REAL SECTOR CONFIDENCE INDEX FROM THE BUSINESS TENDENCY SURVEY OF CBRT

Transcription

1 REAL SECTOR CONFIDENCE INDEX FROM THE BUSINESS TENDENCY SURVEY OF CBRT Ece Oral, S.Özge Tuncel * Central Bank of the Republic of Turkey Research and Monetary Policy Department Ece.Oral@tcmb.gov.tr Ozge.Tuncel@tcmb.gov.tr Abstract The Business Tendency Survey (BTS) and the constructed real sector confidence index of the Central Bank of the Republic of Turkey (CBRT) are introduced in this survey. BTS conducted since December 1987, has some deficiencies, emanating basically from questions on similar subjects that are treated equally by most of the respondents. In addition, the periods of some questions cannot be differentiated and a few questions exhibit permanent pessimism. Moreover, questionnaire, sample, sectoral coverage and timing of BTS do not match with the EU Harmonised Industry Survey. Some modifications in BTS for the purpose of enhancing the effectiveness and making it compatible with the EU harmonised survey are also suggested in this study. Keywords: Business tendency survey; Industrial production index; Cross-correlation; Real sector confidence index. * The views expressed in this paper are those of the authors and do not necessarily correspond to the views of the Central Bank of the Republic of Turkey.

2 1. INTRODUCTION The surveys on expectations primarily designed to signal changes in economic activity are widely used in macroeconomic assessments and forecasts. The main aim of the business tendency surveys conducted in various ways is to find out the general tendency of the cyclical developments and provide economic decision-makers with the necessary information about future expectations. Business Tendency Survey (BTS) of the Central Bank of the Republic of Turkey (CBRT) has been conducted since December It has been prepared with the aim of discovering the opinions of senior managers of the major private sector firms on several economic aspects over the recent past, near future and current situation. The intention of this study is to introduce BTS and the real sector confidence index derived from some questions of BTS in details. Afterwards, the suggested changes to improve the performance of BTS and to harmonise it in the EU Industry Survey are presented. The paper is organized as follows. We discuss BTS in Section 2. The confidence index and its performance are described in Section 3. Section 4 briefly outlines the effectiveness of BTS data and the suggested changes in BTS within the context of the harmonised EU programme of business and consumer surveys. 2. BUSINESS TENDENCY SURVEY BTS is a monthly survey, which is intended to find out the senior managers assessments and expectations on their firm, their sector and the general economy. BTS launched as a pilot study in September 1987 has been carried out since December The frame is a fixed panel composed of 83 industrial enterprises ranked on the First 5 Major Industrial Enterprises of Turkey and the Next 5 Major Industrial Enterprises of Turkey listings of the Istanbul Chamber of Industry (ICI) who have accepted to participate in the survey (see Appendix A). The questionnaire contains 34 questions. The majority of the questions are qualitative, while there are a few quantitative questions on PPI inflation and interest rate expectations. Variables covered in BTS are about demand (total, domestic, export), employment, investment, production, stocks (finished goods, raw materials), cost and selling prices, finance requirement, general economy (business situation, export prospects, inflation and interest 2

3 rate). The questions can be found in Appendix B. The monthly report of the survey is disseminated to the public through web site 3. REAL SECTOR CONFIDENCE INDEX Business tendency surveys provide the necessary data for the measurement of the business confidence. Qualitative and ordinal choice structure of these surveys makes the responses sensitive to the cyclical developments (OECD, 23). The index constructed by combining a set of survey questions in a single composite index is called confidence index because it sums up economic agents assessments and expectations of the economic situation. For constructing a real sector confidence index for Turkey, the survey questions providing qualitative information on the current situation as well as on the expectations for the next three months and assessments for the last three months are used. The variables related to the index are the ones which measure an early stage of production (e.g. new ), respond to changes in economic activity rapidly (e.g. stocks), measure expectations or draw a picture of overall business condition (e.g. general business condition), measure improvement in economic conditions (e.g. investment expenditures) (OECD, 23). The methodology used for the construction of the real sector confidence index is the same as the one derived in OECD (Nilsson, 1999). First, the balance of each question is computed by subtracting the negative choice from the positive choice and then the diffusion index is calculated by adding 1 to this balance (Sutanto, 1999). Second, Industrial Production Index (IPI) has been chosen as the reference series for comparison, since the aim of the index is to forecast the expansion and contraction periods of the economic activity. For the cycles of IPI, the series derived by the cooperative study of the CBRT and OECD is used (See Next, the necessity of standardization, smoothing and seasonal adjustment for each diffusion index is checked. Since the standard deviations of all series range in a narrow band, the series of BTS can be set to have the same fixed scale. Furthermore, irregular components effect is found to be little on each of the diffusion indices. Consequently, no standardization and smoothing seem to be necessary. Even though seasonal influences are found to be present in some of the indices, the series are not seasonally adjusted for the sake of simplicity. As a result, questions 1, 2, 3, 9, 11, 12_2, 15_2, 16_1 and 18_1 are chosen, where _1 stands for the last 3 months and _2 for the next 3 months, to construct the real sector confidence index, namely MBRKGE, on the basis of cross-correlation analysis, peak-trough analysis, 3

