FINAN indicative ratingFINAN indicative ratingFINAN indicative ratingFINAN indicative rating

(1)

FINAN indicative rating

Default prediction rating and organizational performance scorecard

(2)

FINAN indicative rating

Default prediction rating and organizational performance scorecard

Johnny Hillebrand Zwart Student number: 1587633

14 May, 2008

University of Groningen

Master program: Business Administration Specialization: Business & ICT

Supervisors:

drs. E.J. Stokking Faculty of Economics and Business

(3)

Preface

This master thesis is the result of my time working at FINAN financial analysis. The topic of this thesis is the FINAN indicative rating, which has been developed by FINAN financial analysis with the purpose to be included in their commercially available software products.

(4)

Management Summary

The topic of this thesis is the “FINAN indicative rating” and organizational performance scorecard. The idea of developing these tools have come from FINAN financial analysis (from this point referred to as “FINAN” ). FINAN has recognized the wishes of their clients (specifically accounts and

business advisors) to obtain a tool which can support them to determine the creditworthiness of an organization and to guide them in making (financial) decisions to improve an organization’s creditworthiness. FINAN assumes that these tools are marketable and it is this assumption that has lead to the topic of this research. The tools are aimed at business advisors and accountants which use or could potentially use the FINAN financial analyzer software to provide services regarding

creditworthiness of Small to Medium Sized Enterprises (SMEs). This thesis will answer the main research question: “What would an indicative rating to assess creditworthiness of a Dutch SME organizations look like?“. This thesis will describe the development of the FINAN indicative rating and the actual models included in the FINAN indicative rating.

One definition of creditworthiness is: “A creditor's measure of an individual's or company's ability to

meet debt obligations” (investorwords.com, 2007). Creditworthiness therefore can be summarized as an organization’s ability (or intention) to repay debts. This definition however does not make it clear why it is important to consider creditworthiness in the first place.

The reason why creditworthiness is (increasingly becoming) important has to do with the fact that banks play an important role in the funding of organizations. In fact: banks are the most important source of funding for SMEs. Recently changes have occurred in the bank’s lending business. The biggest change has come from the Basle committee which in 1999 released a report on “capital adequacy of financial institutions”. The previous Basel I standards (Basel Committee on Banking Supervision, 1988) required a bank to maintain a 8% equity position for their risky assets. No distinction was made between the different degrees of credit risk of these assets. In the new Basle II standards the assets held by a bank should match the implied credit risk . The following question then arises: how can this implied credit risk be measured?

(5)

risk segments. The result of the previous step determines the expected probability of default of an organization. The probability of default estimates the chance an organization will not (be able to) repay its debts. External ratings: The rating process of rating agencies include quantitative, qualitative and legal analyses. The first step for an organization to receive a rating is to request a rating. After receiving a request for a rating the rating agency will assign an analytical team. The first step for the rating agency is to review public information and internal documents of the organization in order to assess projected performance. The next step is to meet the (debt) issuer: in a direct meeting with the debt issuer the credit rating agency is able to collect more detailed information than from going through information made public by the organization. Finally the rating committee reviews the provided information. The committee will make a recommendation for the rating. The organization that requested the rating can appeal the rating before it is made public and can supply new information for reassessment of the organization. The next step for the rating agency is to issue the rating. At the same time the assigned rating is announced, the rating agencies add the issue to their surveillance system. Usually these ratings are reviewed once every year.

Now that the credit rating processes of banks has been highlighted it is time to return to the subject of the FINAN indicative rating. The FINAN indicative rating will support business advisors and

accountants in their advice giving processes regarding creditworthiness of SMEs. In order for the FINAN indicative rating to be effective, the tool has to align with the working procedures and the general thought process of a business advisor or accountant. One criteria following from objective is that quantitative and qualitative indicators/risk drivers have to be included which are commonly used by business advisors and accountants or which can easily be interpreted and understood. The FINAN indicative rating for this reason has to make assumptions about what an average bank considers to be creditworthiness and how this translates to indicators which are used and/or understood by business advisors and accountants. Fortunately the European Commission (2005) shows that the (seven) most important financial factors used by banks for credit rating purposes will not change must after implementation of Basel II. According to this research it is not the factors that change but the way these factors are interpreted. We can therefore assume that the average (European) bank will roughly look at the same elements as their competitors do. The steps taken to develop the FINAN indicative rating are explained in chapter 3 (most of which are the results of the efforts of Mr. Ran Chen working at FINAN). These steps include: (Pre-) selection of potential performance indicators, data gathering and extraction, data cleansing, Significance tests, correlation tests, selection of risk drivers, estimating a temporary model, re-estimating using AIC and model validation.

(6)

does not stop here. At this point the development continuous in the form of the FINAN score model.

The development of the FINAN score model includes the following steps which are explained in chapter 3: data gathering, data cleansing, selection of potential performance indicators, selecting industry groups, determining cut-off points, determining weights of performance indicators and analyzing the differences between both models.

The FINAN rating model and the FINAN score model are combined into one tool: The FINAN indicative rating. By combining both the rating model and the score model the advantages of both models are combined. The FINAN rating model has a high dicriminary power to determine the probability of default. This means that the model is good at dividing defaulting organizations from the normal performing organizations. The FINAN score model on the other hand has the ability to give the user insight in the scoring process. Effects of decisions are easier to understand and interpret. The following conclusions can be drawn from this research:

- Monoticity between the FINAN score model and the FINAN rating model is maintained. Also in most cases the prediction in rating categories is no more than one. However rating accuracy can be improved by developing a FINAN rating model which includes models for the different industry groups identified. The FINAN rating model and score model will then predict the same rating category in more instances.

- Users have expressed the wish to see the performance of an individual indicator compared to its industry. This could either be by displaying the percentage of organizations scoring higher or lower on the specific indicator or adding a “bar” which indicates the relative position of the performance indicator on it.

- If the FINAN rating is improved by adding more models for different industry groups the FINAN score model itself can be improved by selecting performance indicators which are better fit for assessing creditworthiness of a specific industry group. If that is the case every industry group would have its own FINAN rating and FINAN score model.

