J. Risk Financial Manag. 2020, 13, x doi: FOR PEER REVIEW www.mdpi.comjournal jrfm [623013]

J. Risk Financial Manag. 2020, 13, x; doi: FOR PEER REVIEW www.mdpi.com/journal/ jrfm
Article 1
Corporate Bankruptcy Prediction 2
Daniel Ogachi1, Richard Ndege2 Peter Gaturu3, and Zoltan Zeman4* 3
1 Szent Istvan University 1; [anonimizat] 4
2 Jomo Kenyatta University of Agriculture and Tech nology 2; [anonimizat] 5
3 Jomo Kenyatta University of Agriculture and Technology 3 ; [anonimizat] 6
4 Szent Istvan University 4 ; [anonimizat] 7
* Correspondence: [anonimizat] ; Tel.: +36-3088 -23119 8
Abstract: Predicting bankruptcy of companies has been a hot subject of focus for many economists. 9
The rationale for developing predicting financial distress of a company is to develop a predictive 10
model used to f orecast the financial condition of a company by combining several econometric 11
variables of interest to the researcher. The study sought to introduce deep learning models for 12
corporate bankruptcy forecasting using textual disclosures. The study constructed a comprehensive 13
study model for predicting bankruptcy based on listed companies in Kenya. The study population 14
included all the 64 listed companies the listed companies in the Nairobi securities and exchange 15
market the study used 10 years financial stateme nts of listed companies. Logistic Analysis was used 16
in building a model for predicting the financial distress of a company. Findings revealed that asset 17
turnover, total asset and Working capital ratio had a positive coefficient. On the other hand Inventory 18
turnover, Debt equity ratio, Debtors turnover, Debt ratio and Current ratio have negative coefficients. 19
The study concluded that Inventory turnover, Asset turnover, Debt equity ratio, Debtors turnover, 20
Total asset, Debt ratio, Current ratio and Working ca pital ratio were the most significant ratios for 21
predicting bankruptcy. 22
Keywords: Bankruptcy, Insolvency, Financial Distress, Default, Failure, Forecasting Methods. 23
24
1. Introduction 25
Bankruptcy prediction is a technique of forecasting and projecting on company financial distress of 26
public firms. The purpose of predicting bankruptcy is fundamental in assessing the financial 27
condition of a company and prospects in its operations. Corpora te Bankruptcy prediction is a very 28
crucial phenomenon in economics. The financial soundness of a company is of great importance to 29
the various actors and participants of the business cycle. The participants and interested parties 30
include the policymakers, investors, banks, internal management, and the general public referred to 31
as consumers. Accurate prediction of the financial performance of companies is of great importance 32
to various stakeholders in making important and significant decisions concerning co mpanies. 33
Financial distress is a global phenomenon that affects companies across all sectors of the economy 34
(Zhang, Wang, & Ji, 2013) . 35
Additionally, bankruptcy prediction is essential for investors as well as suppliers or retailers to the 36
business. Credit lenders and investors need to evaluate the financial bankruptcy risk of a company 37

J. Risk Financial Manag. 2020, 13, x FOR PEER REVIEW 2 of 19
before making an investment or credit -gran ting decisions to avoid a significant loss: banks and other 38
credit lenders. A company’s suppliers or retailers always conduct credit transactions with the 39
company, and they also need to fully understand the company’s financial status and make decisions 40
on the credit transaction. To correctly predict a company’s financial distress is of great concern to the 41
various actors of a company. Problems concerning bankruptcy have necessitated the need for studies 42
to establish different stressors to companies to aid i nvestors in making investment decisions. 43
Corporate failures in significant economic companies have spurred research for better understanding 44
to develop prediction capabilities that guide decision making in investments. Financial distress 45
projections in co mpanies is a product of available data from listed companies, public firms that have 46
sunk. Available accounting ratios may be a vital indicator or signal to indicate danger. Typically firms 47
are quantified by many indicators that describe their business per formance based on mathematical 48
models constructed from past observations based on evidence from data. 49
Decisions of a corporate borrower on credit risk Traditionally were exclusively based upon subjective 50
judgments made by human experts, based on past expe riences and some guiding principles. 51
However, two significant problems associated with this approach include the difficulty to make 52
consistent estimates and the fact that it tends to be reactive rather than predictive (Cleofas -Sánchez, 53
Garcí a, Marqués, & Sánchez, 2016) . 54
Bankruptcy prediction is of great importance to all participants in the insurance market, including 55
insurance regulators, policyholders, agents, and insurance companies. As insurance products become 56
more and more familiar to the public, they strengthen the consumers' willingness to buy products. 57
However, since the service period of insurance products happens after the purchase of products, the 58
consumer is very concerned about whether the insurance company will be able to pay in the future 59
when purchasing products of the insurance company. Assessing the solvency of an insurance 60
company in the future during the product service period is very important to the policyholder's 61
purchase decision, and equivalently crucial to the operation of the insurance company. 62
In many instances, policyholders have a habit of thinking that large companies are financially stable 63
and that they are solvency guaranteed, which is not the case always. In assessing the creditworthiness 64
of companies, the various actors use solvency adequacy ratio indicators. In most companies, the 65
companies have a given solvency adequacy ratio used as a yardstick for measuring performance 66
required to be made public. One of the questions that stakeholders ask themselves is whether the 67
indicator is reliable for policyholders to forecast the solvency of a company using the current 68
information. 69
2. Materials and Methods 70
Studies have examined the causes of business failure indicated by values of bankruptcy scores 71
established during the decline stage of the business. In a survey of the 70 Estonian manufacturing 72
firms obtained the causes of bankruptcy from court judgments. The firms classified the reasons and 73
the types of failure, i.e., internal factors that are different from management deficiencies and external 74
factors to the firm. Ohlson’s model and a local (Grünberg’s) bankruptcy prediction models were used 75
to calculate bankruptcy scores for the first and second pre -bankruptcy years. Applying median tests 76
form independent samples to examine whether the different failure types are associated with 77
different failure risk. Findings revealed that multiple causes have a significantly higher bankruptcy 78