4 volatility test and economic significance. The results of the principal component analysis used to choose optimal weights illustrate that scores have approximately equal weights (Ece et al., 25). In order to examine the performance of MBRKGE, its components degrees of consistence with the quantitative realizations are evaluated. It should be noted that, survey series in level form refer to the related quantitative series in de-trended form, while survey series in change form are more compatible with the related quantitative series in first difference form (change over previous period or change over the same period of the previous year) (OECD, 23). In case of quarterly realization series, monthly index is transformed into quarterly series. Detrended series is obtained by eliminating trend component via HP filter. The correlations between MBRKGE questions and the corresponding realizations are given in Table 1. In general, it can be said that they are highly consistent with the realizations. The graphs of the questions as in balance form and the realizations can be seen in Figure 1. Table 1: The Cross-Correlation Analysis of Quantitative Realizations with the MBRKGE Questions Questions of MBRKGE Questions of MBRKGE *Detailed information about the series can be found in Appendix C. **Q-o-Q represents quarter-to-quarter changes and Y-o-Y represents year-to-year changes. Transformation Related Quant. Series** Related Quantitative Series* Lead** Industrial Production Index General economic situation (q1) (Manufacturing, Private) Quarterly Mean Y- o- Y % Change Export prospects (q2) Export of Goods and Services Quarterly Mean Y- o- Y % Change 1.46 Investment expenditures (q3) Machinery Expenditure Quarterly Mean Y-o-Y % Change Total (q9) Industrial Production Index (Manufacturing, Private) Quarterly Mean Detrended.617 Monthly stocks of finished Changes in Stocks goods (q11) Quarterly Mean Employment (q12_2) Production Workers (Private) Quarterly Mean Y-o-Y % Change Volume of output (q15_2) Industrial Production Index (Manufacturing, Private) Quarterly Mean Y-o-Y % Change 1.68 Goods sold in domestic market (q16_1) Private Consumption Quarterly Mean Y-o-Y % Change.724 Raw material stocks (q18_1) After comparing the MBRKGE questions with the related realizations, MBRKGE s performance is tested. MBRKGE, having the property of leading at the turning points of economic activity, is compared with the cycles of IPI. Figure 2 illustrates the historical performance of MBRKGE in this regard. The economic recession between 1988 and 1989, the slump during the Gulf crises in 199, the currency crisis in 1994, the Asian and Russian crisis of 1997, the financial crisis in February 21 following on the liquidity crisis in November 2 and the upturns in 1989, , 1995 and 2 are forecasted by the Cross- Corr. Import Volume of Intermediate Goods - Y-o-Y % Change.55 4