- In the extraction process the BIK codes of records were altered (I assume in the extraction itself, could also be caused by opening the document in excel). The result was that the number 0151 would be turned into the same number as 151. The first however represent industry 01 the latter 15. They are not the same industry and resulted in much manual labor which in affect could result in making mistakes in the extraction.

- Erroneous data would show up after every extraction and requires data cleansing. Data cleaning is time consuming and error-prone.

(7)

1 Problem statement

This chapter provides an introduction to the remaining chapters of this thesis. In this chapter the research set-up and the research topic will be discussed.

Problem statement

FINAN financial analysis has the intention to develop a tool which will be know as the FINAN

indicative rating. This tool will be aimed at business advisors and accountants who provide services to small to medium sized enterprises (SMEs) regarding creditworthiness. FINAN has recognized the wishes and requirements of these business advisors to obtain a tool which can support them in these advice giving processes. It is FINAN financial analysis’s assumption that this tool can be integrated in their commercially available software and that it can be successfully marketed as such.

One definition of creditworthiness is: “A creditor's measure of an individual's or company's ability to

meet debt obligations” (investorwords.com, 2007). Creditworthiness becomes an issue when a SME (debtor) wishes to receive a (additional) loan from (in most cases) a bank. Banks have their own (internal or external) credit rating systems to assess the creditworthiness of clients. These rating systems (and in most cases also the issued ratings) are considered inside information of the bank. The information concerning the rating process of banks is usually not shared with the rated organizations and the outside world. There are limited (perhaps no) tools available which can be used to assess the creditworthiness of a SME. Although specialized rating agencies have existed for decades, their services usually only apply to large organizations.

Having a good creditworthiness will increasingly become important for SMEs because banks now have the obligation to assess the risks of their financing activities. It is to be expected that SMEs will more frequently ask for advice from their business advisor or accountant. It is for this reason that the FINAN indicative rating will be aimed at these business advisors and accountants.

This thesis will describe the development process of the FINAN indicative rating. The main research question can be summarized as: “What would an indicative rating to assess creditworthiness of a

(10)

The main research question refers to the tool FINAN intends to develop: The FINAN indicative rating. This tool should give an indication of the creditworthiness of an organization. The tool will be used by business advisors and accountants in their advice giving process. The tool will help the business advisor or accountant to “rate” and assess the creditworthiness of the client. Rating the client will help the client to understand the rating process of a bank and to gain insight in its creditworthiness. The word “indicative” is used to indicate that the rating is not an official credit rating but an indication of an average bank’s opinion of creditworthiness. To understand this opinion of an average bank the first research question following this main research question is “what are credit ratings”. Sub questions of this research question are based on the fact that a bank uses a statistical algorithm to calculate the loss given default (Basle II requirement). The terms “probability of default” and “exposure at default” are linked to calculating the Loss given default. This explains the third and fourth sub question. Credit ratings are expressed as a “rating category” and therefore a question has been added concerning the rating scales used for credit ratings. The last two sub questions deal with performance indicators and risk drivers. A distinction has been made in this research between these two definitions.

1 What are (credit) ratings ?

- What is a (credit) rating ?

- What is “probability of default”(POD/PD)? - What is the “Loss Given Default”? (LGD)? - What is the “Exposure at default” (EAD)? - How are (credit) rating scales usually defined ?

- How are credit rating scales linked with probability of default ? - What are risk drivers ?

- What are performance indicators and how do they differ from risk drivers?

The FINAN indicative rating is composed of two models: The FINAN rating model and the FINAN score model. The second research question concerns the development of the FINAN rating model. It will answer how the model is constructed and how it has been developed. The development involves selecting (potential) risk drivers which explains the question concerning potentially interesting risk drivers. The last sub question will answer how the actual model will be developed and validated.

2 How is the FINAN rating model composed? - How is the FINAN rating model constructed?

- What risk drivers are relevant for the FINAN rating model?

- What is the meaning and possible added value of these risk drivers? - How will risk drivers be selected?

- Which risk drivers will be selected for the FINAN rating model? - How will the default prediction model be developed?

(11)

The third research question concerns the second model of the FINAN indicative rating: the FINAN score model. The FINAN score model uses a scorecard model as a basis for development. This question will answer how potential performance indicators are selected. The most important sub questions will answer how performance indicators will be translated to scores and how these scores are translated to a weighted average score. This results in the following research question and sub questions.

3 How is the FINAN score model composed?

- How is the FINAN score model constructed?

- What performance indicators are relevant for the FINAN score model? - How will performance indicators be selected?

- How are performance indicators translated to scores?

- How will performance indicators be weighted and translated to a final score? - What do the score and the different sub-scores mean?

- What diagnoses and advises can me made based on these scores?

- What opportunities do Small to medium sized enterprises have to improve their rating?

Once both models have been developed they have to be integrated into one model in the FINAN financial analyzer software. This fourth research question will answer how this can be done and what the final model will look like. It will also review the demands of users and usability.

4 How can the FINAN indicative rating be imbedded in FINAN financial analyzer ?

- Who should use the FINAN indicative rating model? - What information should be displayed to users? - What is the best way to present information to users?

- What is, from a marketing standpoint, the best way to implement the rating in FINAN? - How often should the FINAN indicative rating be updated ?

The last sub question has been added to answer the question concerning the role of an indicative rating in an advise giving process. To answer this question the role of an accountant or business advisor in the credit rating process of a bank has to be reviewed. This advice giving process will give insight in the usability of the FINAN indicative rating.

5 What is the role of the FINAN indicative rating in an advise giving process?

- What does the advise giving process, of a business advisor, look like ?

- What would be the added value of the FINAN indicative rating in this process?

(12)

Applied research methods

Three research methods have been applied to construct the FINAN indicative rating (and the scorecard) and form the basis of this thesis.

Literature research: Recent literature has been reviewed to answer questions as “what is a credit rating?” and “what is the probability of default? “. Chapter three is the main result of this literature review. The chapter defines what credit ratings are, how they are used at banks and what their main (common) elements are.