J. Risk Financial Manag. 2020, 13, x FOR PEER REVIEW 3 of 19
risk than s ingle reasons for the year before the declaration of bankruptcy. Results indicate that 79
numerous reasons lead to a considerably higher insolvency risk as compared to a single cause for the 80
year before bankruptcy disclosure (Lukason & Hoffman, 2014) . 81
Based on the first bankruptcy prediction model from the time of Altman, the prediction has gained 82
prominence and is at the epicentre of all economists and scientists all ove r the world. Early detection 83
of a possible threat to the financial performance of a company is a critical phenomenon in the world 84
of economic analysis. 85
Over the years, several models have been designed all over the world to measure the insolvency of 86
compa nies. One of the shortcomings of such models is the inability to transfer and apply one model 87
from one country to the other due to the difference in the economic conditions of the economy among 88
countries. A well -developed model in Hungary may not work well in another country; therefore, 89
there is a recommendation to develop a predictive model that takes into account the specific 90
conditions of a particular state using the real data on the financial situation (Svabova, Durica, & 91
Podhorska, 2018) . 92
Predicting the trend of stocks is an area of interest to researchers and investors due to the complex 93
nature underlying the data on prices and profitability. Machine learning in recent years has become 94
a popular stock market modelling technique. Three quantitative approaches to stock prediction in 95
the stock market have been established, with the most common method based on general indicators 96
of historical price and technical indicators. The proces s relies on chartist theory, which presumes that 97
past price patterns will reoccur in the future. The other second approach relies on sentiment analysis 98
which involves natural language processing techniques critical in analysing text -based data like 99
publish ed articles (Dong, 2019) . 100
Financial ratios are essential in predicting the bankruptcy of business ventures. Various variables 101
measure the financial soundness of an enterprise. In a study conducted in Indonesia on bank financial 102
ratios, the researcher used capital adequacy ratio (CAR), loan to deposit rate (LDR), non-performing 103
loan (NPL), Operating income operating costs ( BOPO), Return on Assets (ROA), Return on Equity 104
(ROE), and NIM. Using logit regression with 40 banks, LDR had a significant effect on the 105
profitability of banks in Indonesia. CAR, NPL, BOPO, ROE, and NIM had no considerable impact on 106
bankruptcy. Various companies in different sectors have different contributions to the growth of the 107
economy. Financial institutions like banks have the most role to play in such an eco nomy. 108
Banks contribute to economic growth through loan advancement as well as offering other financial 109
services. They also develop payment system mechanisms for all commercial sectors. Trust from the 110
community who are the key customers to their services mo stly influence the confidence of the 111
Activities of the bank. Any disturbance in institutions triggers a strong reaction from the customers 112
as well as the community. The economic crisis has struck banks and other institutions before leading 113
to financial dis tress. Companies face economic and financial difficulty. The imbalance between 114
revenue and expenditure can lead to financial trouble as well as higher capital costs rather than the 115
profit rate on the historical value of the investment. Financially distress ed companies are the ones 116
that are unable to pay their dues as of when they fall due. Several factors trigger bankruptcy, both 117
directly and indirectly. It can be due to non -performing loans or assets or massive withdrawals at 118
one time, which can lead to sy stematic risk. A mismatch of short -term funding structures in a 119
company can lead to liquidity problems. Financial statements measure a company’s performance. 120

J. Risk Financial Manag. 2020, 13, x FOR PEER REVIEW 4 of 19
Financial statements of the bank reflect the past performance of the company as well as used to 121
forecast for future financial performance. Financial ratios help to detect the financial distress of a 122
company. Predicting bankruptcy is very important as it acts as an early warning of financial trouble 123
(Ali Qalati et al., 2019) . 124
Predicting bankruptcy has gained attention for almost a century now and remains one of the hottest 125
topics in economics. The Financial distress prediction aims to design a model that blends the various 126
economic variables to foresee the condition of the firm. Several methods proposed statistical 127
modelling and artificial intelligence (Ziȩba, Tomczak, & Tomczak, 2016) . Textual disclosures 128
introduce deep learning models for bankruptcy prediction. Mai, Tian, Lee, & Ma (2019) established 129
that deep learning models yield superior forecasting on bankruptcy prediction. Blending textual data 130
with ratio analysis can improve the prediction accuracy. 131
Literature suggests that firms with a higher prior history of affirmative corporate social responsibility 132
(CSR), engagement are less likely to file for bankruptcy when they are financially distressed and are 133
expected to experience accelerated re covery from distress. Moral capital reduces bankruptcy 134
likelihood when the firm grows more massive. Additionally, capital mitigates bankruptcy likelihood 135
when the firm relies on intangible assets to operate and when firms operate in a more litigious 136
busine ss environment (Lin & Dong, 2018) . 137
Most institutions and researchers have focused on bankruptcy prediction due to the growth in the 138
complexity of global economies and an increasing number of corporate failures ignited by the 2008 139
crisis. Fisher’s linear discriminant has gained dominance and popularity in terms of accuracy (Garcí a, 140
Marqués, Sánchez, & Ochoa -Domí nguez, 2019) 141
In Lithuanian companies, a bankruptcy prediction model was built to assess the probability of 142
bankruptcy in companies. Private limited companies dominate the country. The study used 73 143
already bankrupt and 72 still operating to deduce a bankruptc y prediction model to be used for 144
predicting bankruptcy of business ventures. Using Mann -Whitney u test techniques, correlations and 145
multivariate discriminant analysis showed that the model was 89% accurate in predicting for 146
bankruptcy of private companies in Lithuania (Šlefendorfas, 2016) . 147
Other predictors of the bankruptcy of companies have been the convolutional neural network that is 148
being applied to identify the bankruptcy vice in a variety of fields. Convolutional neural networks in 149
financial analysis have been used to predict stock price movements. However, it is not a very common 150
technique that has been applied. Only very few studies have used it. Convolutional neural networks 151
approach uses two methods of the balance sheet and the profit and loss account to test for bankruptcy. 152
Hosaka (2019) established that predicting bankruptcy through trained networks is shown to have 153
higher performance as compared to decision trees, intelligent m achines, and linear discriminant 154
analysis, which was according to a study they conducted in the Japanese stock markets using 102 155
delisted companies and 2062 financial statements of listed companies. 156
In centuries research in predicting bankruptcy has been v ery challenging. Models have been built 157
from financial figures, stock market data, and specific firm variables — both low dimensional data 158
and high on company managers and directors in the models of prediction. Relational models are 159
found to have an improv ed prediction over financial models that are simple when detecting those 160
firms that are more riskier than others. Combining relational and economic data gives the most 161
substantial performance increase (Tobback, Bellotti, Moeyersoms, Stankova, & Martens, 2017) . 162

J. Risk Financial Manag. 2020, 13, x FOR PEER REVIEW 5 of 19
Managers are expected to carefully build bankruptcy prediction models and adjust them to the size, 163
type, and risk of the activities of the company (Boratyńska & Grzegorzewska, 2018) . 164
A study was conducted in India, which is an emerging economy to perform corporate d istress 165
prediction where bankruptcy details were not available. The study used firm -specific parameters to 166
capture any signs of distress for the firms. The study used standard Logistic and Bayesian modelling 167
to predict distressed firms in the corporate sec tor of India. The study found out that Bayesian 168
methodology provides for a consistent predictive capability of identifying the early signal of failure 169
in Indian companies (Shrivastava, Kumar, & Kumar, 2018) . 170
Financial misery and business failure is usually an extremely costly and disruptive event. Statistics 171
have been used to predict financial distress prediction in an attempt to predict the future of 172
businesses. Popular approaches to Discriminant analysis and logistic regression are used to predict 173
bankruptcy prediction. Using a variety of cost ratios the results by (Gepp & Kumar, 2015) in his study 174
showed that decision trees and survival analysis models have good prediction accuracy tha t justifies 175
their use and supports further investigation 176
In another study, the researcher analyzed the influence of financial distress on the investment 177
behaviour of companies. The study included companies from Germany, Canada, Spain, France, Italy, 178
the UK , and the USA. The researcher sought to use several institutions from different study 179
environments.using the Generalized Method of Moments (System -GMM), from panel data, the 180
results show that the influence of financial distress on investment is distinct ac cording to the 181
investment opportunities available to companies. So, companies in difficulties with fewer 182
opportunities have the highest propensity to under -invest, while firms in problems with better 183
opportunities do not present different investment behavi our than healthy companies (López – 184
Gutiérrez, Sanfilippo -Azofra, & Torre -Olmo, 2015) 185
In another study to establish whether a sensitivity variable, industry beta has a significant impact on 186
the firm's likelihood of default, the study used logistic regression and multiple discriminant analysis 187
on listed companies in India. The sensitivity va riable for industry factors, industry beta, is found to 188
be statistically significant in predicting defaults. Higher sensitivity to industry factors leads to an 189
increased probability of default (Agrawal & Maheshwari, 2019) . 190
Most bankruptcy research seems to have relied on parametric models like multiple discriminant 191
analysis and logit. The parametric models can only handle a finite number of predictors, which is the 192
most significant limitation of the model. The gradient boosting model has been advocated due to its 193
nature of accommodating for a vast amount of predictors that ca n be ranked in an orderly manner 194
ranging from best to worst based on their predictive power. A study on 1115 US bankruptcy filings 195
and 91 predictor variables, the study established that ownership structure/concentration and CEO 196
compensation treated as non -traditional as reliable predictors while unscaled market and accounting 197
variables as good predictors when studying firm size effects. Macro -economic variables, analyst 198
forecasts, and industry variables were found to be the weakest predictors (Jones, 2017) . 199
Improving corporate financial risk management requires a dynamic financial distress prediction. 200
Early researchers in constructing financial distress models (DFDP) ignored the time weight of 201
samples. A study on Dynamic financial distress prediction (DFDP) proposed two approaches based 202
on tim e weighting and Adaboost support vector machine (SVM) ensemble, which are more suitable 203
for DFDP in case of financial distress concept drift (Sun, Fujita, Chen, & Li, 2017) . 204