5 Balance 1 Industrial Production Index (Annual Percentage Change) Balance 2 Goods and Services Export (Annual Percentage Change) Q1 1989Q3 1991Q1 1992Q3 1994Q1 1995Q3 1997Q1 1998Q3 2Q1 21Q3 23Q1 24Q3 26Q1 1988Q1 1989Q3 1991Q1 1992Q3 1994Q1 1995Q3 1997Q1 1998Q3 2Q1 21Q3 23Q1 24Q3 26Q Balance 11 Changes in Stocks (Secondary axis) Balance 12_ Index of Production Workers (Annual Percentage Change) 1988Q1 1989Q2 199Q3 1991Q4 1993Q1 1994Q2 1995Q3 1996Q4 1998Q1 1999Q2 2Q3 21Q4 23Q1 24Q2 25Q3 1988Q1 1989Q3 1991Q1 1992Q3 1994Q1 1995Q3 1997Q1 1998Q3 2Q1 21Q3 23Q1 24Q3 26Q Balance 15_2 Industrial Production Index (Annual Percentage Change) Balance Machinery (Annual Percentage Change, Secondary Axis) Q1 199Q3 1992Q1 1993Q3 1995Q1 1996Q3 1998Q1 1999Q3 21Q1 22Q3 24Q1 25Q3 1988Q1 1989Q3 1988Q1 1989Q3 1991Q1 1992Q3 1994Q1 1995Q3 1997Q1 1998Q3 2Q1 21Q3 23Q1 24Q3 26Q1 Balance Industrial Production Index (Detrended, Secondary Axis) Balance 18_1 Import Volume of Intermediate Goods (Secondary Axis) Q1 199Q1 1992Q1 1994Q1 1996Q1 1998Q1 2Q1 22Q1 24Q1 26Q1 1991Q1 1992Q3 1994Q1 1995Q3 1997Q1 1998Q3 2Q1 21Q3 23Q1 24Q3 26Q1 Jan.95 Apr.96 Jul.97 Oct.98 Jan. Apr.1 Jul.2 Oct.3 Jan.5 Apr.6 Figure 1: Questions of MBRKGE and Related Realizations Balance 16_1 Private Consumption (Y-o-Y Change)

6 confidence index. Therefore, the index seems to be useful for monitoring and detecting changes in economic activity. The correlation between the cycles of IPI and MBRKGE is found to be.5885 at lag 3. Figure 2: Real Sector Confidence Index and the Cycles of IPI 12 Real Sector Confidence Index (MBRKGE) Cycles of IPI Economic Recession Gulf War 1994 Crises Russian Crises Bottom of 21 Crises 9 8 Dec.87 Marc.89 Jun.9 Sep.91 Dec.92 Marc.94 Jun.95 Sep.96 Dec.97 Marc.99 Jun. Sep.1 Dec.2 Marc.4 Jun.5 Sep.6 MBRKGE is also compared with the official composite leading indicator (MBÖNCÜ - SÜE) of the Turkish economic activity, consisting of four BTS questions and three macroeconomic series. Figure 3 shows that the cyclical turning points of MBRKGE are in line with those of MBÖNCÜ-SÜE. According to the cross-correlation analysis there is a coincident relation between these two indicators with a correlation of.916. However, it is important to note that MBÖNCÜ-SÜE and MBRKGE include some of BTS questions in common such as export possibilities, stocks of finished goods and total amount of employment. Besides, the question on raw material stocks of MBRKGE can correspond to the official statistics used in MBÖNCÜ-SÜE, namely import volume of intermediate goods. Figure 3: Business Confidence Index and Leading Indicator of Industrial Production Index MBRKGE MBÖNCÜ-SÜE (right scale) Dec.87 Marc.89 Jun.9 Sep.91 Dec.92 Marc.94 Jun.95 6 Sep.96 Dec.97 Marc.99 Jun. Sep.1 Dec.2 Marc.4 Jun.5 Sep.6