Data analysis: Databases containing information regarding Dutch SME organizations have been referred to in order develop the FINAN indicative rating. Data has been extracted, filtered and analyzed to develop these tools. The results of this can be found mainly in chapter four which describes the development of the FINAN indicative rating and the performance scorecard.

The goal of this data analysis is to develop a model. To be specific this research has mainly focused on developing certain aspects of the FINAN score model. The FINAN rating model has been used as an input for this development (Although: help has been given to select potential risk drivers for the FINAN rating model, data has been cleansed and results have been reviewed). For this reason the development of the FINAN rating will be described in this thesis. The main elements of this research have been: to determine industry groups, to determine cut-off points for the scoring of performance indicators, determining the weights for the weighted average score and determining advice and diagnosing “rules” which are part of the FINAN score model. Chapter three will describe the development of the FINAN rating model and the development of the FINAN score model.

(13)

2 FINAN financial analysis background

This chapter will provide the reader with a short introduction of the organization where the development of the FINAN indicative rating took place and the program in which it has been integrated. FINAN financial analysis is the organization behind the FINAN indicative rating. The main product of FINAN financial analysis is the FINAN financial analyzer application.

2.1 FINAN Financial analysis

The main product of FINAN financial analysis is the software application called FINAN financial analyzer. This software application is available in the following versions: advice, advice plus, valuation and valuation plus. There also is a special version of the FINAN financial analyzer for banks.

Figure 1 Overview of the FINAN financial analyzer application

(14)

easily create forecasts and different scenarios. These scenarios and other information can be used to form and print a report which can be provided to clients or other stakeholders.

The valuation and valuation plus versions have specific functionality for a business advisor or

accountant to calculate the value of an organization. The application includes different methodologies to calculate the value of an organization.

The banking version of FINAN financial analyzer usually comes equipped with custom build models. These models are built to the specifications a bank supplies. The basis of these models is formed by the standard banking model of FINAN financial analysis. Banks use FINAN financial analyzer to analyze the financial position of their clients. The bank’s analysis usually includes the bank’s own rating models which can be integrated in the custom FINAN banking model. The FINAN financial analyzer is also used to store the information into the database of the bank.

(15)

2.2 FINAN financial analyzer and the FINAN indicative rating

The FINAN financial analyzer is a software application used by accounts and business advisors to make financial analyses. The application requires the user to input historical (financial) information. The application allows the user to make forecasts and construct different scenarios. The FINAN indicative rating uses the (financial) information and the scenarios to calculate the expected probability of default and to assigns scores to performance indicators of all of the historical and forecasted years. The FINAN indicative rating is a combination of the FINAN rating model and the FINAN score model. The FINAN indicative rating will be described in this paragraph.

The indicative rating requires additional information concerning the industry the organization is operating in. The industry of the organization has to be selected from a list including the name and

“BIK-code (Dutch chamber of commerce industry classifications)” of every type of industry. In the example below the industry for a wholesaler of food products has been chosen. Based on this

information industry group 4 (Retail and wholesale) has automatically been selected. If preferable the industry group can be overridden manually.

(16)

Six quantitative performance indicators and one qualitative performance indicator are “graded” in the FINAN score model. The assigned scores to the quantitative performance indicators are dependent of the selected industry group. The quantitative performance indicators that receive a score are: solvency, liquidity, profitability, sales, paying behavior and interest coverage. The information required to assign scores to the quantitative performance indicators comes from the profit and loss account and the balance sheet. The performance indicators receive a score for both historical years and forecasted years for every created scenario. The qualitative performance indicator requires the user to make an objective assessment of several qualitative variables. The weighted average score of this assessment is used to form the score of the qualitative performance indicator. The weighted average score of the seven performance indicators determines the final score (The FINAN score). The assigned scores for both the individual performance indicators as the final score range from 1-10. Based on the scores given to individual performance indicators the program automatically generates advices. The generated advices can be reviewed in the folder “advices”.

(17)

3 Theoretical background

This chapter will provide a theoretical background on the subject of (credit) ratings. This chapter will review the following subjects: internal- versus external ratings, rating scales, the probability of default and the rating principles of Krahnen and Weber (2001).

3.1 Credit rating systems

Banks play an important role in the funding of organizations, according to the European Commission (European Commission, 2005) the most important sources of funding for small and medium-sized enterprises (SMEs) are banks. The EC has identified changes in the banks’ lending business which affect the way banks are conducting business with these SMEs. The EC refers to these changes as “a

significant transformation”.

The biggest change identified by the EC has come from the Basle committee which in 1999 released a report on capital adequacy of financial institutions. The report proposed changes to the international standards on capital adequacy. After this initial release the committee has released additional

proposals in the following years. The proposals have been revised during the Basel II process based on responses from financial markets and other stakeholders. The previous Basel standards (Basel I) required a bank to maintain a 8% equity position for their risky assets. No distinction was made between the different degrees of credit risk of these assets (Basel Committee on Banking Supervision, 1988).

In the new Basle II standards (Basel Committee on Banking Supervision, 2005) the assets held by a bank should match the implied credit risk (European Commission, 2005). To estimate the credit risk of new and outstanding loans a bank basically has two options: the first option is to use information provided by rating agencies and another option is to develop an internal rating model. Krahnen and Weber (2001) mention that the rating of borrowers is a widespread practice in capital markets; a

rating summarizes the quality of a debtor and informs about repayment prospects. According to the European Commision (2005) the rating is an assessment of a specific debtor’s creditworthiness.

(18)

internal risk assessment and external rating refers to a credit supplier that uses external risk

assessments (Cantor, 2001). Another key difference between external and internal ratings systems is the fact that external ratings are usually published to be used by other stakeholders free of charge. Internal ratings on the other hand often are not even disclosed to the rated organization.

3.1.1 Internal credit rating systems

The workings of an internally developed rating model are summarized in the European Commission’s report (2005). This paragraph provides an overview of the European Commission’s description of an internal rating system.

In the first step of the rating process a bank gathers information about potential clients. Usually the information comes from the borrower and only in a few cases the information is collected externally. Borrowers are required to provide the bank with detailed information about certain aspects of their organization. The bank will use this information together with historically gathered information about defaults to determine the default risk of the organization.