J. Risk Financial Manag. 2020, 13, x FOR PEER REVIEW 6 of 19
The dwindling in the profitability of listed companies not only intimidates the interests of the 205
enterprise and internal workforce but also leads to significant financial losses to investors. T herefore 206
companies must establish early predictive signs of financial difficulties in companies that will help 207
in issues relating to corporate governance. A study on 107 listed companies in the Shanghai Stock 208
Exchange and the Shenzhen Stock Exchange to dev elop the phenomenon of financial distress 209
interviewed companies that received the label of special treatment between 2001 to 2008. Data mining 210
techniques were used to build a model for establishing financial trouble in companies. One of the 211
critical contri butions of the paper was to discover that return on total assets, earnings per share, the 212
net profit margin of total assets, and cash flow per share, play an essential role in the prediction of 213
deterioration in profitability. Therefore the study provided a suitable method for forecasting the 214
financial distress of companies (Geng, Bose, & Chen, 2015) . 215
A study by Lian (2017) aimed to establish the role of the customer -supplier relationship on the 216
financial distress of the supplier used a customer -supplier sample between 1980 and 2014. The aim 217
of the study was to The findings revealed that a supplier's probability of financial trouble is positiv ely 218
related to the financial distress status of a customer. The study further showed that the relationship 219
is more pronounced when customer -supplier relationships are stronger and when a major customer 220
is likely to flop in the future, and when the supplier makes unique products. It is, therefore, vital to 221
understand the customer -supplier relationship when analyzing the probability of financial distress 222
of a firm. 223
A study was conducted on Predicting financial distress of agriculture companies in the European 224
Union by Klepac & Hampel (2017) 250 agriculture business companies interviewed, with 62 of them 225
having defaulted in 2014. From the results, increasing the distance to bankrupt cy leads to a decrease 226
in the average accurateness of the financial distress prediction. Therefore, there is a significant 227
difference flanked by the active and distressed companies in terms of liquidity, rentability, and debt 228
ratios (Klepac & Hampel, 2017) . 229
In another study to predict the financial distress companies in the trading and services sector in 230
Malaysia, the researcher used using financial distress companies as the dependent variable and 231
macroeconomic variables and financial ratios as the independent variables. Based on the results from 232
a Logit analysis, the study established that turnover ratio, debt ratio, total assets, working capital 233
ratio, net income to total assets ratio and base lending rate are the indepe ndent variables used to 234
predict financially distressed companies in the trading and services sector in Malaysia (Alifiah, 2014) 235
Whether to use accounting or market -based information to predict corporate default has b een a long – 236
standing research debate. Integrating a regime -switching mechanism, we establish a hybrid 237
bankruptcy prediction model with various loadings on accounting and market -based approaches to 238
re-examine bankruptcy prediction. Recommendations include Cr editors to increase the loading on 239
market -based information when large and liquid corporations are considered. 240
In the present states of the economy, there is an increasing number of organizations that are facing 241
financial difficulties, which may, at times, lead to bankruptcy. The deficiencies of customary 242
determining models inspire this examination. Partial Least Squares Logistic Regression allows for 243
incorporating a large number of ratios in the model and also solves the problem of correlations taking 244
into account the missing data in the matrix. Results obtained confirm the superiority of this method 245

J. Risk Financial Manag. 2020, 13, x FOR PEER REVIEW 7 of 19
compared to conventional methods of projecting for bankruptcy because the model allows 246
considering all the indicators in predicting financial distress (Ben Jabeur, 201 7). 247
Banks frequently adopt expert systems in supporting their decisions when advancing credit. One of 248
them being the Machine learning techniques that have been used for decades in issuing loans. Banks 249
are required to provide the logic behind their choic es in addition to being able to project the 250
performance of companies and individuals when assessing corporate applicants for loans. 251
One of the methods they use is the Data Envelopment Analysis (DEA) to evaluate several decisions 252
making units (DMU)ranked based on the best practice in their sector. Linear programming is 253
imperative as it is used in calculating corporate efficiency used as a measure of differentiating 254
between financially sound companies and those that are economically distressed. Results base d on a 255
study that sampled 742 listed Chinese companies observed over ten years suggest that Malmquist 256
DEA offers discernments into the competitive position of a company in addition to accurate financial 257
distress predictions based on the DEA efficiency meas ures (Li, Crook, & Andreeva, 2017) . 258
Ratio analysis financial indicators are the most popular variables that are u sed in bankruptcy 259
prediction models. They often exhibit heavily skewed results due to the presence of outliers. It is not 260
very clear on how different approaches affect the predictive power of models that predict bankruptcy. 261
One of the challenges faced in m odels is the lack of a clear cut of how to handle outliers and extremes 262
that affect the power of models — two ways of reducing outlier bias by omission and winsorization. 263
The categorization of financial ratios is an effective way of handling outliers conce rning the predictive 264
performance of bankruptcy prediction models. 265
Predicting financial distress in empirical finance has received a lot of attention from researchers 266
throughout the globe. Sampling small and medium enterprises in France using Logit model, A rtificial 267
Neural Networks, Support Vector Machine techniques, Partial Least Squares, and a hybrid model 268
integrating Support Vector Machine with Partial Least Squares, it has been established that within a 269
year of financial distress, Support Vector machine should be preferred because it is the best and 270
accurate method for predicting for bankruptcy. In the case of considering two years, then the hybrid 271
model outperforms Support Vector Machine, Logit model, Partial Least Squares, and artificial Neural 272
Networks had an overall accuracy of prediction by 94.28%. Financially distressed firms are found to 273
be smaller, more leveraged, and with lower repayment capacity. In addition to that, they have lower 274
profitability, liquidity, and solvency ratios. Creditors should therefore correctly evaluate the financial 275
position of firms and be keen on any signs that may lead to negative growth to avoid capital loss and 276
costs related risks (Mselmi, Lahiani, & Hamza, 2017) . 277
In a study by (Laitinen & Suvas, 2016) , to establish the influence of Hofstede's original cultural 278
dimensions on the prediction of financial distress, 1,255,768 non -failed and 22,594 failed yearly firm 279
observations were ob tained from 26 European countries. A model known as the logistic regression 280
model was used to predict the future financial position of a company in an international context. 281
Empirical findings revealed that Hofstede's dimensions significantly moderate the effects of 282
economic predictors in failure prediction. However, solvency (equity ratio) and return on assets ratio 283
(ROA), which is used to measure company success play a vital role in bankruptcy prediction models 284
irrespective of the position of moderating effects that they play at times. Solvency and profitability, 285
therefore, are imperative forecasters of bankruptcy in international financial modelling. The 286

