the economist, the owner or other

employees with COVID, etc.

 In some businesses, the questionnaire was sent

electronically, but due to the lack of quality human

capital, these businesses encountered difficulties in

using information technology and did not

understand how this information should be

completed and sent online.

The rest of the paper is structured as follows;

Section 2 contains the objectives of the study, in

which two research questions and two hypotheses

are raised.

Section 3 presents the literature review by

analyzing how foreigner researchers studied in a

long period of pandemic the performance of SMEs

in their country or in other countries, evidencing

results and their conclusions and comparing them

with other studies.

Section 4 presents the data processing

methodology, the sample selection process of the

observed population, codification process of the

variables taken in consideration.

Section 5 presents data analysis using logistic

regression with the help of statistical software SPSS

v21 and statistical testing of raised assumptions.

WSEAS TRANSACTIONS on BUSINESS and ECONOMICS

DOI: 10.37394/23207.2023.20.30

Lorenc Kociu, Alban Korbi

E-ISSN: 2224-2899

319

Volume 20, 2023

Section 6 presents the conclusions of this study

and further recommendations for future researchers

and decision-making institutions.

2 Objectives of the Study

The main objective of this study is to highlight the

impact of the COVID-19 pandemic situation on

Albanian SMEs in the perception of the most

important types of risks, to which they should pay

great attention.

In the context of this objective, the research

questions are raised:

- Do SMEs attach importance to all types of

risks or do they have preferences for certain

types of risks?

- What are the main directions in which the

risk management faced by SMEs should be

focused?

In focus of the above research questions and the

general objective of this paper, the following

hypothesis has been raised:

H0 - SMEs attach importance to all types of

risks, having no preferences in their

identification and evaluation.

H1 - SMEs do not attach importance to all

types of risks, having preferences in their

identification and evaluation.

3 Literature Review

The businesses included in the SME group operate

in the same environment as large businesses, such as

the economic, political, technological, social and

cultural environment. But they cannot be compared

or compete with large businesses for many reasons

such as insufficient capital, human resources,

technological capacities, etc.

SMEs face a very stiff competition due to

frequent changes in legislation, rapid development

of information and communication technology and

innovation, globalization of markets, movement of

human capital. But, according to [3] [4] regardless

of these obstacles, SMEs are able to survive and

develop due to their agility and adaptability such as

the proximity of their relationship with the

customer, their openness to new ways of doing

business, but many micro, small and medium

businesses are very sensitive to external shocks.

During the COVID-19 pandemic period, all

SMEs in the world encountered extraordinary,

unforeseen difficulties. They faced great difficulties

in liquidity, significant reduction in sales, difficulty

in finding employees, difficulty in supplying the

right quantities and at the right time, etc. More than

half of SMEs have faced severe losses in their

revenues. One third of SMEs fear to be out of

business without further support within 1 month,

and up to 50% within three months.

A recent study on the impact of the pandemic in

Europe, [5], investigates jobs most at risk and finds

that "at least two of three jobs at risk are in an SME,

and more than 30 percent of all jobs at risk are

found within microenterprises consisting of nine

employees or fewer". In Australia, SMEs account

for 68% of all jobs at risk, [6].

By April 2020, 20 million places will be closed

in the USA work, 11 million from to which

belonged to SMEs. In New Zealand during the

months of March - April 2020, a 4% decrease in the

level of employment to small businesses was found.

From a study done from, [7], authors estimated

that from 17 countries to taken in the study it turned

out that SME long period they had COVID-19 a rate

bankruptcy of 12.1% absent to any political

intervention, compared to one rate of 4.5% before

the COVID-19 pandemic situation.

According to [8] from analysis of SMEs in 19

countries in Europe it turned out that those facing

the effects of the COVID-19 pandemic not only

estimated the need for cash and liquidity, but also

estimated their financial capacities to take debt.