7 4. POSSIBLE CHANGES IN BTS The CBRT intends to do several modifications in BTS, which will probably be effective by the beginning of 27, in order to improve its performance and harmonise it within the context of the EU programme. In this regard, the sampling methodology, the questions and the timing of the survey are examined comprehensively Current Sampling Methodology and Planned Modifications The survey is aimed at the senior managers of the industrial enterprises that are ranked among the First 5 Industrial Enterprises of Turkey and the Next 5 Major Industrial Enterprises of Turkey lists prepared by the ICI. Participation in the survey is voluntary. Sampling method adopted is non-probabilistic purposive sampling method. The sample is revised once a year as to cover the new companies included in the ICI's latest biggest firms ranking, and kept fixed thereafter until the next revision. Data is collected via postal and electronic mail. The current sample size comprises 83 enterprises and the recent response rate is 55 percent. The survey covers enterprises most of which operate in the manufacturing sectors of 15 to 36 according to NACE Rev. 1 classification, namely, food, textiles, forest, paper products, chemistry, stone, metals and machine-vehicle. Machine-vehicle, textiles, food and chemistry sectors make up approximately 75 percent of the entire participation in the BTS. The share of the participants from the mining-quarrying and energy sectors, which belong to nonmanufacturing grouping, is about 8 percent. The prospective modifications related to the sampling methodology include enlargement of the sector coverage of the sample for improving its representativeness. Possible sample design will somehow be based on the Business Registers (BR) of Turkish Statistical Institute (TURKSTAT). The considered frame comprises private companies that have at least 1 employees and contributes to 8% of the value added in their sectors at 4-digit sector level. The sample unit is the local unit i.e. legal entity in BTS. This sample unit can be kept in the new survey. The participants from the companies that operate in more than one sector is requested to reply the questions on the basis of the main activity of their company. Sample can be drawn by either cut-off or stratified sampling methods. In this sample, the number of observations in the sub-sectors at 3-digit level should be considered to be adequate except for the ones that comprise relatively fewer private companies. 7

8 In the current implementation survey results are disseminated without weighting, whereas the results of the sectors should be weighted by their production values in the new setting Current and Candidate Survey Questions Based on our experience till now, it can be said that there are limitations in some of the BTS questions. Respondents perceive the questions on a similar subject in the same manner so that they give close answers. They also cannot distinguish the last 3 months trends from the next 3 months expectations. Furthermore a few questions exhibit permanent pessimism. Besides, BTS questions do not match with the EU Harmonised Industry Survey. Therefore it is essential to reconstruct the questionnaire of the survey. For this purpose, current survey questions that are not included in EU Industry Survey are examined comprehensively in order to test their performance. After evaluating the performance of the current survey questions, our recommendations on possible changes in the questionnaire are briefly outlined Performance of the Current BTS Questions The correlation analysis of domestic demand related questions are shown in Table 2-3 and foreign demand related questions in Table 4-5, respectively. Table 2: Trend of the Last Three Months Table 3: Trend of the Next Three Months New domestic (q13_1) Production (q15_1) New domestic 1.98 (q13_1) Production (q15_1) Goods sold (q16_1) Work in progress (q19_1) Goods sold (q16_1) Work in progress (q19_1) New domestic (q13_2) Production (q15_2) New domestic (q13_2) Production (q15_2) Goods sold (q16_2) Work in progress (q19_2) Goods sold (q16_2) Work in progress (q19_2)

9 Table 4: Trend of the Last Three Months Table 5: Trend of the Next Three Months Export (q1)* New export (q14_1) Volume of exported Goods (q17_1) Export (q1)* New export (q14_1) Exported Goods (q17_1) Export opportunities (q2) New export (q14_2) Exported Goods (q17_2) Export opportunities (q2) New export (q14_2) Volume of exported Goods (q17_2) *Note that q1 shows this month s trend There exist significantly high correlations between questions within each group (domestic and foreign demand) indicating that the respondents cannot make a clear distinction between the questions on similar subjects. There is also another point to be highlighted about the information content of BTS. Table 6 shows that there exist high correlations between the last and next three months periods for some questions. In other words, the managers seem to preserve the same structure in their assessments for the same variables over the last and next three months period. Table 6: Correlations Between the Last and the Next Three Months' Trends Employment (d12-1, d12-2).85 New from domestic market (d13-1, d13-2).7 New from export market (d14-1, d14-2).8 Production (d15-1, d15-2).68 Goods sold in domestic market (d16-1, d16-2).66 Exported goods (d17-1, d17-2).75 Raw- material stocks (d18-1, d18-2).56 Work in process (d19-1, d19-2).7 Stocks of finished goods (d2-1, d2-2).67 Average unit cost (d21-1, d21-2).98 Average price for new domestic (d22-1, d22-2).96 Average price for new export (d23-1, d23-2).9 The last point which must be taken into account is the Bias, appeared as permanent optimism or permanent pessimism. Whereas the participants are systematically pessimistic in answering questions related to monthly developments of the financial requirement, export 9