J. Risk Financial Manag. 2020, 13, x FOR PEER REVIEW 8 of 19
contributions of regulating effects and further variables on the overall performance of prediction 287
models are not resilient due to the dominant role of the equity ratio across cultures. 288
289
In project management, bankruptcy prediction is also very vital in projecting for a financial crisis 290
during the execution of the project. Financial crisis prediction mode ls have received special attention 291
from contractors. In spite of the possibility of knowing the signs of the financial difficulties of a project 292
through financial distress prediction, no attempt has been made to predict for the financial hardship 293
that a co ntractor can suffer before reaching a financial crisis including highly visible legal events, 294
such as bankruptcy, default, and delisting. This means that there is a significant gap between the 295
models and real -world applications in terms of the forecast per iod and definition of the financial 296
crisis. The prediction performance of the proposed model was evaluated using the financial 297
statements of contractors in South Korea from 2007 to 2012. Findings revealed that by predicting 298
financial distress of the contra ctor from the early stages of a construction project to the end -stage with 299
high accuracy, this model could help project owners and stakeholder to avoid damage caused by the 300
financial crisis during a project (Choi, Son, & Kim, 2018) 301
It is argued that the forecasting horizon for bankruptcy should be one year. Beyond th is period, the 302
accuracy of the prediction model is usually low. The ability to use models to provide for bankruptcy 303
forecasting is an essential characteristic for financial institutions that apply the prudential accounting 304
concept. There is a need to impro ve forecasting models for five years by quantizing how the financial 305
soundness of companies change within time. Results from (du Jardin, 2017) indicate that whichever 306
technique is used to design prediction models, t he accuracy of the models can be improved when the 307
period exceeds two years 308
Theoretical models proposed argue that financial distress in small – and medium -sized enterprises 309
(SMEs) emanate from the interaction financial distress likelihood and the degree of the consequences 310
borne whenever a financial failure occurs. Evidence from empirical research involving five 311
European countries, where the bankruptcy laws are representative of prevailing institutional 312
traditions, supports this model. Studies have reveale d that financial distress by a company depends 313
not only on variables with an influence on the amount of time and costs incurred during the 314
insolvency process but also on the likelihood of financial distress but also (Keasey, Pindado, & 315
Rodrigues, 2015) . 316
In the accounting and finance domains, financial distress prediction is of splendid utility for all of the 317
monetary stakeholders. The challenge of correct assessment of commercial enterprise failure 318
prediction, mainly under situations of economic crisis, is regarded to be complicated. Although there 319
has been much profitable research on financial distress detection, seldom probabilistic methods were 320
carried out. An in -depth analysis was conducted using real -world bankruptcy data, displaying that, 321
in addition t o a probabilistic interpretation, the GP can efficiently enhance the financial disaster 322
prediction overall performance with excessive accuracy when in contrast to the other approaches. 323
The study moreover generates an entire graphical visualization to impro ve the grasp of the 324
exceptional attained achievements, efficaciously compiling all the conducted experiments in a 325
significant way (Antunes, Ribeiro, & Pereira, 2017) . 326
Altman’s Z -score developed in 1968 has been used widely in making decisions on financial failure by 327
companies withing different countries. Agribusiness is a vital enterprise in Lithuania, and recent 328

J. Risk Financial Manag. 2020, 13, x FOR PEER REVIEW 9 of 19
developments of consolidation and long -term government subsidies assess the financial health of 329
such agencies important not only for the owners, however, for the different stakeholders as well. In 330
a study on Financial distress predict ion on the three listed Lithuanian agricultural companies, it was 331
found out that the model correctly locates companies into “safe” and “grey” zones, which gives initial 332
information for the stakeholders. Further, the study established that financial and non -financial 333
factors constituting Z -score could provide additional details for forecasting a firm’s performance 334
(Kiaupaite -Grushniene, 2016) . 335
In the design of a monetary financial disaster prediction model, financial ratio selection, and classifier 336
design play the most critical roles. A methodology based totally on expert opinion, statistical concept, 337
and computational intelligence method has been widely applied . In this study, a hybrid shape 338
integrating mathematical idea and computational talent technique was once developed using a 339
genetic algorithm (GA) with statistical measurements and fuzzy good judgment based fitness 340
features for essential ratio selection. In the experiments, two monetary ratio sets, one extracted from 341
the recommendations of different research and the different got by employing the use of the GA 342
toolbox in the SAS statistical software program package, have been utilized to take a look at the 343
proposed ratio choice schemes. A distinction between the improved hybrid shape and different well – 344
applied structures was additionally given. Used to gauge the performance of the prediction model 345
proposed were the experimental results of financial data based on less than four year period prior to 346
bankruptcy occurrence (Chou, Hsieh, & Qiu, 2017) . 347
Introduction to predictive bankruptcy is an objective and realis tic problem facing companies and 348
firms, and because of its frequency, it has discovered a specific niche in monetary and investment 349
literature following the motto "prevention is better than cure." In this respect, more than a few 350
fashions have been present ed based totally on motives and motives for bankruptcy. Numerous 351
research has been committed to discovering high -quality experimental techniques in predicting the 352
economic crisis. As a result, exceptional patterns have been generated uniquely to predict th e 353
financial crisis. Prediction of financial disaster is signific ant for all corporations due to the fact it has 354
a profound effect on the economic system and raises expenses inflicting many social problems. There 355
are many strategies and methods via which companies and monetary analysts can predict 356
bankruptcy. A combination of various ratios used for bankruptcy prediction and classification 357
fashions can help to choose financial ratios and amplify prediction accuracy. 358
Neural networks are one of the numerous methods in predicting financial distress of industrial 359
groups, which is used right here considering elements such as accuracy in predicting and health of 360
model for predicting financial distress in the industry. Concerning management, time series 361
prediction is one of the applications of neural networks. Corporate financial disaster is typically 362
superb in capital market liquidity and economic development. When financial distress occurs, b anks 363
generally limit bankrupt companies’ credits, and in exchange for loans, they demand more 364
exceptional pastime to compensate for their increased risk. Given the reverse impacts of financial 365
distress on capital markets and the economy, researchers and be neficiaries have tried to create and 366
advance various predicting models using distinct procedures to minimize its effects and incurred 367
losses (Salehi & Pour, 2016) 368
369
370