According to [9] in the Union European SMEs

cope with more losses than large businesses based

on the percentage in relation to total assets (6-11%

for SMEs, 2-4% for larger firms).

So, as it turns out from a part of conducted studies

during the year 2020, the period in which the crisis

from the COVID-19 pandemic arrived at its peak, it

turns out that SMEs face risks to many, to what

often brought until bankruptcy to a considerable

number of SMEs. SMEs, unlike large businesses

(large enterprises), suffer from numerous financial

problems and lack of human capital. Precisely, due

to numerous financial problems and the lack of

quality human capital, SMEs find it difficult to

apply or use risk management tools, [10]. It is

impossible for SMEs to use the same instruments as

those of large businesses, because they can be too

expensive or too complex, [11].

Risk management is an important process for the

survival of SMEs, but according to [12] they are

suspicious of and sceptical to apply genuine risk

management strategies. According to [13] [14] it is

emphasized that one of the reasons for the failure

and bankruptcy of SMEs is the weak and poor risk

management process, and the lack of planning for

the process of risk identification and assessment.

WSEAS TRANSACTIONS on BUSINESS and ECONOMICS

DOI: 10.37394/23207.2023.20.30

Lorenc Kociu, Alban Korbi

E-ISSN: 2224-2899

320

Volume 20, 2023

It is clear that for European SMEs the literature

is diverse, while for Albanian SMEs this literature is

poor. Especially for the period of the Covid-19

pandemic, for the Albanian reality, this literature is

almost non-existent.

4 Methodology

4.1 Logistic Regression Analysis

According to [15] quantitative and qualitative

methods should be used to guarantee the success of

an article. According to [16] there are many

methods to collect data, such as observations,

experiments, historical data analysis and case

studies. Furthermore, according to [17] it is

emphasized that case studies are valuable when

qualitative data must be provided to achieve the

purpose of the study. In this context, the study is

mainly based on the collection of qualitative data in

order to explain the main problems of this study as

accurately as possible.

The logistic regression model is successfully

used to process the qualitative data, which is

provided through the questionnaire. Logistic

regression is an extension of logical models. Logical

models were applied for the first time by, [18], to

avoid some shortcomings arising from the

application of multivariate discriminant analysis

techniques (Multivariate Discriminant Analysis-

MDA) for predicting the bankruptcy of

organizations. Multivariable discriminant analysis

techniques required restrictive assumptions and

allowed working with disproportionate samples.

Also, in discriminant analysis models, standard

coefficients cannot be interpreted as the slope of the

regression equation, therefore they cannot have an

impact on the relative importance of different

variables. These shortcomings were eliminated with

the use of conditional logic models. Logical models

give a value between zero and one, which

conveniently indicates the probability of the event

occurring. Also, according to logical models, the

estimated coefficients can be interpreted as

important in terms of the level of significance for

each independent variable in the impact they have

on the dependent variable.

Logistic regression is an integral part of the

category of statistical models called General Linear

Models. Logistic regression is used to analyze

problems in which one or more independent

variables intervene that affect the dependent

variable of dichotomous type, in this case the latter

is considered as their dependent case variable, [19].

From a mathematical point of view, logistic

regression differs from linear regression. In linear

regression, the value of the dependent variable is

predicted by one or several predictor variables,

while with logistic regression we get a logistic

equation in which the logarithm of the odds ratio

(odd ratios) is given by a linear dependence of the

independent variables that participate in the process.

The logistic regression equation will be of the

form: ln(p/(1-p) = B0 + B1X1 + B2X2+ ....+ BnXn and

shows the probability of the event occurring under

the influence of all independent variables. One of

the independent elements the most important of the

multiple logistic regression model is the "odds ratio"

or the ratio of chances.