10 and average unit costs. However, the persistency in optimism or pessimism can be corrected by comparing the gap between the diffusion indices and their long-term averages Suggested Changes in the BTS Questionnaire All of the EU Industry Survey questions (monthly and quarterly) should be included in the new survey (European Commission, 26). In this context, questions 12_2, 15_1, 15_2, 22_2 and 24 should remain in the survey, while wording of questions 5, 9, 1, 11 and 25 should be examined and appropriate modifications should take place. In addition, questions 1, 3, 13_1, 13_2, 14_1, 14_2, 21_1, 21_2, 3 and 32, which exhibit satisfactory performance should be retained Timing of BTS Currently, the questionnaires for the surveyed month (t) are sent in the last working day of the previous month (t-1), and the participants are requested to return the completed forms by the 3rd day of the following month (t+1). A reminding is sent to the participants who do not sent their reply until the 3rd of the following month. Following the reminder, the fieldwork is ended within a week. No revision is made for the late responses since January 24. Returns are subject to editing and validation. The timing in EU harmonized system is as follows: The monthly surveys are carried out in the first fortnight of each month and the results in the form of excel files are sent by to the European Commission before the end of the month. The quarterly surveys are carried out in the first fortnight of the first month of each quarter (January, April, July and October) and the results are sent to the Commission under the same format before the end of that month (European Commission, 26). 1

11 Acknowledgment We are grateful to Cevriye Aysoy for her valuable comments. REFERENCES Ece, D., Hamsici, T., Oral, E., (25), Building up a real sector business confidence index for Turkey, Joint European Commission OECD Workshop on International Development of Business and Consumer Tendency Surveys, November 25, Brussels. ( European Comission (26), The Joint Harmonised EU Programme of Business and Consumer Surveys, User Guide 26. Nilsson R. (1999), Business Tendency Surveys and Cyclical Analysis, Business Tendency Surveys, Proceedings of the First Joint OECD-ADB Workshop, Manila, November. OECD (23), Business Tendency Surveys: A Handbook. Sutanto A. (1999), Business Confidence Index, Consumer Confidence Index and Index of Leading Indicators: An Experiment for Indonesia, Business Tendency Surveys, Proceedings of the First Joint OECD-ADB Workshop, Manila, November. 11

12 Appendix A. List of Metadata Items for Business Tendency Survey Top level Child level Business/Consumer Survey Specifics Source Name of collection / source used Organisation and contact person Types of data sources used (household, establishment / enterprise survey) Source periodicity Business Tendency Survey Dilek TALI, MSc. Assistant Manager Department of Statistics Division of the Real Sector Data The Central Bank of the Republic of Turkey İstiklal Caddesi No.1 61 Ulus Ankara TURKEY Tel: (9-312) Fax: (9-312) Dilek.Tali@tcmb.gov.tr Enterprise Survey Monthly Statistical concepts and classifications used Key statistical concepts used The first 23 questions, Questions 29 and 32 are presented in a 3- Definition of indicators, variables collected Classification(s) used point Likert scale format. Question 24 is a multiple-choice question. Questions between 25 and 28 are about ranking some factors. Questions 3, 31, 33 and 34 are quantitative questions. The questionnaire translated into English is given below in Appendix B. ISIC is used. Data characteristics and collection Unit of measure used Sampling frame Percentage of positive replies (more optimistic, more than adequate, higher, above normal, up), unchanged replies (the same, adequate, normal) and negative replies (less pessimistic, less than adequate, lower, below normal, down) indicate the direction of change of a variable (question). The frame used has a list type. The list consists of the industrial enterprises that are ranked among the First 5 Industrial Enterprises of Turkey and the Next 5 Major Industrial Enterprises of Turkey lists prepared by the Istanbul Chamber of Industry. Size of actual frame list The number of enterprises in the frame is approximately 832. Characteristics included in frame list Identification number, name, address, sector code, phone and fax of the enterprise; contact person's name, title and address; average number of employees; turnover; exports; beginning inventory and ending inventory; own funds (paid-in capital). Frame list update The frame is updated yearly and the date of the last update / revision is made in May 24. Sampling method(s) Non-probabilistic purposive sampling method based on the participation of selected volunteers (fixed panel). Sample size The average number of units included in the survey is between 4 and 5. Sample representativeness The sample represents the industrial enterprises that are ranked among the First 5 Industrial Enterprises of Turkey and the Next 5 Major Industrial Enterprises of Turkey lists prepared by the Istanbul Chamber of Industry. Most of the sub-sectors are not covered in the sample. Response rate The average response rate for the most recent survey(s) is 5 Non-response procedures Reporting unit Survey method percent. Non-responding units are telephoned in order to obtain information. Enterprises. 49 units by postal mail, 423 units by 12