J. Risk Financial Manag. 2020, 13, x FOR PEER REVIEW 10 of 19
371
Academicians and practitioners have conducted intensive research regarding models for bankruptcy 372
prediction and default events to manage cred it risk. Traditional statistics techniques (e.g., logistic 373
regression and discriminant analysis) as well as and early artificial intelligence models (e.g., artificial 374
neural networks) have evaluated bankruptcy. In the study, machine learning models (suppor t vector 375
machines, bagging, boosting, and random forest) were tested to forecast for bankruptcy one year 376
before the event and compare their performance with results from the neural networks, logistic 377
regression and discriminant analysis data for the years 1985 to 2013 on North American firms, 378
analyzing more than 10,000 firm -year observations. Insightful findings revealed a substantial 379
improvement in the accuracy of the prediction using machine learning techniques. 380
Comparing the best models, with all predic tive variables, the machine learning technique related to 381
random forecast led to 87% accuracy, whereas logistic regression and linear discriminant analysis led 382
to 69% and 50% accuracy, respectively, in the testing sample. We find that bagging, boosting, an d 383
random forest models outperform the other techniques and that all prediction accuracy in the testing 384
sample improves when the additional variables are included. (Barboza, Kimura, & Alt man, 2017) . 385
In a study to determine the probability of the determinants of corporate financial distress, a large 386
dataset was used that consists of almost 31,000 Greek non-public companies, we observe the 387
determinants of the likelihood of corporate finan cial distress. Using a multi -period logit model, we 388
discover that leverage, size, profitability, retained earnings to total assets and liquidity ratio, an 389
export dummy variable, the tendency to pay out dividends and the boom rate in actual GDP are 390
sturdy predictors of the chance of financial distress for Greek private firms. A mannequin, which 391
includes these variables, exhibits the perfect in-sample and out -of-sample overall performance in 392
phrases of efficiently classifying corporations that went bankrupt a s extra likely to go bankrupt. The 393
predictive capacity of the model stays when we expand the forecast horizon, suggesting that the 394
model works well over brief and longer time horizons (Charalambakis & Garrett, 2019) . 395
Bankruptcy prediction and credit risk analysis is one of the most significant issues in the subject of 396
accounting and economics selection making. Developing an excellent classification rule induction 397
(CRI) framework for bankruptcy prediction and credit danger evaluation in a fantastic time is 398
imperative to prevent the commercial enterprise communities from being bankrupt. Traditional 399
statistical methods and synthetic talent techniques play an influential position in predicting financial 400
distress and credit risks. Most of the before research works were carried out on quantitative methods, 401
while few studies have proposed on qualitative techniques to improvise the overall performance of 402
bankruptcy prediction models. The discovery of financial distress predictio n in a qualitative way is 403
an essential venture because it depends on the subjective expertise of the experts. In this paper, a 404
unified framework for qualitative and quantitative financial distress analysis, the usage of Ant 405
Colony Optimization (ACO) primar ily based ant-miner algorithm, was proposed. Three one-of-a- 406
kind natured datasets are used to present an accurate result. For this experiment, we have 407
accumulated a qualitative_bankruptcy dataset and benchmarked through the UCI repository. The 408
proposed method is successfully utilized, and the overall performance analysis shows that the ant – 409
miner approach is better than present classifiers particularly Logistic Regression (LR), Multilayer 410
Perceptron (MLP), Random Forest ( RF) and Radial Basis Function (RBF) in phrases of a variety of 411
performance analysis factors. Furthermore, the proposed ant -miner model is determined to be a more 412

J. Risk Financial Manag. 2020, 13, x FOR PEER REVIEW 11 of 19
appropriate technique for financial disaster prediction when, in contrast to other typical statistical 413
and artificial Genius techniques (Uthayakumar & Vengattaraman, 2017) . 414
In a study by (Manzaneque, Priego, & Merino, 2016) , the paper explores some mechanisms of 415
corporate governance (ownership and board characteristics) in Spanish listed agencies and their 416
influence on the likelihood of financial distress. An empirical learn about was once performed 417
between 2007 and 2012 the usage of a matched -pairs lookup design with 308 observations, with 1/2 418
of them labelled as distressed and non -distressed. Based on the previous study modelling Pindado, 419
Rodrig ues, and De la Torre (2008), broader thinking of bankruptcy are used to outline enterprise 420
failure. Based on a logistic model, the outcomes confirm that in hard situations before the bankruptcy, 421
the have an effect on board ownership, and the proportion of impartial administrators on commercial 422
enterprise failure possibility is comparable to those exerted in more extreme cases. These effects go 423
one step further to offer a negative relationship between the board dimension and the possibility of 424
financial distress. This result is interpreted as a shape of growing range and to enhance the right of 425
entry to the information and resources, e specially in contexts where the ownership is noticeably 426
centred, and massive shareholders have a higher power to impact the board structure. However, the 427
results verify that ownership concentration does not have a widespread impact on financial distress 428
possibility in the Spanish context. It is argued that massive shareholders are passive as regard s more 429
excellent monitoring of management and, alternatively, they do no longer have ample incentives to 430
hold lower back the financial distress. These findings have necessary implications in the Spanish 431
context, where it places several modifications in the regulatory checklist necessities which have been 432
carried out with appreciation to corporate governance, and where there is no empirical proof related 433
to this respect . 434
According to (Arroyave, 2018) Logit and discriminant analyses have been used for company financial 435
distress prediction in several studies due to the fact the last century. In the latest years, there have 436
been dozens of research evaluating the quite a few models available, along with the ones referred to 437
above and additionally, probit, synthetic neural networks, help vector machines, amongst others. 438
This paper for the first time for Colomb ia provides a comparative analysis of the effectiveness of 439
countless models predicting corporate bankruptcy. Such models have in the past, been usually used 440
concerning European and North American markets, whereas here they are utilized to the financial 441
ratios of three firms located in Colombia. The fundamental objective is to corroborate the validity of 442
these fashions in terms of their ability to predict company failure in the Latin American context, 443
especially for two bankrupt Colombian companies and one wholesome one. The evaluation is carried 444
out using financial disaster forecasting fashions extensively proposed in the literature and used 445
systematically in developed countries. 446
Accurate prediction of company economic distress is essential for managers, creditors, and buyers to 447
take corrective measures to minimize loss. Many quantitative methods have been employed to 448
advance empirical fashions for predicting corporate bankruptcy. However, there are many facts 449
disclosed in the c ompanies' monetary statements, what events need to be selected for building the 450
empirical models to maximize predictive accuracy. In this study, extra than 20 models primarily 451
based on six features, ranking techniques are tested on North American organizat ions and Chinese 452
listed companies. The experimental outcomes are useful to increase monetary fashions with the aid 453

J. Risk Financial Manag. 2020, 13, x FOR PEER REVIEW 12 of 19
of selecting the acceptable quantitative methods and features selection approach (Zhou, Lai, & Yen, 454
2012) . 455
Financial data has been widely used to graph financial disaster prediction models. All lookup works 456
that have studied such fashions expect that monetary statements are reliable. However, the actuality 457
is a bit different. Indeed, companies can also tend to present their financial debts depending on 458
particular circumstances, in particular when looking to exchange the understanding of the hazard 459
incurred by using their partners, and as a consequence, distort or alter some of them. Consequently, 460
one can also surprise to what extent such “manipulations,” called revenue management, might also 461
influence any mannequin that relies on accounting data. 462
This is why we find out about how income administration may additionally have an effect on 463
monetary variables and how they can indirectly distort predictions made through failure models. For 464
this purpose, we used a measure that makes it possible to investigate practicable account 465
manipulations and now not fantastic manipulations. Our outcomes exhibit that when these 466
distortions are measured and used with different economic variables, fashions are more accurate than 467
those that rely on real financial data. They also manifest that the enhancement of model accuracy is 468
undoubtedly due to a reduction of type -I error, the most costly mistake in monetary phrases (Du 469
Jardin, Veganzones, & Séverin, 2019) . 470
The paper is typically committed to the financial distress prediction fashion s and their capacity to 471
assess a financial distress probability for Lithuanian companies. The find out about confirmed that 472
the most frequent type of organisations in Lithuania is a private constrained company, therefore, the 473
primary objective was to analy se such companies' monetary statistics and by way of the usage of 474
these results, create a new financial distress prediction model, which would permit to predict the 475
financial disaster likelihood as precisely as possible. One hundred forty -five organization s (73 476
already bankrupt and seventy -two nevertheless operating) had been chosen as an essential sample, 477
and through the use of multivariate discriminant evaluation stepwise method, a linear function 478
Z^sub GS^ has been created. 156 one of a kind economic ratios have been selected as the primary 479
input data by using the use of correlation calculation between financial distress and nonetheless 480
running companies and Mann – Whitney U test techniques. The results confirmed that 89% of 481
companies had been labelled correctly, which states that the mannequin is robust adequate to predict 482
financial distress probability for restricted private corporations working in Lithuania in a sufficient 483
accuracy (Šlefendorfas, 2016) 484
Corporate leverage responds in a different way to employees’ rights in bankruptcy, depending on 485
whether or not it is pushed by using strategic issues in wage bargaining or via credit score constraints. 486
Using new information on employees’ rights in bankruptcy, we estimate there has an effect on 487
leverag e, exploiting time -series, cross -country, and firm -level variation in the data. For financially 488
unconstrained firms, effects accord with the strategic debt model: leverage will increase extra in 489
response to rises in corporate property values or profitabilit y if employees have sturdy seniority in 490
liquidation and vulnerable rights in a restructuring. Instead, in financially restrained companies, 491
leverage responds less to these shocks if personnel have better priority (Ellul & Pagano, 2019). 492
Some ratios were se lected to help in the prediction of financial distress in companies based on the 493
usefulness and relevance in previous studies. Ratio analysis is essential in a 10 -year trend analysis. 494
In this study, the ratios of interest to the researcher included total a sset debt ratio, current ratio, quick 495