Unlike linear regression, logistic regression

coefficients are interpreted indirectly. Odds ratios

are calculated and show that when an

independent variable changes by one unit and all

other independent variables do not change, the

chance of impact on the dependent variable will

change by a factor of type . So, this factor for the

independent variable indicates the relative amount

by which the chance of the outcome increases (odds

ratio>1) or decreases (odds ratio<1), when the value

of this variable increases by one unit.

There are two cases of hypothesis validation:

1. When there is a null hypothesis (H0),

according to which no predictor variable affects the

dependent variable, and to test this hypothesis if it is

true, all the coefficients of the logistic regression

equation must be equal to zero, βi = 0, otherwise

this hypothesis is rejected or rejected,

2. When there is a non-null hypothesis (H1),

according to which the predictor variables affect the

dependent variable, and for its testing, at least one

of the coefficients of the regression equation must

be different from zero, so    then we say that

the hypothesis we want to verify, it stands or in

other words, it is accepted.

The logistic regression equation will be of the

following form;

 

                

 󰇛󰇜

In addition to the fact that the coefficients of the

logistic regression equation must be different from

zero, the significance levels for each coefficient

must be less than the allowed level of 0.05, which

means that the model is significant at the 95% level

of reliability. The best case would be when the level

of significance for each coefficient would be Sig =

0.000.

WSEAS TRANSACTIONS on BUSINESS and ECONOMICS

DOI: 10.37394/23207.2023.20.30

Lorenc Kociu, Alban Korbi

E-ISSN: 2224-2899

321

Volume 20, 2023

Also, in the focus of hypothesis testing, other

statistical indicators such as -2Log Likelihood, Chi-

square, as well as their level of significance was

analyzed. For this purpose, the complete test of their

relationship was performed in the statistical

program. -2Log Likelihood indicates the error of the

logistic regression model, which is analogous to the

“sum of squares”. When an independent variable

has a relationship with the dependent variable the

value of -2Log Likelihood starts to decrease. The

difference between the value of -2Log Likelihood in

the first step of the regression model with the value

of -2Log Likelihood at the end of the process of the

regression model will evidence the value of the

statistical indicator Chi-square(), for its

significance level (α) and the corresponding degree

of freedom (df). In the case when the level of

significance (Sig), corresponding to the Chi-square

value of the logistic regression model is less than

the value α = 0.05, for the same degree of freedom

(df), then it can be said with complete certainty that

the hypothesis is proven with the logistic regression

model.

4.2 Determining the Selected Sample of

Population

Determining the size of the selected sample is one

of the most important points of the study. The

selected sample must represent the population under

study from the point of view of the characteristics to

be studied. According to [20] the sample is a part of

the population and carefully selected to represent

the entire population and listed three reasons in

support of sampling:

1) Greater accuracy of results

2) Greater speed of data collection

3) Availability of population elements

For this study, the sample was randomly selected

from the population. Random sampling means that

each element of the population has the same chance

of being selected and that the selection of one

element does not affect the chances of another

element being selected.

Table 3 shows the number of SMEs in the South

region, divided according to their activity. By

determining the total number of the population that

will serve as a study subject, it is easy to determine

the sample size.

Table 3. Distribution of SMEs by activity

Total

number

of SMEs

Manufactur

Building

Business

Service

4,124

835

322

1883

1084

100%

20.25%

7.82%

45.67%

26.26%

The exact determination of the sample size was

made by relying on the statistical formula,

simultaneously determining an error interval

between 5% and 10%.

  

󰇛󰇜 (2)

n - indicates the size of the sample to be studied

N - indicates the full size of the population from

which the sample will be selected

e - indicates the margin of error

If we wanted to have a 95% confidence level in

this study, we should have surveyed:

  

󰇛  󰇜

󰇛  󰇜

 󰇛󰇜

In fact, 360 businesses were interviewed, making it

possible to maintain the reliability level of 95%. The

following table shows the distribution of the sample

of the interviewed population, adhering to the

specific weight of each business group to the total

population, table 4.