The information gathered by a bank includes both quantitative and qualitative information. The information of an SME is usually gathered from financial statements or annual reports. The most important quantitative factors for the bank are financial figures and ratios. The importance (the weight) of qualitative factors is very dependent on the size of the organization and the size of the desired loan. In general banks will request more information from larger firms or when an organization requests a relatively large loan.

Figure 3 How borrowers are rated, European Commission (2005)

(19)

The result of the previous step determines the expected probability of default of an organization. The probability of default estimates the chance an organization will not (be able to) repay its debts.

3.1.2 Internal risk rating process

Crouhy et al (2001) show how a typical internal rating system at a bank could be organized. The authors’ propose an (prototype) internal risk rating model which will be discussed in this paragraph. The authors suggest using a two-tier system which is firstly used to establish an obligator rating

“which can be mapped to a default probability bucket”. Secondly a facility rating should be calculated that determines the loss given default.

The first step, suggested by Crouhy et al (2001), is to make a financial assessment. In this step a credit analyst will analyze (financial) statements of the organization and assess the financial health position of the organization. Amongst others the credit analyst will look for: trends of financial indicators, the quality of assets, cash reserves of the organization, and leverage. In this step a distinction is made between three areas of interest: 1) earnings and cashflow, 2) asset values, liquidity and leverage and 3) financial size, flexibility and debt capacity. For each of these three areas a separate rating should be calculated and by combining these an initial rating should be calculated. In some cases (industries) the three main assessment areas should be weighted more heavily or lightly. If this is the case, the

assessment should be benchmarked against other companies from the same types of industry.

The second step considers the impact of management and other qualitative factors. The impact of qualitative factors on the obligator rating can either be negative or positive. Qualitative factors are usually more difficult to measure. Usually they are not in the form of a ratio or a numeric value with meaning. The qualitative factors require more expertise and knowledge from the credit risk assessor in order for the credit analyst to perform an objective analysis. Often the credit analyst will have to request additional information because the necessary information cannot be found in the financial report.

(20)

Step 3b involves making a tier assessment. The tier assessment determines the relative position of an organization within its own industry. According to the authors the relative position of an organization within its own industry is an important survival factor, particularly during downturns. The same criteria used to make an industry assessment could also be used for the tier assessment. The authors make a distinction between four tier groups: Organizations in tier 1 are major organizations with a dominant share of the relevant market, tier 2 organizations are important or above average

organizations, tier 3 are average organizations and tier 4 organizations are weak industry organizations and have a declining customer database. An organization should be ranked against its relative

competition (On either a global or a local scale).

Step 3c combines the assessment of the health of an organization and the position of a business within its industry (industry / tier position). By combining the results from step a and b the vulnerability of a company can be assessed. The authors suggest using a table to determine the “best possible” obligator rating. The table gives a maximum OR rating for every industry/tier rating combination. The table acts as a cap on the OR rating. For example the obligator rating of well performing organizations in an weak industry will be capped to the maximum OR rating for that industry.

The fourth step deals with the quality of financial information. According to the authors this step should not be used to improve the obligator rating but should only be used to determine the maximum possible obligator rating. The quality of the financial statements is largely determined by the

capabilities (and size) of the composing organization (Accounting firm / administrative office). It is important that these capabilities are aligned with the size and complexity of the borrower.

The fifth step consists of making a country risk assessment. According to the authors: country risk exists when more than a prescribed percentage of the obligor’s cash flow or assets are located outside of the local market. Usually a bank will posses information about macro- and microeconomic events in different countries in order to assess the country risk. Step 5 also limits the “best possible” obligor rating.

The sixth step is the first step to determine adjustments to the organization’s facility rating. Step 6 looks at third-party support for the adjustment of the facility rating. When it is clear that a third party or the owner has provided (financial) support and is willing to do so in the future, the facility rating should be improved. Based on the quality of the support, the facility rating of the organization can be upgraded or downgraded.

(21)

rating from step 6 with the remaining “term to maturity” in order to determine the adjustment to the facility rating. The term to maturity is the remaining life of a financial instrument.

The eighth step examines how well the facility is structured. In this step aspects such as the

“covenants” and conditions of the facility are reviewed. The lending purposes and/or structure may influence the strength and quality of the credit. The effect of this step on the (adjusted) facility rating can be positive or negative.

Step 9 recognizes that the presence of security (Collateral) should (heavily) affect the severity of the loss given default of a facility. The quality and depth of the security determine the extent of reducing any loss. Securities should be valued against their liquidation value. Securities can be devised into several categories. If the organization possesses securities from several categories than usually the worst category containing a security with significant reliance should be chosen. The collateral category should only reflect the securities that are being held for the facility that is being rated.

3.1.3 External credit rating systems

Crouhy et al (2001) review the definitions of credit ratings used by two major (worldwide) credit rating agencies. Standard & Poor’s (S&P) describe a credit rating as: “A credit rating is S&P’s

opinion of the general creditworthiness of an obligor, or the creditworthiness of an obligor with respect to a particular debt security or other financial obligation, based on relevant risk factors.'' Moody’s give the following description of a credit rating: “...an opinion on the future ability and legal

obligation of an issuer to make timely payments of principal and interest on a special fixed income security.” The rating process of these rating agencies includes quantitative, qualitative and legal analyses. The quantitative analyses are focused on the financial analyses of an organization and are based on an organization’s financial reports. The qualitative analyses focuses on aspects such as the quality of management, the firm’s competiveness within its industry, expected growth of the industry, the vulnerability (of the industry) to technical changes, regulatory changes and the labor relations.

According to Crouhy et al (2001) S&P and Moody’s are regarded as unbiased evaluators. However, according to Cantor (2001) the role of the rating agencies are inherently controversial. Cantor continuous by mentioning that “It is the rating agency’s task to make independent and sometimes

(22)

3.1.4 External credit rating process

Crouhy et al (2001) present the rating process of Standard & Poors (S&P) and note that the rating process of other rating agency’s are very similar to that of S&P. This paragraph will summarize the description of this external rating process.