J. Risk Financial Manag. 2020, 13, x FOR PEER REVIEW 13 of 19
ratio, turnover ratio, working capital ratio, and net income to total assets ratio. ROE, ROA and Net 496
profit margin were selected as the dependent variables denoting the financial performance of 497
companies. The study pop ulation for this research included all the listed companies in the Nairobi 498
securities and exchange market (NSE). Currently, there are 64 listed companies in Kenya. Included 499
in the listed companies also include companies that were delisted at some point due to financial 500
distress. The study used 10 years of financial statements of listed companies. The financial statements 501
were obtained from the Capital Markets Authority as well as the Nairobi Securities. Logit Analysis 502
was used in building a model for predic ting the financial distress of a company. Logit Analysis was 503
necessary for the study because it provided for the probabilities of occurrence of the outcome. 504
The study was guided by the model below; 505
Yi = α +β1X1 + β2X 2 +μI………………………………….(1) 506
Where, 507
Xi, X2…Xn =the independent (explanatory) variables (Asset Turnover, Debt to Equity Ratio, Debtors 508
Turnover, Total Asset, Debt Ratio, Current Ratio, Quick Ratio, Inventory Turnover Ratio, Working 509
Capital Ratio, 510
Yi=Dependents variables (Return on Assets, Retu rn on Equity, and Net profit margin 511
Yi = 1 if a company is financially distressed 512
Yi = 0 if a company is not financially distressed 513
The first equation based on logistic regression can be denoted as 514
ln 𝑃
1−𝑃=α +β1X 1 +β2X 2 +μ (1) ……………………………………… (2) 515
therefore the probability of a company becoming financially distressed will be given by 516
𝑝=1
1+𝑒−(α+B1 X1+B2 X 2+ … BnXn……………………………………… (3) 517
Values with a figure of 0.5 and above denote that the company is financially distressed while numbers 518
below 0.5 show that a company is not economically distressed. 0 indicates an indifferent state of the 519
company. On the other hand, negative coefficients re duce the probability of financial distress while 520
positive factors increase the chance of occurrence of bankruptcy prediction. The study used SPSS 521
software to aid in data analysis. 522
523
3. Results and Discussions 524
Table 1: Collinearity Statistics of the variabl es 525
Variable Tolerance VIF
Inventory turnover .959 1.043
Asset turnover .925 1.081
Debt equity ratio .978 1.022
Debtors turnover .958 1.044
Total asset .947 1.056
Debt ratio .917 1.091
Current ratio .932 1.073

J. Risk Financial Manag. 2020, 13, x FOR PEER REVIEW 14 of 19
Quick ratio .372 2.685
Working capital ratio .969 1.032
526
Based on the multicollinearity analysis, the quick ratio was excluded from the subsequent 527
investigation that is the stepwise logit analysis due to its high multicollinearity. The independent 528
variable chosen under this model included inventory turnover, asset turnover, debt -equity ratio , 529
debtors turnover, total assets, debt ratio, current ratio and working capital. ROE and ROA were the 530
dependent variables of the study. Stepwise logit analysis was conducted to evaluate the impact of a 531
number of independent variables on the likelihood tha t companies will be financially distressed. 8 532
independent variable model was drawn to denote their relationship with the dependent variable. 533
534
Table 2: Test Statistics
N 550
Chi-Square 119.969
df 3
Asymp. Sig. .000
The final model was statistically significant, with a chi -square value of 119.969 and 3 degrees of 535
freedom and sig value (p<0.005)=0.000. This indicates that the model was able to distinguish between 536
financially distressed and non -financially distressed companies. 537
538
Table 3: C lassification table 539
Predict
Distressed Percentage Correct
Observed 0 1
Distressed 0 51 9 82.0
1 8 52 84.0
Overall Percentage 83.0
The model correctly classified 85% of overall cases or also known as the percentage accuracy in 540
classification which is higher than the 50% when the analysis was conducted without any of the 541
independent variables that are used in the model. The classificat ion table was as shown above. 542
Table 4: Logit Analysis Results 543
Iv B S.e Wald Sig-value

Inventory turnover -0.068 0.178 5.245 0.000***
Asset turnover 2.269 0.935 7.865 0.006***
Debt equity ratio -4.987 1.452 6.458 0.003***
Debtors turnover -0.075 0.009 8.456 0.001***
Total asset 2.853 0.759 9.985 0.003***
Debt ratio -3.296 2.498 8.321 0.002***
Current ratio -0.059 0.085 6.429 0.033**
Working capital ratio 0.086 0.026 6.382 0.010**
***statistically significant at 1% level 544
**Statistically significant at 5% level 545

J. Risk Financial Manag. 2020, 13, x FOR PEER REVIEW 15 of 19
The findings from the table above show 8 predictor variables that contribute to the logistic analysis 546
model. The predictors include Inventory turnover, Asset turnover, Debt equity ratio, Debtors 547
turnover, Total asset , Debt ratio, Current ratio and Working capital ratio. The dependent variable for 548
the study included the ROA and ROE. Wald statistic was conducted to show the contribution of each 549
variable to the model. The p -value is significant to the model in establishi ng the level of significance 550
and contribution of each variable. Variables with sig -value P<0.005 contribute significantly to the 551
model. Asset turnover, Total asset and Working capital ratio have positive coefficients. This shows 552
that they increase the cha nces of bankruptcy. They have a more significant contribution to predicting 553
bankruptcy in companies. Higher values in the mentioned ratios can lead to financial distress in 554
companies. On the other hand Inventory turnover, Debt equity ratio, Debtors turnove r, Debt ratio 555
and Current ratio have negative coefficients. Negative coefficients reduce the risk of financial distress 556
in listed companies in the Nairobi Securities and Exchange Market. 557
558
The logistic regression model can be as shown below; 559
𝑃=1
1+𝑒−(−0.068X1 +2.269X2 −4.987X3 −0.075X4 +2.853X5 −3.296X6 −0.059X7 +0.086X8 560
Where, 561
X1=Inventory turnover 562
X2=Asset turnover 563
X3=Debt equity ratio 564
X4=Debtors turnover 565
X5=Total asset 566
X6=Debt ratio 567
X7=Current ratio 568
X8=Working capital ratio 569
Given the value of X1 -X8, the cost of B can be established. Value of greater than 0.5 shows the 570
possibility of a company going into financial distress. This study, therefore, identified Inventory 571
turnover, Asset turnover, Debt equity ratio, Debtors turnover, Total asset, Debt ratio, Cu rrent ratio 572
and Working capital ratio as the most significant ratios for projecting for bankruptcy. The findings 573
show that financial ratios can be used to predict financially distressed companies in the Nairobi 574
securities and exchange market. 575
4. Conclusion s 576
This study found out that the predictor variables that can be used to predict financial distress 577
companies in the NSE included Inventory turnover, Asset turnover, Debt equity ratio, Debtors 578
turnover, Total asset, Debt ratio, Current ratio and Working cap ital ratio as the most significant ratios. 579
P-value of greater than 0.5 shows a possibility of a company going into financial distress while 580
smaller amounts show the absence of financial trouble in companies listed in the NSE. 581
582
Funding: This resear ch receiv ed no external funding 583
Acknowledg ements: I would like to acknowledge the Tempus Public Foundation for awarding me a PhD 584
scholarship to study in Hungary. Special thanks to Lecturers and staff at the Szent Istvan University. 585