Table 4. Distribution of the selected sample

according to activities

Total

number

of SMEs

Manufacture

Building

Business

Service

364

166

100%

20.25%

7.82%

45.67%

26.26%

In the southern region, the economic activity is

carried out almost by SMEs, and referring to table 4,

20.25% of the businesses of the SME group carry

out manufacturing activity such as the production of

agricultural and livestock products. 7.82% of SMEs

are included in the construction sector, where it

should be emphasized that during the pandemic

period the construction sector almost did not work.

45.67% of SMEs are involved in trade, where it

should be noted that most of the goods are imported

from neighbouring countries such as Greece, Italy,

Serbia, Kosovo and the Republic of North

Macedonia. Albania is a net importing country,

which means that the value of imports is greater

than the value of exports. 26.26% of SMEs are

included in the service sector. The service sector

also includes companies operating in the tourist

WSEAS TRANSACTIONS on BUSINESS and ECONOMICS

DOI: 10.37394/23207.2023.20.30

Lorenc Kociu, Alban Korbi

E-ISSN: 2224-2899

322

Volume 20, 2023

sector, since the southern region of Albania is one of

the main tourist destinations in Albania.

4.3 Codification of the Variables

Qualitative ordinal type data were processed with

statistical software SPSS v.21, using logistic

regression to identify the chance of influence on the

dependent variable of ordinal data, which are also

independent variables. The independent variables

were evaluated using a 5-level Likert scale, which is

widely used to measure qualitative data. Meanwhile,

the dependent variable is presented as data of a

categorical-dichotomous nature (which only takes

the values NO = 0 and YES = 1). For the processing

of these data with the help of SPSS v.21, it is

necessary to codify them in order to process the data

as accurately and simply as possible. The process of

this codification is the independent and dependent

variables, as follows:

 "PrefRisk" – Dependent variable, which

indicates whether SMEs have preferences or not

regarding the identification and assessment of

risks

 "BiznRisk" - preference for the

identification and assessment of business risk

 "FinancRisk" - preference for identifying

and evaluating financial risk

 "TechnolRisk" - preference for the

identification and assessment of technological risk

 "CapHumRisk" - preference for identifying

and evaluating human capital risk

 "PolitRisk" - preference for identifying and

assessing political risk

It should be said in advance that these types of

risks were selected after a preliminary evaluation

work with SMEs to identify which risks they had

complete or incomplete knowledge of.

5 Data Analysis

The first step of creating a logistic regression model

using the statistical software SPSS is to evaluate the

accuracy of the predictive model with the help of

the "Overall Percentage" parameter. From this

analysis it turned out that the predictive model tends

to be 89.6% correct every time it is used, table 5.

This value allows this model to be used as a

predictor, but it cannot be said with full accuracy

just by analyzing this table how much is the

influence of each predictor variable on the

dependent variable, for the reason that the following

tables must also be analyzed to reach a correct

conclusion.

Table 5. Overall percentage a

Observed

Predicted

PrefRisk

Percentage

Correct

.00

1.00

Step

PrefRisk

.00

1.00

323

100.0

Overall Percentage

89.6

Table 6 summarizes from a statistical point of

view the logistic regression coefficients, or

otherwise called pseudo- . This is due to the fact

that logistic regression does not have the same

regression as least squares (OLS regression). The

Cox Snell-R-Square coefficient tends to

approximate a multiple based on the probability

of the event occurring, the smaller it is, the more

accurate the predictive model is. Since Cox Snell R

Square = 0.351, it means that the change in the

dependent variable is explained by the logistic

regression model to the extent of 35.1%. While the

Nagelkerke R Square coefficient is a more reliable

measure of the relationship between the dependent

variable and the independent variables in the logistic

regression model, compared to Cox Snell R Square,

it is considered as estimated. This coefficient

must always be greater than the value of Cox Snell R

Square. According to the summary table 6,

Nagelkerke R Square = 0.721, which means that the

prediction model is influenced 72.1% by the

independent variables.