The first step for an organization to receive a rating is to request a rating. Usually the request is placed when an organization wishes to issue debt (usually bonds). Raeburn (2003) of the Association of Corporate Treasurers mentions a difference between solicited and unsolicited ratings, which is not mentioned by Crouhy et al. An unsolicited rating is a purely statistically rating based on published information and usually is not paid for by the rated organization. Solicited ratings on the other hand are initiated on the request of an (debt) issuer. The solicited ratings are published after (confidential) discussions with an organization’s management and the organization is able to hand over confidential and unpublished information to the rating agency. Solicited ratings are almost entirely paid for by (debt) issuers and used by other stakeholders. (Raeburn, 2007 ). An unsolicited rating will in some cases be turned into a solicited rating on the request of an organization’s management. The reason for requesting a solicited rating usually is a potential higher rating. Based on sample ratings from S&P, Poon (2003) concludes: there is a significant difference in distributions between solicited and unsolicited ratings. These lower ratings are commonly explained by the fact that less information is available to the rating organization in the case of unsolicited ratings. The rating agency will assign a lower rating for the lack of information. If that is the case an organization can increase its rating by providing more detailed information to the rating agency and request a (solicited) rating. The rating agency has the opportunity to improve its analysis with the help of (additional) information concerning the quality of management and other qualitative variables and will issue a new rating. For this reason the established major credit rating agencies have a very strong position and have an equally strong reputation of being a providing source of information for credit suppliers. Lending rates in these financial markets are very dependent on the rating of the organization issuing the debt. It is therefore not very surprisingly that Fitch Ratings (another major credit rating agency ) wrote in an published statement that 95% of the companies and financial institution that they rate requested their rating and agreed to pay Fitch’s fee even though the entity would have almost always already be rated by both Moody’s and S&P (Fitch Ratings, 2002).

(23)

After reviewing public information the fourth step is to meet the (debt) issuer. In a direct meeting with the debt issuer the credit rating agency is able to collect more detailed information than from going through information made public by the organization. Boot (2006) states that there is an information-asymmetry between borrowers and lenders. Managers of an organization usually have more reliable information about their organization. However, they find it difficult to transfer this information to the market in a way that it is considered trustworthy by lenders. According to Boot a possible a possible function of a CRA could be its ability to screen an organization and deliver information that is considered trustworthy by lenders. Boot concludes however that the results of research trying to confirm this standpoint provide no clear evidence to support this standpoint.

The fifth step involves a rating committee meeting to review the provided information. The committee has relevant expertise related to the type of industry of the organization. The committee will make a recommendation for the rating. The organization that requested the rating can appeal the rating before it is made public and can supply new information for reassessment of the organization. However, Kliger and Sarig (2000) note that this is not standard practice amongst all rating agency’. For instance: Moody’s policy is to simultaneously announce a final rating to the issuer and to the public.

The final step for the rating agency is to issue the rating. This means the rating is made public and becomes available for lenders and other stakeholders At the same time the assigned rating is announced, the rating agencies add the issue to their surveillance system (Kliger and Sarig, 2000). Usually ratings are reviewed once every year. Inputs for these reviews are new financial reports, new information from the organization’s management and other business information.

Figure 4 summarizes all the steps of the external rating process.

(24)

3.2 Rating scales and the probability of default

Ratings issued by credit rating agencies are usually expressed as a letter or a combination of letters (AAA, BBB, etc). Organizations with a good credit rating (BB and above) have a low(er) probability of default than organizations with an average (BB) or bad credit ring (BBB and lower).

Krahnen and Weber (2001) define a rating as the mapping of the expected probability of default into a discrete number of quality classes, or rating categories. The expected probability of default (PD) is a continuous variable between zero and one. This expected PD indicates the expected chance that principle or interest due will not be (re)paid in time. The expected probability of default is one component of the expected loss. The other component of the expected loss is the loss given default (LGD).

According to Cantor (2001) there is no single, universally accepted, quantitative standard for each credit rating category. Cantor continuous by mentioning that there is no expected loss benchmark that is likely to be accepted by all rating agencies. However, most rating agencies do use similar rating symbols. These rating symbols are in many cases based on the rating symbols used by moody’s or Standard & Poor’s (S&P). Moody’s uses the following 10 symbols: Aaa, Aa, A, Baa, Ba, B, Caa, Ca, C, and D and S&P & Fitch ratings use: AAA, AA, A, BBB, BB, B, CCC, CC, C, and D. Another popular format of presenting rating quality is the “traffic light model”. In a traffic light model the “traffic light” is either green or red. Green logically refers to having good credit quality and red to having bad credit quality. An optional third category is an orange “light” indicating doubtful or uncertain creditworthiness.

Ratings compress a great deal of information into one symbol (Cantor, 2001), they provide financial markets the ability to compare “credit quality” on a global scale. An important aspect of ratings is that they contain only a limited number of categories. Organizations and/or bonds with the same rating do not necessarily have the exact same chance of defaulting. Kliger and Sarig (2000) summarize this by stating: equally rated bonds are not claimed to be of identical quality and ratings cannot be inverted

into unique default probabilities. This implies that the degree of precision is determined by the number of rating categories and the range they represent. To represent improvements in default rating

(25)

categories. An assigned rating expresses the bandwidth of the probability of default. In some cases the probability of default is defined as the average of a rating category. Even more importantly to the credit issuer is the loss given default. According to the European Commission (2005); banks use ratings as the main input for calculating the expected loss implied by a given loan. The rating is a key indicator of the cost a bank incurs for a given loan (expected loss). Schuermann (2001) gives an overview of the loss given default and expected loss in his 2001 paper for The Wharton Financial Institutions Center. The expected loss (EL) is defined as the Probability of default (PD) x the loss given default (LGD) x exposure at default (EAD) resulting in the following formula:

EL = PD x LGD x EAD

PD: The probability of default: is the chance an organization will default within a certain time frame

(usually within a one year time frame).This chance is expressed as a percentage. For example: a 6% chance the organization will default within a one year time frame.