J. Risk Financial Manag. 2020, 13, x FOR PEER REVIEW 16 of 19
Conflicts of Interest: The author s declare no conflict of interest. 586
Appendix A 587
Profitability ratios
Net profit margin Gross profit/Sales
EBITDA margin EBITDA/Sales
EBIT margin EBIT/Sales
Net profit margin Net income/Sales
ROE Net income/Equity
ROA Net income/Total assets
Liquidity ratios
Current liquidity ratio Current assets/Current liabilities
Quick liquidity ratio (Current assets –inventory)/Current
liabilities
Absolute liquidity ratio (cash ratio) Cash and cash equivalents/Current
liabilities (immediately chargeable)
Activity ratios
Receivable turnover rate Sales/Receivable
Inventory turnover rate Cost of goods sold/Inventory
Net-working capital turnover rate Sales/ (Current assets –current liabilities)
Asset turnover rate Sales/Total assets
Equity turnover rate Sales/Equity
Fixed asset turnover rate Sales/Fixed assets
Current assets turnover rate Sales/Current assets
Debtors Turnover Ratio Net Credit Sales/Average Accounts
Receivable
Creditors Velocity Total Purchases/Total Trade Creditors
Progress Ratios
Assets growth rate (Total assets –Total assets −1)/Total assets −1
Net-profit growth rate (Net income–Net income −1)/Net income −1
Sales growth rate (Sales–Sales −1)/Sales −1
Asset structure ratios
Share of current assets to total assets Current assets/Total assets
Share of inventories to current assets Inventory/Current assets
Share of cash and cash equivalents to
current assets Cash and cash equivalents/Current assets
Share of fixed assets to total assets Fixed asset/Total assets
Debt coverage ratio
Current liabilities ratio Current liabilities/Total liabilities
Interest coverage ratio EBIT/Interest
Debt ratio Total liabilities/Equity
Leverage Total liabilities/Total assets

J. Risk Financial Manag. 2020, 13, x FOR PEER REVIEW 17 of 19
References 588
Le,Agrawal, K., & Maheshwari, Y. (2019). Efficacy of industry factors for corporate default prediction. 589
IIMB Management Review , 31(1), 71–77. https://doi.org/10.1016/j.iimb.2018.08.007 590
Ali Qalati, S., li, W., Iqbal , S., Hind, H., Wen Yuan, L., & Yassir Hussain, R. (2019). Quality of Services 591
and Satisfaction of Customer:Mediated by Consumer Buying Decision Benchmark sensitivity 592
in performance evaluation of funds View project SMEs Performance in Pakistan View project 593
Quality of Services and Satisfaction of Customer:Mediated, 11(12), 183 –189. 594
https://doi.org/10.7176/EJBM 595
Alifiah, M. N. (2014). Prediction of Financial Distress Companies in the Trading and Services Sector 596
in Malaysia Using Macroeconomic Variables. Proced ia – Social and Behavioral Sciences , 129, 90–98. 597
https://doi.org/10.1016/j.sbspro.2014.03.652 598
Antunes, F., Ribeiro, B., & Pereira, F. (2017). Probabilistic modelling and visualization for bankruptcy 599
prediction. Applied Soft Computing Journal , 60, 831–843. https://doi.org/10.1016/j.asoc.2017.06.043 600
Arroyave, J. (2018). A comparative analysis of the effectiveness of corporate bankruptcy prediction 601
models based on financial ratios: Evidence from Colombia. Journal of International Studies , 11(1), 602
273–287. https://doi.org/10.14254/2071 -8330.2018/11 -1/21 603
Barboza, F., Kimura, H., & Altman, E. (2017). Machine learning models and bankruptcy prediction. 604
Expert Systems with Applications , 83, 405–417. https://doi.org/10.1016/j.eswa.2017.04.006 605
Ben Jabeur, S. (2017). Bankruptcy prediction using Partial Least Squa res Logistic Regression. Journal 606
of Retailing and Consumer Services , 36(November 2016), 197 –202. 607
https://doi.org/10.1016/j.jretconser.2017.02.005 608
Boratyńska, K., & Grzegorzewska, E. (2018). Bankruptcy prediction in the agribusiness sector: 609
Lessons from qua ntitative and qualitative approaches. Journal of Business Research , 89(June 610
2017), 175 –181. https://doi.org/10.1016/j.jbusres.2018.01.028 611
Charalambakis, E. C., & Garrett, I. (2019). On corporate financial distress prediction: What can we 612
learn from private firms in a developing economy? Evidence from Greece. Review of Quantitative 613
Finance and Accounting , 52(2), 467–491. https://doi.org/10.1007/s11156 -018-0716 -7 614
Choi, H., Son, H., & Kim, C. (2018). Predicting financial distress of contractors in the construc tion 615
industry using ensemble learning. Expert Systems with Applications , 110, 1–10. 616
https://doi.org/10.1016/j.eswa.2018.05.026 617
Chou, C. H., Hsieh, S. C., & Qiu, C. J. (2017). Hybrid genetic algorithm and fuzzy clustering for 618
bankruptcy prediction. Applied Soft Computing Journal , 56, 298 –316. 619
https://doi.org/10.1016/j.asoc.2017.03.014 620
Cleofas -Sánchez, L., Garcí a, V., Marqués, A. I. , & Sánchez, J. S. (2016). Financial distress prediction 621
using the hybrid associative memory with translation. Applied Soft Computing Journal , 44, 144– 622
152. https://doi.org/10.1016/j.asoc.2016.04.005 623
Dong, Z. (2019). Dynamic Advisor -Based Ensemble (DYNABE): a Case study in stock trend prediction of 624
critical metal companies . PLoS ONE (Vol. 14). https://doi.org/10.1371/journal.pone.0212487 625
du Jardin, P. (2017). Dynamics of firm financial evolution and bankruptcy prediction. Expert Systems 626
with Applications , 75, 25–43. https://doi.org/10.1016/j.eswa.2017.01.016 627
du Jardin, P., Veganzones, D., & Séverin, E. (2019). Forecasting Corporate Bankruptcy Using Accrual – 628
Based Models. Computational Economics , 54(1), 7–43. https://doi.org/10.1007/s10614 -017-9681 -9 629
Ellul, A., & Pagano, M. (2019). Corporate leverage and employees’ rights in bankruptcy. Journal of 630

