Applying Financial Distress Prediction Models in Saudi Arabia Businesses - Research Proposal Example

Applying Financial Distress Prediction Models in Saudi Arabia Businesses Introduction The financial impact of business failure is colossal, especially affecting the interest of the stakeholders of various publically traded companies. Research has indicated that prior to undergoing business failure the company’s financial position is commonly under distress. Thus, it is essential to identify these financial distresses at the early stages itself so that the stakeholders and investors interest could be safeguarded (Dothan 2006). The recent global recession has also exposed the vulnerability of various established companies throughout the world. The recession has even left a dent on the perceived stabilised economy and business of Saudi Arabia as well (Cerra et al 2009). Therefore, the proposed research focuses on using various statistical models to predict financial distress in Saudi Arabian companies, which can help in identifying such negative economic situation in the future and preventing them. In order to conduct this research, data from around 30 listed companies would be collected and analysed. Further, a structured questionnaire would also be designed to extract the necessary information from these companies. The collected data would be analysed using various statistical techniques such as Logistic regression, Linear Discriminant Analysis and Artificial Neural Network. Further, a model would be created to predict financial distress in these companies. The proposed model would be based on predicting distress on the basis of liquidity, debt, profitability and other such variables. The statistical results for this model would help in identifying and predicting financial distress in companies, which would help in controlling financial losses at the onset of financial troubles in a company. Aim An essential asset to any business, particularly those related to financial investment and lending, is a reliable and accurate Business Failure Prediction (BFP) model. Such models are of immense benefit, especially in times of financial instability where new business ventures can be precarious. Therefore, there has been a substantial increase in interest in the area of business failure prediction, from both industry and academia as it helps in recognising the benefits and associated value that these models can offer to all sectors concerned. In Saudi Arabia there is a growing need for an established and accurate business failure model especially for those related to family businesses, as this is the most common type of business found in the country. In the present global financial climate this need is emphasised by the current lack of literature and the continued emphasis for viable models that would be operable in the real world. This project would focus on developing such a business failure prediction model using the available resources that highlights the past and current research, as well as the key factors involved when investigating family businesses in Saudi Arabia. Thus, the project aims to: Develop a business failure model for Saudi Arabia business, and particular family-owned businesses Amend and develop the existing literature to make it suitable for businesses in Saudi Arabia with Islamic finance focus Investigate whether the companies based on Islamic financial system are more subject to financial distress in comparison with other companies based on conventional system Objectives It is clear from established research and statistical evidence that there is an emphasised need for a feasible and accurate business failure prediction model to be introduced into the market. This has been further highlighted by the extremely costly failures of high profile businesses like Merrill Lynch, Lehman Brothers, Washington Mutual and Wachovia during the 2008 credit crisis which costed these global banks over $300 billion (DePamphilis 2009). Experts also believe that the economic condition and the business health of Saudi Arabian companies are very fragile and unstructured. Further, around 95 per cent of the businesses in the country are family-owned which offers unique challenges and sustainable issues (Mahayny 2007). In the recent times, established Saudi-Arabia businesses such as Al-Jasim Engineering, Alyami Contracting and The Software House witnessed many difficulties (Bradley 2005). Further, numerous Saudi Arabia business also witnessed failures which led to significant commercial and social impacts. These failures have increased the interest in business failure prediction for both large and small businesses from both industry and academia (Bradley 2005). There is a visible lack of study and research into this particular niche area of the market where the limited amount and access to relevant information can seriously impact the development of such models. Therefore, our objective is to: Develop a model that is useful in the prediction of financial distress of Saudi Arabia business. In doing so it will be possible to: Determine variables that are important in explaining sources of financial distress of family operated Saudi Arabia business. Develop a decision tree to predict financial distress amongst these family operated Saudi Arabia business. Statement of significance In light of the many business ventures that have failed and the many bankruptcies there has been a substantial growth in individuals