Table 6. Model summary

step

-2 Log

likelihood

Cox Snell R

Square

Nagelkerke R

Square

1087.675 a

0. 351

0.721

Table 7 shows the variables included in the

logistic regression. As can be seen, all the

independent variables, "FinancRisk",

"CapHumRisk", "PolitRisk", "BiznRisk" and

"TechnolRisk", which are included in this model

have great statistical significance because their

significance levels are at very low (Sig = 0.000 <

0.05), which means that they refer to the 95%

confidence level.

WSEAS TRANSACTIONS on BUSINESS and ECONOMICS

DOI: 10.37394/23207.2023.20.30

Lorenc Kociu, Alban Korbi

E-ISSN: 2224-2899

323

Volume 20, 2023

Table 7. Variables of regression equation (4).

S.E

Wald

Sig.

Exp(B)

95%CIfor

EXP(B)

Lower

Upper

Step

1 a

FinancRi

2.312

.677

11,680

.001

10,096

2.681

38,019

CapHum

Risk

2.205

.665

10,976

.001

9,068

2.461

33.414

PolitRisk

1,964

.672

8.554

.003

7.127

1.911

26,578

BiznRisk

2,670

.691

14.911

.000

14.439

3.724

55,985

Technol

Risk

2.256

.666

11,486

.001

9,548

2,590

35.206

Constant

31,559

10.02

9.912

.002

.000

Variable(s) entered on step 1: FinanRisk, CapHumRisk,

PolitRisk, BiznRisk, TeknolRisk.

The variables that are included in this table mean

that they are the variables with a significant impact,

which is determined by the relevant logistic

coefficients, on the possibilities that SMEs have to

predict the chance that they have to prefer different

types of risks. These logistic coefficients serve to

build the logistic regression equation (4):

 

     

   

  (4)

To verify the null hypothesis (H0), that no

predictor variable affects the dependent variable, all

the coefficients of the logistic regression equation

must be 0 (   ), which is not true according to

the logistic regression equation (4).

But, in the logistic regression equation, it is seen

that all the coefficients are different from zero

(  ). This means that the alternative (no-null)

hypothesis (H1) is acceptable and statistically

verifiable with a confidence level equal to 95%,

according to which the predictor variables influence

the dependent variable and the influence of each

predictor variable is evidenced by the coefficient 

next to each of these variables.

Referring to equation (4) of the logistic

regression, it is noted that all the coefficients of this

equation are different from zero. Also, the

significance level values are almost at the zero level,

which shows that statistically this prediction model

is significant within the 95% confidence interval.

These mentioned above are summarized in Table 8.

Table 8. Summary of statistical parameters of H1

The

coefficients

Value

Sig

Statistical

Significance

Testing H1

B FinanceRisk

≠ 0

2.312

0.001

Sig<0.05

Accepted

B CapHumRisk

≠ 0

2.205

0.001

Sig<0.05

Accepted

B PolitRisk ≠ 0

1,964

0.003

Sig<0.05

Accepted

B BiznRisk ≠ 0

2,670

0.000

Sig<0.05

Accepted

B TechnolRisk

≠ 0

2.256

0.001

Sig<0.05

Accepted

Also, in the focus of testing the alternative

hypothesis H1, the relationship that exists between

the statistical parameters -2Log Likelihood and Chi-

square (󰇜, as well as their level of significance,

will be analyzed. For this purpose, a complete test

of their relationship has been made in the statistical

program, which is presented in table 9.

Starting the data processing with the value of -

2Log Likelihood = 1129.739 and finishing the data

processing to obtain the most statistically significant

independent variables with the value of -2Log

Likelihood = 1087.675, it results that the value of

Chi-Square (󰇜will be the difference of the value

of -2Log Likelihood at the beginning of the model

with the value of -2Log Likelihood at the last step of

the model, i.e. 1129.739 – 1087.675 = 42.064,

which is also evidenced in table 10. This value of

Chi-square results in a significance level

(Sig=0.000), which is compared to the default level

α = 0.05, in order for the model to be significant

within the 95% confidence interval.