J. Risk Financial Manag. 2020, 13, x FOR PEER REVIEW 18 of 19
Financial Economics , 133(3), 685–707. https://doi.org/10.1016/j.jfineco.2019.05.002 631
Garcí a, V., Marqués, A. I., Sánchez, J. S., & Ochoa -Domí nguez, H. J. (2019). Dissim ilarity -Based Linear 632
Models for Corporate Bankruptcy Prediction. Computational Economics , 53(3), 1019 –1031. 633
https://doi.org/10.1007/s10614 -017-9783 -4 634
Geng, R., Bose, I., & Chen, X. (2015). Prediction of financial distress: An empirical study of listed 635
Chin ese companies using data mining. European Journal of Operational Research , 241(1), 236–247. 636
https://doi.org/10.1016/j.ejor.2014.08.016 637
Gepp, A., & Kumar, K. (2015). Predicting Financial Distress: A Comparison of Survival Analysis and 638
Decision Tree Techniqu es. Procedia Computer Science , 54, 396 –404. 639
https://doi.org/10.1016/j.procs.2015.06.046 640
Hosaka, T. (2019). Bankruptcy prediction using imaged financial ratios and convolutional neural 641
networks. Expert Systems with Applications , 117, 287 –299. 642
https://doi.or g/10.1016/j.eswa.2018.09.039 643
Jones, S. (2017). Corporate bankruptcy prediction: a high dimensional analysis. Review of Accounting 644
Studies , 22(3), 1366–1422. https://doi.org/10.1007/s11142 -017-9407 -1 645
Keasey, K., Pindado, J., & Rodrigues, L. (2015). The dete rminants of the costs of financial distress in 646
SMEs. International Small Business Journal: Researching Entrepreneurship , 33(8), 862–881. 647
https://doi.org/10.1177/0266242614529317 648
Kiaupaite -Grushniene, V. (2016). Altman Z -Score Model for Bankruptcy Forecasti ng of the Listed 649
Lithuanian Agricultural Companies, (Icaat), 222 –234. https://doi.org/10.2991/icaat -16.2016.23 650
Klepac, V., & Hampel, D. (2017). Predicting financial distress of agriculture companies in the EU. 651
Agricultural Economics (Czech Republic) , 63(8), 347–355. https://doi.org/10.17221/374/2015 – 652
AGRICECON 653
Laitinen, E. K., & Suvas, A. (2016). Financial distress prediction in an international context: 654
Moderating effects of Hofstede’s original cultural dimensions. Journal of Behavioral and 655
Experimental Finance , 9, 98–118. https://doi.org/10.1016/j.jbef.2015.11 .003 656
Li, Z., Crook, J., & Andreeva, G. (2017). Dynamic prediction of financial distress using Malmquist 657
DEA. Expert Systems with Applications , 80, 94–106. https://doi.org/10.1016/j.eswa.2017.03.017 658
Lian, Y. (2017). Financial distress and customer -supplier relationships. Journal of Corporate Finance , 43, 659
397–406. https://doi.org/10.1016/j.jcorpfin.2017.02.006 660
Lin, K. C., & Dong, X. (2018). Corporate social responsibility engagement of financially distressed 661
firms and their bankruptcy likelihood. Advances in Accounting , 43(July), 32 –45. 662
https://doi.org/10.1016/j.adiac.2018.08.001 663
López -Gutiérrez, C., Sanfilippo -Azofra, S., & Torre -Olmo, B. (2015). Investment decisions of 664
companies in financial distress. BRQ Business Research Quarterly , 18(3), 174 –187. 665
https:// doi.org/10.1016/j.brq.2014.09.001 666
Lukason, O., & Hoffman, R. C. (2014). Firm Bankruptcy Probability and Causes: An Integrated Study. 667
International Journal of Business and Management , 9(11), 80 –91. 668
https://doi.org/10.5539/ijbm.v9n11p80 669
Mai, F., Tian, S., Le e, C., & Ma, L. (2019). Deep learning models for bankruptcy prediction using 670
textual disclosures. European Journal of Operational Research , 274(2), 743 –758. 671
https://doi.org/10.1016/j.ejor.2018.10.024 672
Manzaneque, M., Priego, A. M., & Merino, E. (2016). Corporate governance effect on financial distress 673

J. Risk Financial Manag. 2020, 13, x FOR PEER REVIEW 19 of 19
likelihood: Evidence from Spain. Revista de Contabilidad -Spanish Accounting Review , 19(1), 111– 674
121. https://doi.org/10.1016/j.rcsar.2015.04.001 675
Mselmi, N. , Lahiani, A., & Hamza, T. (2017). Financial distress prediction: The case of French small 676
and medium -sized firms. International Review of Financial Analysis , 50, 67–80. 677
https://doi.org/10.1016/j.irfa.2017.02.004 678
Prusak, B. (2018). Review of Research into Enterprise Bankruptcy Prediction in Selected Central and 679
Eastern European Countries. International Journal of Financial Studies , 6(3), 60. 680
https://doi.org/10.3390/ijfs6030060 681
Salehi, M., & Pour, M. D. (2016). Bankruptcy prediction of listed companies in Te hran Stock Exchange. 682
Dynamic Factor Models , 35(5), 317–360. https://doi.org/http://dx.doi.org/10.1108/MRR -09-2015 – 683
0216 684
Shrivastava, A., Kumar, K., & Kumar, N. (2018). Business distress prediction using the Bayesian 685
logistic model for Indian firms. Risks , 6(4). https://doi.org/10.3390/risks6040113 686
Šlefendorfas, G. (2016). Bankruptcy prediction model for private limited companies of Lithuania. 687
Ekonomika , 95(1), 134. https://doi.org/10.15388/ekon.2016.1.9910 688
Sun, J., Fujita, H., Chen, P., & Li, H. (2017). Dyna mic financial distress prediction with concept drift 689
based on time weighting combined with Adaboost support vector machine ensemble. 690
Knowledge -Based Systems , 120, 4–14. https://doi.org/10.1016/j.knosys.2016.12.019 691
Svabova, L., Durica, M., & Podhorska, I. (2018). Prediction of Default of Small Companies in the 692
Slovak Republic. Economics and Culture , 15(1), 88–95. https://doi.org/10.2478/jec -2018 -0010 693
Tobback, E., Bellotti, T., Moeyersoms, J., Stankova, M., & Martens, D. (2017). Bankruptcy prediction 694
for SMEs using relational data. Decision Support Systems , 102, 69–81. 695
https://doi.org/10.1016/j.dss.2017.07.004 696
Uthayakumar, J., & Vengattaraman, T. (2017). Swarm intelligence b ased classification rule induction 697
(CRI) framework for qualitative and quantitative approach: An application of bankruptcy 698
prediction and credit risk analysis. Journal of King Saud University – Computer and Information 699
Sciences . https://doi.org/10.1016/j.j ksuci.2017.10.007 700
Zhang, Y., Wang, S., & Ji, G. (2013). A rule -based model for bankruptcy prediction based on an 701
improved genetic ant colony algorithm. Mathematical Problems in Engineering , 2013 . 702
https://doi.org/10.1155/2013/753251 703
Zhou, L., Lai, K. K., & Yen, J. (2012). Empirical models based on features ranking techniques for 704
corporate financial distress prediction. Computers and Mathematics with Applications , 64(8), 2484– 705
2496. https://doi.org/10.1016/j.camwa.2012.06.003 706
Ziȩba, M., Tomczak, S. K., & Tomcz ak, J. M. (2016). Ensemble boosted trees with synthetic features 707
generation in application to bankruptcy prediction. Expert Systems with Applications , 58, 93–101. 708
https://doi.org/10.1016/j.eswa.2016.04.001 709
710
© 20 20 by the authors. Submitted for possible open access publication under the terms
and conditions of the Creative Commons Attribution (CC BY) license
(http://creativecommons.org/licenses/by/4.0/).
711