LGD: The loss given default is expressed as a percentage of the exposure at default. The loss given

default is the percentage of the expected exposure that cannot be retrieved when the organization defaults. For example 50% of the exposure could be covered by collateral.

EAD: The exposure at default is the amount owed to the credit issuer at the time a default occurs.

During the lifetime of the issued credit the obligator repays the debt. This means that during the lifetime the total exposure changes and therefore is not equal to the amount originally issued. To finish the example lets assume that at the start of year 1 € 50.000 Euro has been borrowed and at the start of year 2 € 10.000 will be repaid.

In the example the expected loss at the start of year one equals 0.06 (PD) x 0.5 (LGD) x

€ 50,000 (EAD) = € 1,500.00. The expected loss at the start of year two equals 0.06 x 0.5 x (50,000 – 10,000) = € 1,200. Under normal circumstances the expected loss would gradually deteriorate during the lifetime of the loan.

(26)

3.3 Risk drivers and performance indicators

This paragraph will discuss the terms risk driver and performance indicator which will be regularly used in the following chapters and paragraphs. The main difference between these two definitions has to do with their application. The term risk driver or risk factor is used when an risk driver/factor refers to factors which highly correlate with defaults. The term performance indicator is used when the performance of an organization (regarding creditworthiness) is mentioned.

Credit Risk drivers: The term credit risk drivers (factors) will be used when referring to statistically proven variables that determine, or at least influence, the probability of default. As explained earlier, variables determined this way do not necessarily have to be supported by theories that explain their “high discriminary power”. This term will therefore be used when referring to variables of the FINAN rating model.

Performance indicator: The term performance indicator is borrowed from the balanced score card system which was originally developed by Kaplan and Norton in 1992. The balanced score card (BSC) was developed to combine financial with non-financial measure in one performance measuring tool. The balanced score card is used by organizations/managers to measure and track key performance indicators in order to align them with the organization’s strategy. The term “performance indicator” is used here when referring to variables of the FINAN score model.

3.3.1 Common quantitative and qualitative risk drivers used by banks

According to Grunert et al (2005) factors commonly used for predicting corporate bankruptcy with financial factors concern the capital structure, profitability, and liquidity of an organization. Due to their “high discriminary power“ the models which include these factors are widely accepted. Grunert, Norden and Weber conclude that only a few of these models are based on theory that explains why and how certain financial factors are linked to corporate bankruptcy. The models are further limited by the fact that the financial factors are usually backward looking measures and because it is usually not clear how well these models perform with samples the model was not designed for.

The European commission ( 2005) states that 75% of the mid-sized and large banks consider

indebtedness of firms to be of high or very high importance as a rating input factor (of SMEs). 50% of these banks also place the same importance on liquidity and profitability.

(27)

qualitative factors (leverage, profitability and liquidity ratios, management experience, industry perspectives) However, the weighted schemes of these risk factors differ considerably across banks. The usage of both quantitative and qualitative risk factors is further supported by Grunert et al (2005) whom state that the combined use of financial and non-financial factors lead to a significantly more accurate default prediction than the single use of financial or non-financial factors.

For start-up companies, the information collected by a bank to process a credit request is substantially different from that of an existing company. (European Commission, 2005) Because start-ups cannot provide historical financial data, qualitative input factors account for about 60% of the rating. In such cases: business plans, management credentials and the degree of management’s financial expertise are viewed as key information. Due to the fact that financial (quantitative) information is hardly available for these start up organizations banks have to rely more on qualitative information.

According to Crouhy et al, the assessment of management (although subjective in nature) investigates how likely it is that management will achieve operational success, and its risk tolerance. The rating process includes meetings with the management of the issuer to review operating and financial plans, policies and strategies

The EC states that following the logic of prudent risk management many banks are inclined to assume the worst if any information is missing regarding a borrower’s current situation. Information provided to banks should be delivered timely and error free, otherwise this might have a negative effect on the rating. Providing poor quality information or not providing the information timely might be seen as a “warning signal” by the bank. According to the EC: the collection of qualitative information at banks often involves face to face meetings where banks try to confirm that an organization is well managed. Approximately 50% of banks that participated in the European Commission’s survey emphasized the high importance of (an SME’s) management quality as a rating input factor.

According to Crouhy et al (2001) the Industry and the relative position of a borrower within its industry has a strong effect on the organizations position. Organizations that perform weak in

vulnerable industries are major contributors to credit losses. The authors give the following examples which can be used to assign an industry rating: competiveness, the trade environment, the regulatory framework, restructuring, technical change, financial performance, long term trends, and vulnerability to macroeconomic events.

(28)

received group support in the past and whether or not it is likely the organization will receive further group support in the future.

Crouhy et al (2001) also state that the financial results and key ratios of companies in cyclical industries should be adjusted. This means that in very positive economic periods the performance of an organization should be adjusted downward and vice versa in very weak economic periods.

The European Commission (2005) shows that the (seven) most important financial factors used by banks for credit rating purposes will not change must after implementation of Basel II. According to this research it is not the factors that change but the way these factors are interpreted. Statistical techniques will be used which convert quantitative information into default probabilities. According to this research the seven most important quantitative factors in order of importance (most weight) are: leverage, profitability, private means, liquidity, Efficiency, company size and growth. This same survey of the European Commission shows that three important qualitative factors are: management quality, market situation and legal form.

3.3.2 Qualitative performance indicators

This paragraph will provide additional qualitative performance indicators recognized by several studies on the subject of creditworthiness and bankruptcy. Qualitative performance indicators as opposed to qualitative risk drivers can be used to “steer” the organization. The management of an organization uses both financial performance indicators and qualitative to make decisions which in turn affect the financial and qualitative performance of an organization.

De Noni et al (2007) describe their developments of both a quantitative and a qualitative model for default prediction. These models were designed to determine the credit risk of small to medium sized enterprises (SMEs). The final qualitative model of De Noni et al(2007) uses the following risk drivers:

- Planning and strategy: Planning and strategy looks at the following sub-components: Business strategy production strategy, human resources, labor and market value. The main area of interest is the organization’s capacity to plan for the future and to form a strategy which aligns with these plans.