interests to carry out further research in this area of business failure prediction, from both the industry and academia. To be in a position to gain from current and past research carried out in Australia and overseas and make use of the data and the various tools could facilitate new and accurate business failure prediction models that will be rendered extremely valuable to both Saudi Arabia business operators and their advisers and financiers. The unique aspects of Saudi businesses most of which are family owned and operated requires the “business models” that are created to take into account that factor in identifying success factors, or alternatively in development of failure prediction models that need must necessarily incorporate factors that are uniquely associated with family operation. Accordingly ‘family associated factors’ will be added to those previously identified financial factors in developing our failure prediction model. This study may contribute significantly in the below mentioned areas: Financial entities such as banks, investment operators, credit unions, and other financial institutions should be able to realistically gauge the success and failure rate of those Saudi Arabian businesses they are providing credits. These financial entities may modify their lending policies to protect their interests in case of business failures and if the businesses are unable to repay the loan amount (Altman & Hotchkiss 2005). Loss of large sum of money could be avoided in case the investors are already aware of the actual financial health of a company. Such a financial model would help the investors to analyse the economic condition of the company and thereafter, make investment decisions (Brealey & Meyers 2000). Business relationships have always been a powerful tool and many Saudi Arab businesses engage and develop long-term relationships with other businesses (such as suppliers), which ensures a higher success rate in the future. This model would help in identifying this trend and establishing the credibility of a company based on its long-term relationships with other businesses (Damodaran 2002). Aside from the various industry sectors that will profit from accurate BFP models, individuals dealing with businesses have also potential to gain from using accurate BFP models in order to only deal with successful businesses. Furthermore, it is a realistic goal for businesses to use a BFP model to predict their own failure or survival, and utilising this information as a measure of financial health for management decision making can prove highly useful in making such predictions. From a common perspective, accurate BFP models will increase general confidence in investment, lending and the development of profitable Saudi Arabia business relationships. Thus, lending and investment in successful businesses will have the opportunity to increase, which will result in increased stable economic growth for the benefit of everyone involved (Brown & Kapadia 2007). Conceptual framework In order to conduct such a large-scale research, it is important to develop a comprehensive framework that would be used as a guide to prove the research hypothesis. The paper focuses on developing a business failure model for Saudi Arabia business, especially family-owned businesses. However, even before undertaking the research, it is imperative to understand the concept of business failure and how it impacts the financial health of a company together with the economic condition of a country. The research would also discuss the detail the business failures from countries such as Australia and the recent failures of US businesses during the 2008 recession. The report would thereafter narrow its focus on family-owned businesses and the major reasons for the failure of such businesses. It would emphasise on the family-owned businesses in Saudi Arabia. After establishing the importance of conducting business failure prediction, the proposed paper would develop a model based on various statistical tools and apply them to analyse the financial health of some selected Saudi Arabia-based companies. The statistical results from the proposed model would help in understanding the financial health of the companies and predicting the possibility of business failure. Further, in order to conduct such a comprehensive research, it is essential to undertake the following steps: Identifying the problem statement Defining the purpose of the research Conducting literature review Providing methodology Collecting data Analysing data Discussing results Literature review Due to financial distress and bankruptcy many investors, shareholders, creditors, vendors, clients and even employers suffer heavy losses (Giesecke 2005). The recent global recession has indeed exposed the vulnerability of many established companies and failure of some businesses such as Lehman Brothers in fact shocked the entire world (Reinhart & Rogoff 2009). Further, earlier failure of large businesses all over Europe and US such as Enron, Swissair and WorldCom has also dented investor confidence and most investors are wary of investing large sum of money into one organisation (White 2006). Such examples of large business failure has raised the issue that not just medium or small businesses are prone to failures, but even big enterprises are not protected from financial distress. It is therefore important to identify the various reasons of financial distress and develop a