C-FaR is a model using current data by the comparables approach to draw conclusion. It is not an easy task to find the current data. Five-year period is a long-term strategic horizon for the firms. This period contains drastic volatility especially in an emerging market, like Turkey. This volatile climate limited us to choose five-year period data set for application because this volatility affects Turkish firms considerably from strategic planning to decision- making. Therefore it is very common to see that Turkish firms concentrate more on short term planning than their worldwide counterparts. Five-year period started from the fourth quarter of 1998 to the third quarter of 2003. The thesis intends to develop C-FaR model to quantify the one-quarter ahead risk exposures of all non-financial firms in Turkey. Defense industry is not attractive for the investors because of high volatility. Therefore defense companies such as Aselsan, Otokar and Netas will be good examples for the application of C-FaR model. The management can concentrate on only one figure to evaluate their total risk exposure.        

The overall effect of risk components’ can be evaluated by the volatility of firms’ operating cash flow. Earnings before interest and tax (EBIT) was chosen to measure the operating cash flows. A forecasting model is needed to find forecast errors, that is, expected changes in the firms’ operating cash flows. Each firm will have 20 forecast errors for the five-year time period. These are not enough to comment on the firms’ cash flow volatility. Consequently, group of identical firms are needed because forecasting model depends on only one particular firm to estimate volatility in cash flows does not meet the expectations scientifically. Group of identical firms mean more forecast errors to facilitate making the decisions. The EBIT values, which are the indicators of the operating cash flows, are to be regressed against four lags of itself for the selected time period. These values are different for every firm. As a result, EBIT values were divided by the total assets. The forecast errors of the forecasting model are deviations in EBIT/TA. These forecast errors will be grouped according to two characteristics. These are market capitalization and stock price volatility. First forecast errors will be divided into three sub-samples according to market capitalization. Then each sub-sample will be divided into three sub-samples again according to stock price volatility. In the end, there will be nine sub-samples containing a sufficient number of forecast errors. The bucket of nine forecast errors will help in making decisions. That is, the histogram of the forecast errors can be judged from different points. These histograms of forecast errors enable us to come up with a single number showing non-financial firms’ total risk exposure.

Advantages and disadvantages, power, and the test of the model were explored and the detailed findings were presented for future studies in this chapter as well. The details of the application will be presented in the following sections.

3.2 Data Collection:

Stock markets are not only economical actors but also reliable sources of information for public. They put the transparent, reliable and audited information about the firms into the use of investors. The Istanbul Stock Exchange (ISE) database was designated as a primary source of information for that reason. ISE offers not only the historical data about Turkish firms but it also revises firms’ financial statements in each quarter.

C-FaR model demands comparable or peer non-financial firms to comment on a particular firm. ISE offers financial information about major companies in Turkey. These companies belong to different industries. The first step in the data collection is to decide which industries should be chosen to evaluate non-financial firms. Therefore food and beverage, textile and leather, wood, paper and printing, chemical, petroleum and plastic, non-metal mineral product, basic metal, metal products and machinery, information technology, defense, second national industries were selected.

The five-year period was chosen to collect past data for the firms in the sample. Each firm has totally 20 quarterly data for the period under examination. The ISE was founded in 1986 with fewer firms than it has now. As time went by, it lures many other firms who were seeking capital. New firms were quoted to ISE year by year. There were 156 non-financial firms, which were publicly traded in ISE in 1998. The firms, in alphabetical order, are presented in Appendix A. Some firms in 156 are not traded in ISE today for several reasons. There are also some firms with incomplete data. That is, they were not traded for a particular time in 1998-2003. Therefore they were removed from the sample. The remaining 147 firms were all included in the sample. The quarterly financial data were obtained through their income statement and balance sheets.

Earnings before tax and interest (EBIT) is the basic measure for the operating cash flow of the non-financial firms. EBIT can be obtained from firms’ income statement. The profit before tax and financial expenses are the two components of the EBIT. Financial expenses figure should be added to profit before tax figure to find EBIT.

There is no problem to find the profit of the firms for each quarter. Income statements show the cumulative magnitude over the year. That is, second quarter’s profit includes both the profit in the first and second quarter. The first quarter’s profit should be deducted from the second quarter’s profit in order to find the second quarter’s net profit. This is the same for third and fourth quarter of the firms as well. The only net value of the firms’ profits can be attainable in the first quarters. When you look at the profit figure in the income statement of the first quarter, the figure points the firms’ net profit for that quarter. The firms can have either profits or losses. Briefly, when looking at the profit before tax figure from income statement, attention should be concentrated on whether the firm profits or have losses. That is, the sign of the figure is very important.

In fact, there are some important problems when finding the financial expenses figure. This figure is in the income statement and behaved like the profit before tax figure. Financial expenses figure has normally a negative sign. When this figure is deducted from profit before tax, in fact it is added to profit before tax. Nevertheless, data collection process showed that there are some quarterly financial expenses figures, which have positive signs. That is, the firms have not financial expenses but rather gain in that period.

 The situation can be explained by two factors: Foreign exchange rate volatility and volatility in cost of goods sold. The foreign exchange rates fluctuate due to the floating exchange rates policy of Turkish Central Bank (CB) in Turkey since February 2001. As a result of this policy, firms’ financial expenses show volatility. The designated currency in the previous quarters may drop against the TL in the later quarters. With this information in mind, some positive financial expenses figures were obtained for some quarters. These figures should be deducted from profit before tax figure. These figures were really deducted from profit before tax because they have positive sign. This is the same for the cost of goods. Firms, which have export or import relations, have difference in cost of goods because of foreign exchange fluctuations. The difference is sometimes reflected as positive financial expenses.

The EBIT figure shows differences for each firm. This figure is not sufficient to comment on operating cash flows. In order to overcome the scale differences problem between firms, EBIT is divided by total assets. The new scale can be used for the comparative purposes since then. The EBIT/TA scale can be regarded as analogous to market-to book, price to earnings or price-to-cash flow rates.

          This figure plots the 2940 EBIT/TA of firms for 5-year time horizon. EBIT was calculated by subtracting “financial expenses” from “profit before tax” by the income statements. And total assets were taken from balance sheets. All EBIT/TA values were presented in Appendix B in each line. EBIT/TA values were used as a basic measure for the firms’ operating cash flow.  

3.3.1 Auto Regression Analysis:

The basic goal of the study is to find the deviations in cash flows of the firms in the sample. In order to do so, the forecast of the cash flows is needed. That is, there should be a model to forecast the next quarter’s cash flow. The simple multivariate regression analysis was used as a forecasting model. EBIT/TA t demonstrates the EBIT to total assets ratio in quarter t in the model. This was regressed against four lags of itself, which equates in auto regression in Microsoft Excel 2000. The result is given in the equation 3.1.

EBIT/TA_t,i = bo+b1*EBIT/TA_t-1,i + b2*EBIT/TA_t-2,i + b3*EBIT/TA_t-3,i +          b4*EBIT/TA_{t-4,i        +}    e                                                                           (3.1)