(29)

ability to communicate with the customer and to deliver products which fit the market and the organizational strategy.

- Human resource management: Human resource management consists of the following sub-components: governance style, learning, decisional ability, staff turnover and days of absence. The human resource management component mainly looks at gaining maximum efficiency from the available employees.

- Production Organization: This component is divided in the following sub-components: production planning, technological strategy, physical environment, delivery punctuality and production efficiency. A certain overlap exists with the previous mentioned component of “planning and strategy” when it comes to measuring the production planning and

technological strategy. This component mainly looks at the organization’s ability to translate strategy and marketing goals into desirable products while maintaining and improving production efficiency.

- Innovation: The innovation component consists of R&D, Innovative environment, involvement in the production, knowledge management and innovation and originality. Innovation has become one of the most important areas for an organization to position itself on the market and therefore important for organizations to encourage and promote. However, innovation is very difficult to quantify and specify. Observing the level of innovation in an organization therefore will be very difficult.

A study held by ten Vergert and van der Weide (2001) for the ING bank and the Dutch ministry of economics sheds light on qualitative causes of bankruptcy. The causes which have been

acknowledged by this study are:

- Dependency on crucial employees: Key employees (including management) could be difficult to replace when they leave the organization. The posses knowledge and skills essential to the organization. Without these employees the organization is weakened.

- External causes: External causes can have a great impact on the performance and opportunities of an organization. The external causes include: political, economical en technological

developments, market developments, increased competition, increasing/decreasing markets, and the arrival of substitution products.

- Fraud: every organization is vulnerable to fraud. Measures should be taken to prevent and detect fraud.

(30)

- Risky financial policy: many organizations that went bankrupt had risky financial policies. Examples are high fixed costs and a relatively low owner’s equity

- Poor accounting information: without the proper accounting information managing an organization is very difficult. Wrong decisions are easily made because (financial)

consequences are not transparent. Often this will result in problems with sales and purchasing. - Social problems: social problems can have a major impact on a organization’s performance.

Examples of these social problems are: low involvement and commitment of employees, bad working relationships and poor internal communication.

Luckily for managers and organizations this same study also gives examples of factors which will guide an organization to success. In order to be successful an organization has to:

- Make good arrangements with suppliers: this mainly concerns the price of the supplied products and the terms of deliveries.

- Not be dependent on one or a few clients: If an organization becomes to dependant on one or a few clients than this organization is continuously at risk because it is not certain if it can retain these clients and therefore has no guaranteed income.

- Be continuously alert of new competitors: An organization that is aware of its competitors can take necessary actions.

- Have well structured organizational departments: the objectives of individual departments have to be clear and departments need to be able to work together efficiently.

(31)

3.4 Rating principles

Krahnen and Weber (2001) propose a foundation for something they refer to as “generally accepted rating principles”. They describe these principles as properties a good (internal) rating system should obey. This paragraph will describe these generally accepted rating principles. Although these

principles are developed for internal rating systems at banks they can be also applied (or adapted) in some extend to external credit risk rating systems. This paragraph will review the rating principles recognized by Krahnen and Weber (2001).

1) Comprehensiveness: A rating system should be flexible enough to be able to rate any organization. For a bank this means that it should be able to rate its current clients, but also past clients and future clients. Basically this means the bank’s rating system needs to able to rate organizations of any risk type (Location, industry, etc.). For external rating agencies this is even more important because they usually do not focus on specific industries, something a specialized bank could do.

2) Completeness: A rating system could be made so complex that it can cover any type of organization. However, it is more practical to develop different rating systems for industries with different characteristics. On the other hand, the rating should not be divided into too many subsets for every existing industry. The authors recommend to balance both aspects and suggest to make the reasons for choosing a certain numbers of rating systems transparent.

3) PD-definition: The (expected) probability of default (PD) should be properly defined. This means

that it should be defined what is considered to be a default and what is not. Also the time horizon used to calculate the PD should be given. The time horizon defines the period the estimated default chance refers to (For instance: the chance an organization will default within a one-year timeframe).

4) Monotonicity: Monoticity in the case of a rating means: when two companies have the same PD they should receive the same Rating. If the PD of one of these two companies is lower, than this organization should receive a higher or equal rating. It also implies that the company with a higher rating should by definition have a lower PD than that of one with a lower rating.

5) Fineness: The required level of fineness is determined by the purpose the rating system is used for. A rating system can have different purposes and therefore different requirements for its level of fineness. The most exact form of a rating system is one that models the probability of default. However, this also implies that such a rating system would have an infinite number of rating

(32)

reason the authors suggest allowing a rating system to communicate PDs in different degrees of fineness.

6) Reliability: If the PD of an organization remains stable than the organization should receive an identical rating regardless of the rater and the point in time when the rating is performed. According to the authors this does not mean that a rating might not change when the creditworthiness of an

organization changes along the economic cycle. However, the credit rating of an organization should not change if the creditworthiness does not change.

7) Back-testing: This rating principle requires that the result of back-testing rating data should not differ much from the results expected based on the estimated PDs. For back-testing purposes the rating system should not be divided into too many subsets of rating systems and the rating system should not be changed too often. The authors give the following necessary conditions for the appropriateness of the rating system.

- Ex-post default rates within any given rating category should be larger than that of a higher rating category.

- Even if it is not know whether a cardinal relation between rating and PD can be assumed, the above condition will test ordinality.

- Ex-post default rates should increase with the time horizon.

- Default rates of companies based on a time horizon of five years have to be equal or greater than those based on a time horizon of one year.

- For companies with corporate bonds outstanding, credit spreads may be compared to internal credit ratings.

- Across companies, is should be possible to compare the risk-ordering implied by the market with the risk-ordering implied by credit-ratings. Besides back-testing, credit ratings have to obey certain structural and technical necessities

8) Informational efficiency: A rating should include all available information to be efficient. Furthermore, one should not be able to predict rating changes based on rating history. In other words, if at this time it is already known that the creditworthiness of an organization will increase or decrease in the future then this information should be used to rate the organization at this point and the rating would not change in the future based on these facts.