model to predict financial crisis in a given organisation. Various researchers have suggested different models to identify financial distresses in a company, which they also proved through empirical results. Early signs of distress can be identified by employing various statistical models (Kahl 2002). These models are developed as per the requirements of a particular era. For instance, in the 1960s, the researchers mostly used models that focused on financial ratios and their impact on financial distress (Pindado & Rodrigues 2005). While Beaver (1966) employed dichotomous classification testing for identifying the financial ratios that could predict business failure, Altman (1968) focused on multivariate statistical model. In the later years, Martin (1977) and Ohlson (1980) used logit model to predict bank failure. However, research indicated that these models were not perfect and the researchers might have used their own discrimination in these models (Morris 1977). Thus, in the recent few years with the development in technology, adoption of neural computing has gained ground in predicting business failure. These days, most researchers use neural computing technique to predict financial distress and bankruptcy (Zurada 1998). However, research on predicting financial distress is in a nascent stage and still requires considerable research in the field. The failure of big corporate in the recent times together with various scandals in the corporate world has made it imperative for investors to understand the financial distress in a company before investing (Clarke 2003). Further, it has also been noted that most of these models are for predicting financial distress in western companies. There are rarely any researches that focus on Middle Eastern economies such as Saudi Arabia. Saudi Arabia offers unique business environment as it is predominately captured by family-owned businesses. Such businesses are much more prone to failures due to their unstructured business format and lesser available public information. Thus, it is much more imperative to analyse and understand the financial health of a Saudi Arabia-based company before investing into it (Hyvari 2006). However, the initial research found that there are hardly any studies that focused on predicting financial distress in Saudi Arabia businesses. Thus, this research would focus on developing a model that would help in predicting financial distresses in Saudi Arabia companies. Methodology There are a number of different variables associated with this research and data, particularly on the various accounting sectors needs to be collected and collated so as to provide a viable model for Saudi Arabia businesses. This research requires data from listed companies. If necessary a suitable questionnaire may be designed to extract the necessary information. Data collection The process of data collection would begin with identifying various financial ratios which offer an overview of economic and structural characteristics of the Saudi Arabia business firms. Besides collecting data from the annual reports of the companies, other public sources such as the Saudi Arabia Stock Exchange should also be used. Saudi Arabia Stock Exchange (Tadawul) conducted a Business Longitudinal Survey (BLS) to follow the growth and performance of around 500 firms of all sizes over the five year period from 2004 to 2009. The available data (differentiated into family firm and non-family firm) on the basis of profit, margin, total assets (current and non-current), total liabilities (current and current), net worth, return on assets, return on net worth, long term debt to equity, current ratio, interest coverage, income, age of firm, sales of goods and services, industry gross product etc. is available for these firms and can enable further research into identifying key factors responsible (Fletcher & Goss, 1999). In this project a sample of firms from different sectors will be used some of which will have gained a success rating based on profit performance while others will be distressed either bankrupt or based on loss performance. After having completely investigated and systematically analysed the dataset to remove all erroneous values and other abnormalities, the dataset will then be classified into training and test dataset. Thereafter, a mixture of statistical techniques would be applied to develop a suitable model for business failure prediction (Tan 2001). Data analysis Various statistical techniques would be applied to develop a suitable model for business failure prediction. Two particular techniques, Logistic regression and Linear Discriminant Analysis, would be of prime importance for developing the model and determining identifiers for financial distress prediction. In addition to these Artificial Neural Network and the latest and most developed techniques such as decision tree will also be incorporated. Logistic regression: It is a variation of ordinary regression that is rendered useful when the observed outcome (e.g. level of financial distress) is restricted to two definite values (Kumar & Ganesalingam, 2001). This usually represents the occurrence or non-occurrence of some outcome event, i.e., 0 or 1. This allows the production of a specific formula that predicts the probability of the occurrence as a function of the independent variables. Therefore the logistic distribution constrains the estimated probabilities to lie between 0 and 1. Logistic regression