Table 9. Iteration History a,b,c,d

Iteration

-2 Log

likelihoo

The coefficients

Consta

Financ

Risk

CapH

umRi

Polit

Risk

Bizn

Risk

Tech

nolRi

Step

1164.408

-14,280

1.067

1.045

.962

1.199

1.052

1092.693

-25,802

1,889

1,828

1.651

2.161

1,853

1087.740

-30,881

2.260

2.162

1,929

2.608

2.208

1087.675

-31,549

2.311

2.204

1,963

2,669

2.256

1087.675

-31,559

2.312

2.205

1,964

2,670

2.256

1087.675

-31,559

2.312

2.205

1,964

2,670

2.256

a. Method: Enter

b. Constant is included in the model.

c. Initial -2 Log Likelihood: 1129.739

WSEAS TRANSACTIONS on BUSINESS and ECONOMICS

DOI: 10.37394/23207.2023.20.30

Lorenc Kociu, Alban Korbi

E-ISSN: 2224-2899

324

Volume 20, 2023

Table 10. Omnibus Tests of model coefficients

Chi-square

Sig.

Step 1

steppe

42,064

.000

Block

42,064

.000

pattern

42,064

.000

Table 11 shows the distribution of Chi-Square

according to the level of coefficient α and for the

corresponding degree of freedom (df), which in this

case belongs to level 5. From the comparison of the

standard value of Chi-Square = 11.070 in table 11

which belongs level α = 0.05 for the degree of

freedom (df = 5), and the value of Chi-Square =

42.064 in table 10 that belongs to our regression

model for the degree of freedom (df = 5), it turns out

that 42.064>11.070.

Table 11. Distribution of χ2

The level of α

0.5

0.10

0.05

0.02

0.01

0.001

0.455

2.706

3.841

5.412

6.635

10,827

1.386

4.605

5.991

7,824

9.210

13.815

2.366

6.251

7.815

9,837

11.345

16.268

3.357

7,779

9,488

11,668

13.277

18,465

4.351

9.236

11,070

13.388

15,086

20.517

This means that for the Chi-square value of

42.064, the level of significance is less than the

standard level α=0.05, and in fact the level of

significance for this regression model is Sig=0.000.

Therefore, the alternative hypothesis H1 is

acceptable and statistically tested: SMEs do not

attach importance to all types of risks, having

preferences in their identification and evaluation.

While the hypothesis H0 is rejected.

6 Conclusions and Recommendations

At the end of this paper, it can be said that Albanian

SMEs do not appreciate all the risks they face.

However, they have preferences in the perception of

identifying and evaluating some special risks of

their economic activity. This was also verified with

the logistic regression equation (4), according to

which all coefficients of the equation are different

from 0. This means the verification and statistical

testing of hypothesis H1. By analyzing the

independent variables, the regression coefficients

and their odds ratio {Exp(B)} it results that:

 The variable that shows the preference of

SMEs in the identification and assessment of

financial risk, "RiskFinc" has a positive logistic

regression coefficient (BFinancRisk = 2.312) and odds

ratio recorded as Exp(B)=10.096. This high level of

odds ratio means that when SMEs find changes in

financial risk, then this change will affect the chance

they have to prefer the perception of this type of risk

by 10.096 times, when all other variables of forecast

do not change.

 The variable that shows the preference of

SMEs in the identification and assessment of human

capital risk, "CapHumRisk" has a positive logistic

regression coefficient (BCapHumRisk = 2.205) and odds

ratio evidenced as Exp(B)=9.068. This high level of

odds ratio means that when SMEs find changes in

human capital risk, then this change will affect the

chance they have to prefer the perception of this

type of risk by 9.068 times, when all the variables of

others of the forecast do not change .