(33)

10) Data management: Information has to be easily available in order to perform statistical analyses. Data management is essential for back-testing and successful system development.

11) Incentive compatibility: This principle requires the organization to minimize

misrepresentation by the rater. A rater might have an incentive to change the rating and might lower or higher a rating to acquire a rating that fits him/her (financially) the best. These unwanted activities have to be prevented. This can either be done in the way the rating system is designed or this can be done by thoroughly thinking about the way the rating system is implemented in an organization.

12) Internal compliance: In order to identify systematic biases in the rating of organizations their historical information has to be kept in a back-testing file.

(34)

4. Rating development

The FINAN indicative rating is aimed at providing business advisors and accountants with an advice giving tool which can be used to rate the creditworthiness of a “small to medium sized enterprise” (SME). The goal is to give insight in the overall performance of the organization with a strong emphasis on rating the organization’s creditworthiness. This chapter will describe FINAN’s rating development process. The development process of the FINAN indicative rating consists of two main steps. The first step is the development of a statistical rating model (The FINAN rating model) which estimates the probability of default of an organization (“indicative rating”). The second step consists of developing a scorecard model (The FINAN score model) which looks at the overall organizational performance of an organization. Both models are integrated into one tool (The FINAN indicative rating). The basic steps of the scorecard development process development are: selection of potential interesting performance indicators, describing the potential added value of individual performance indicators, selecting performance indicators and rating model development. Measuring the

performance of an organization is an important aspect of the FINAN indicative rating. This chapter

will therefore first discuss how “performance“ should be measured.

4.1 Organizational performance and financial health:

According to Devinney et al (2005) the multidimensionality of performance covers the many ways in

which organizations can be successful. Furthermore; according to the authors the dimensionality itself has several levels relating to: 1) who are stakeholders of interest for that particular measure, 2) how the firm interacts with its competitive environment and 3) what is the timeframe over which performance needs to be measured?

Stakeholders: According to Devinney et al (2005) the decisions made by an organization’s

management are tradeoffs between stakeholder’s interests. The authors view an organization as a place where stakeholders come together to create something of personal, societal and economic benefits and expect a return for any investment made. The authors also note that organizations which rely more heavily on bank financing will in generally be expected to manage investments more in line with the demands of these stakeholders and engage in investments with fewer risks. The authors conclude that measurement of performance requires accounting for the relevance of those measures to focal

stakeholders. An important aspect of the FINAN indicative rating will be its focus on the

(35)

have an important role in intermediating between banks and SMEs. In its most basic form this role includes giving advice about improving creditworthiness. In a more advanced situation the business advisor or accountant could even support the SME in a loan negotiation process at a bank. Neither creditworthiness or financial performance stand on their own, both are influenced by the input and interests of other stakeholders.

Competitive environment: Devinney et al (2005)conclude that, regarding competitive environment, the measurement of performance must take into account each firm’s own strategic positioning in relation to its competitive environment. Different measures need not be consistent, because organizations themselves are not. The FINAN indicative rating will for this reason include both quantitative as qualitative performance indicators. Measuring a organization’s performance compared to its competitive environment will be done in twofold. The financial performance of the organization will be measured based on qualitative performance indicators such as solvency and are compared with that of the primary industry of the organization. The indicators which set an organization apart from its primary competitors will have to be assessed by the business advisor or accountant. The FINAN indicative rating will have to provide guidelines to make a qualitative assessment of the competitive environment of the organization and that of other qualitative aspects of the organization.

Timeframe: The timeframe refers to the time period over which performance should be measured.

(36)

4.2 FINAN rating model

FINAN has developed a statistical default prediction model which will be known as the FINAN rating model. This default prediction model has been developed by an econometrician (Mr. Ran Chen) working at FINAN for this purpose. This chapter will describe the model development process used by FINAN. This chapter includes the (pre-) selection of potential risk drivers, the actual data extraction and cleansing, significance tests on risk factors, correlation tests on risk factors and the actual model development. The information in this paragraph summarizes an internal FINAN document (Appendix 3). Only the information relevant for this paragraph has been used (or added).

4.2.1 (Pre-) Selection of potential risk drivers

During the rating model development a list of potential performance indicators has been made (appendix 1) which includes common financial ratios. With the help of FINAN’s econometrician a pre-selection of performance indicators has been made from this list. Reasons for selecting a

performance indicator are its financial purpose, its economic value and its potential ability to predict defaults. The pre-selected performance indicators are:

- Solvency (owners’ equity / total assets): As mentioned in chapter 3, 75% of mid-sized and large banks consider the indebtedness of an organization as an important rating input factor. Solvency represents an organization’s ability to meet its long-term liabilities.

- Current ratio (current assets / current liabilities): The current ratio is one of the liquidity ratios. Liquidity refers to the organizations’ ability to meet their short term obligations A large share of banks also include the liquidity of an organization in their credit risk rating system.

Liquidity represents the ability of an organization to repay short term debts.

- Quick ratio ((current assets – inventories) / current liabilities): An alternative ratio for measuring liquidity is the quick ratio. The quick ratio is similar to the current ratio with the difference that it subtracts the inventories form the current assets before dividing it with the current liabilities. Inventories are essential to many organizations and often cannot be easily discarded to repay short-term debts. The true value of inventories is often also difficult to determine or are difficult to compare with other organizations. These could all be reasons to prefer the quick ratio to the current ration.

- Short term debt / total assets: This is not a very common financial ratio, however, early on in the credit risk rating model development process this turned out to be a risk driver which had a very high significance difference between defaulted organizations and non-defaulted

FINAN indicative ratingFINAN indicative ratingFINAN indicative ratingFINAN indicative rating

FINAN indicative rating

FINAN indicative rating

FINAN indicative rating

FINAN indicative rating

Default prediction rating and organizational performance scorecard

FINAN indicative rating

FINAN indicative rating

FINAN indicative rating

FINAN indicative rating

Default prediction rating and organizational performance scorecard

Preface

Management Summary

Table of contents

1 Problem statement

2 FINAN financial analysis background

3 Theoretical background

4. Rating development