models also highlight which variables are significant in explaining financial distress (Fletcher & Goss, 1993). Discriminant Analysis: This separates distinct sets of objects or observations, and then allocates new observations (or objects) to these defined groups (or sets). This can be further clarified by an example where two groups: distressed and non-distressed companies can be compared. In general, discriminant analysis is exploratory in nature, and is often used when a causal relationship is not completely understood allowing distinctive value to be given to the new interpretations (Sori & Jalil 2009). Neural Networks function: It has been implemented through an association with a back propagation algorithm and used in finance related applications for some time now. Neural Network uses the inputs and the target defined by the user to learn a pattern. The target is a 1 or 0, 1 for a well performing company and 0 for a distressed or bankrupt company (Habbershon &Williams, 1990). Through the collation of such data ANN builds relationships between the different inputs, and according to what the ANN has gained from its research it will then adjust weights to each significant ratio and provide the user output as close to the target value as possible. The benefits of using neural networks over other systems of research are that the neural network is very suited to dealing with large and noisy data sets. The Classification and Regression Tree (CART) algorithm is the dominant Recursive Partitioning Algorithm (RPA) that is one of the more useful tools in helping to predict financial distress. This particular algorithm generates a set of tree-based classification rules; that is, constructs a decision tree. When applied to BFP, CART results in a binary classification tree that assigns firms to groups. Reference Altman, E. “Financial Ratios, Discriminant Analysis and the Prediction of Corporate Bankruptcy.” The Journal of Finance 22 (1968): 589-609. Altman, E., Hotchkiss, E. “Corporate Financial Distress and Bankruptcy: Predict and Avoid Bankruptcy, Analyze and Invest in Distressed Debt.” 3rd Edition, New Jersey: John Wiley & Sons (2005). Beaver, W. Financial Ratios as Predictors of Failure. Journal of Accounting Research 5 (1966): 71-111. Bradley, J. R. "Saudi Arabia exposed: Inside a kingdom in crisis." New York: Palgrave Macmillan (2005). Brealey, R., Meyers, S. “Principles of Corporate Finance.” 6th Edition, McGraw-Hill, New York (2000). Brown, G., Kapadia, N. “Firm-Specific Risk and Equity Market Development.” Journal of Financial Economics 84 (2007), 358-388. Cerra, V., Panizza, U. and Saxena, S. “International Evidence on Recovery from Recessions.” IMF Working Paper 09/183 (2009). Washington: International Monetary Fund. Clarke, F., G. Dean and K. Oliver. “Corporate Collapse: Accounting, Regulatory and Ethical Failure”. Cambridge University Press. 2003. Damodaran, A. Investment Valuation. 2nd Edition, New York: Wiley Finance (2002). DePamphilis, M. “Mergers, Acquisitions, and Other Restructuring Activities.” 5th edition (2009). Dothan, M. “Costs of Financial Distress and Interest Coverage Ratios.” The Journal of Financial Research 29 (2006): 147-162. Fletcher, D and Goss , E. “Forecasting with neural Network: an application using bankruptcy data”. Information and Management 24 (1993): 159-167. Giesecke, K. “Default and Information.” Journal of Economic Dynamics and Control, 2005. Habbershon, Timothy G. and Williams Mary l. “A Resource-based Framework for Assessing the Strategic Advantage of Family Firms.” Family Business Review 12 (1999): 1-25. Hyvari, I. ‘‘Success of projects in different organizational conditions’’. Project Management Journal 37(2006): 31-41. Kahl, M. “Economic Distress, Financial Distress, and Dynamic Liquidation.” The Journal of Finance 57 (2002): 135-168. Kumar, K and Ganesalingam S. “Detection of Financial Distress using Multivariate Methods. Managerial Finance (MCB University Press)”. Special Issue edited by K. Kumar, Vol.27, No5, (2001): 45-55. Mahayny, K. “Difficulties Facing Family Owned Businesses in the Arab Region”. A paper presented to the conference “Business Development & Family Businesses: Managerial Foundations & International Accounting Standards”, Arab Tax Society, Cairo, Egypt (2007), 10-11 February, Nile Hilton. Martin, Daniel. “Early Warning of Bank Failure: A Logit Regression Approach”. Journal of Banking and Finance (1977): 249-276. Ohlson, J. “Financial Ratios and Probabilistic Prediction of Bankruptcy.” Journal of Accounting Research 18 (1980): 109-131. Pindado, J. and Rodrigues, L. “Determinants of Financial Distress Costs.” Financial Markets and Portfolio Management 19 (2005): 343-359. Reinhart, C.M. and Rogoff, K.S. “The Aftermath of Financial Crises”. American Economic Review (2009). American Economic Association, Pittsburgh, PA. Sori, Z. M. and Jalil, H. A. “Financial Ratios, Discriminant Analysis and the Prediction of Corporate Distress”. Journal of Money, Investment and Banking 2009: 5-15. Tan, C N W. “ANN application in financial Distress prediction and Foreign exchange trading”. Whilberto Publishing, Gold Coast, Australia. 2001. White, G. B. “Perceptions Of Accountants: What Are They After Enron And WorldCom?” Journal of College Teaching & Learning 3 (2006): 71-76. Zurada, J. “Neural Networks Versus Logit Regression Models for Predicting Financial Distress Response Variables.” The Journal of Applied business Research 15(1998): 21-28. Read More