 The variable that shows the preference of

SMEs in the identification and assessment of

political risk, "PolitRisk" has a positive logistic

regression coefficient (BPolitRisk = 1.964) and odds

ratio evidenced as Exp(B) = 7.127. This high level

of odds ratio means that when SMEs detect changes

in political risk, then this change will affect the

chance they have to prefer the perception of this

type of risk by 7.127 times, when all other variables

of the forecast do not change .

 The variable that shows the preference of

SMEs in the identification and assessment of

business risk, "BiznRisk" has a positive logistic

regression coefficient (BBiznRisk = 2.670) and odds

ratio recorded as Exp(B)=14.439. This high level of

odds ratio means that when SMEs detect changes in

business risk, then this change will affect the chance

they have to prefer the perception of this type of risk

by 14,439 times, when all other variables of the

forecast do not change.

 The variable that shows the preference of

SMEs in the identification and assessment of

technological risk, "TeknolRisk" has a positive

logistic regression coefficient (BTeknolRisk = 2.256)

and odds ratio evidenced as Exp(B)=9.548. This

high level of odds ratio means that when SMEs

detect changes in technological risk, then this

change will affect the chance they have to prefer the

perception of this type of risk by 9.548 times, when

all other variables of the forecast does not change .

At the end of this paper, some important

recommendations emerge:

WSEAS TRANSACTIONS on BUSINESS and ECONOMICS

DOI: 10.37394/23207.2023.20.30

Lorenc Kociu, Alban Korbi

E-ISSN: 2224-2899

325

Volume 20, 2023

 SMEs should pay attention to budgeting and

cash flows, debtor and creditor relationships, tax

obligations and all other financial information,

making it possible to control their financial risk.

 SMEs should pay attention to sales volume,

sales price, input costs, general economic climate,

which affects the best management of business risk.

These elements had a very big impact on SME

during the period of the COVID-19 pandemic.

 SMEs should pay special attention to human

capital, which extends in different directions within

the business, such as recruiting personnel, matching

the requirements of the workplace with the skills

and abilities of the employee, implementing the

code ethics, the ability to cooperate, etc.

 SMEs should pay attention to the

implementation, management, maintenance and

renewal of the technology used. This means

recognizing the need for technology and its cost as

part of the business development strategy. During

the pandemic period, the most successful businesses

were those that used online sales using the Internet,

and those that continued the traditional way of

selling.

 The public sector must be a strategic partner

of SMEs, especially in difficult economic and

financial times such as the COVID-19 pandemic

period. During such periods, the public sector

should strongly support SMEs with complete

financial packages and fiscal ease until they realize

positive cash flows and have financial stability.

References:

[1] INSTAT (Albanian Institute of Statistics),

"Albania in figures 2021"

[2] INSTAT (Albanian Institute of Statistics),

"Business Registers 2021"

[3] Berry, A. "The Role of the Small and Medium

Enterprise Sector in Latin America and

Similar Developing Economies." Seton Hall

Journal of Diplomacy and International

Relations 2002: 107-114

[4] Laforet, Sylvie, and Jennifer Tann.

"Innovative Characteristics of Small

Manufacturing Firms." Journal of Small

Business Enterprise Development, 13(3) ,

2006: 363-380

[5] McKinsey (2020), Safeguarding Europe's

livelihoods McKinsey ,

https://www.mckinsey.com/industries/public-

sector/our-insights/safeguarding-europes-

livelihoods-mitigating-the-employment-

impact-of-covid-19

[6] McKinsey (2020), How to rebuild and

reimagine jobs amid the coronavirus crisis ,

https://www.mckinsey.com/industries/public-

sector/our-insights/how-to-rebuild-and-

reimagine-jobs-amid- the-coronavirus-crisis .

[7] Gourinchas, P. and S. Kalemli-Özcan (2020),

“COVID-19 and business failures”, mimeo,

UC Berkeley.

[8] Bircan, C. et al. (2020), Coronavirus credit

support: Don't let liquidity lifelines become a

golden noose, VOX CEPR Policy Portal,

https://voxeu.org/article/coronavirus-credit-

support-don-t-let-liquidity-lifelines-become-

golden-noose

[9] Revoltella, D et al (2020), EU firms in the

post-COVID-19 environment, VOX, CEPR

Policy Portal, https://voxeu.org/article/eu-

firms-post-covid -19-environment .

[10] Brustbauer,J.(2016). Enterprise risk

management in SMEs: Towards a structural

model. International Small Business Journal,

34(1), 70-85.

[11] Pereira, L., et al. (2015). A risk diagnosing

methodology web-based platform for micro,

small and medium businesses: Remarks and

enhancements. Communications in Computer

and Information Science, 340-356.

[12] Florio, C., & Leoni , G. (2017). Enterprise

risk management and firm performance: The

Italian case. The British Accounting Review,

49(1), 56-74.

[13] Sipa, M. (2018). The factors determining the

creativity of the human capital in the

conditions of sustainable development.

European Journal of Sustainable

Development, 7(2), 1-13. Doi:

10.14207/ejsd.2018.v7n2p1.

[14] Wasiluk, A. (2017). Pro-innovative

Prerequisites for Establishing the Cooperation

between Companies (in the Perspective of

Creation and Development of Clusters). In K.

Halicka & L. Nazarko (ed.), Iwona Gorzeń-

Sustainable Development. 7th International

Conference on Engineering, Project, and

Production Management (Vol. 182, 755–762).

Amsterdam: Elsevier Science Bv .

[15] Leopoulos, Vrassidas N, Kostantinos A

Kirytopoulos, and Charalambos Malandrakis.

"Risk management for SMEs: Tools to use

and how." Production Planning & Control ,

2006: 17(3): 322-332.

WSEAS TRANSACTIONS on BUSINESS and ECONOMICS

DOI: 10.37394/23207.2023.20.30

Lorenc Kociu, Alban Korbi

E-ISSN: 2224-2899

326

Volume 20, 2023

[16] Ghauri, Pervez, and Kjell Gronhaug. Research

Methods in Business Studies (4th Edition).

New Jersey: Prentice Hall, 2010

[17] Berg, Bruce L, and Howard Lune. Qualitative

Research methods for the Social Sciences (8th

Edition). USA: Pearson, 2011

[18] Ohlson, James A. "Financial Ratios and the

Probabilistic Prediction of Bankruptcy."

Journal of Accounting Research, Vol 18, No

1, 1980: 109-131

[19] Mano, Romeo. "Logistic Regression and an

Application in Microfinance." Microthesis.

Tirana: Department of Mathematics, 2005

[20] Blumberg, et al. Business Research Methods.

New York: McGraw-Hill Education, 2005

Contribution of Individual Authors to the

Creation of a Scientific Article (Ghostwriting

Policy)

Lorenc Kociu was responsible for gathering the

quality data from questionnaires and processed them

with statistical software SPSS v.21 and editing the

paper.

Alban Korbi was responsible for literature review,

methodology and supervision.

WSEAS TRANSACTIONS on BUSINESS and ECONOMICS

DOI: 10.37394/23207.2023.20.30

Lorenc Kociu, Alban Korbi

E-ISSN: 2224-2899

327

Volume 20, 2023

Sources of Funding for Research Presented in a

Scientific Article or Scientific Article Itself

No funding was received for conducting this study.

Conflict of Interest

The authors have no conflicts of interest to declare

that are relevant to the content of this article.

Creative Commons Attribution License 4.0

(Attribution 4.0 International, CC BY 4.0)

This article is published under the terms of the

Creative Commons Attribution License 4.0

https://creativecommons.org/licenses/by/4.0/deed.en

_US