13 Top level Child level Business/Consumer Survey Specifics Timeliness The average delay between the end of the reference period and the date of public release of data is one week. Reference period The length of the reference period to which the survey data relates, is one month Fieldwork period One month Base period There is no base period for confidence indicators calculated in index form. Date base period last updated - Link to release calendar - Other data and collection characteristics The sample consists of enterprises from private sector only. The survey has been adopted from the Industrial Trends Survey of the Statistical population and scope of the data Statistical population Geographic coverage Sector coverage Population coverage - Other coverage issues - Confederation of British Industry (CBI). The target population used for the survey is the major private industrial enterprises. The survey includes all the national territory. Mining, food, textiles, forestry, paper products, chemicals, stone & soil, metals, machinery and energy. Manipulation and dissemination Aggregation & consolidation Estimation procedures Imputation Transformations Weights Seasonal adjustment - Dissemination format(s) The first step is to convert the numbers of answers to each of the three reply options into percentages. The net balance is then calculated. Diffusion index is calculated by adding 1 to the balances. The real sector confidence index is calculated by taking averages of the selected diffusion indices. Sample weights are not used. No imputation is used. Balances are the difference (in percentage points of total answers) between positive and negative options. Each diffusion index is constructed by adding 1 to the balanced values. No weighting is used. The data from the survey are published by the Central Bank of the Republic of Turkey's websites ( Other aspects Other comments - 13

14 Appendix B. Questions of BTS 1. Your opinion about the general More Optimistic Same More Pessimistic course of business in your industry, compared with previous month 2. Your opinion about More Optimistic Same More Pessimistic export prospects over the next three months, compared with previous month 3. How much investment expenditure More Same Less do you expect to realize over the next 12 months compared to last 12 months 4. Your capacity utilization compared More Same Less with previous month 5. What is the level of your productive More than Adequate Adequate Less than Adequate capacity in accordance with your demand expectations for the next 12 months 6. What is the level of your sales Higher Same Lower revenues compared with previous month 7. What is the level of your past-due receivables compared with previous Higher Same Lower month 8. What is the financial requirement of your Higher Same Lower firm for the next month compared with previous month Excluding seasonal variations: 9. Total amount of received Above Normal Normal Below Normal this month 1. Amount of export Above Normal Normal Below Normal received this month 11. Amount of monthly stocks of finished Above Normal Normal Below Normal goods this month 14

15 Excluding seasonal variations, what is the last 3 months trend and what will be the next 3 months trend for the following items: Trend of the next Trend of the last three months three months 12. Total employment Up Same Down Up Same Down 13. The amount of new Up Same Down Up Same Down received from the domestic market 14. The amount of new Up Same Down Up Same Down received from the exports market 15. The volume of output Up Same Down Up Same Down 16. The volume of goods sold Up Same Down Up Same Down in domestic market 17. The volume of exported goods Up Same Down Up Same Down 18. The volume of raw-material Up Same Down Up Same Down stocks 19. The volume of work Up Same Down Up Same Down in progress 2. Stocks of finished goods Up Same Down Up Same Down 21. Average unit cost Up Same Down Up Same Down 22. Average price for the new Up Same Down Up Same Down received from the domestic market 23. Average price for the new Up Same Down Up Same Down export 24. According to your existing received or production plan, Month Month Month Month Month Month Month how many months is your production programme 25. Over the next quarter, Order- Sales Labor (*) Capacity-Plant Credit-Finance Input-Cost which factor(s) might limit the production, rank according to degree of importance (*) Qualified / Non-qualified 15

16 26. Over the next quarter, Price comp. Delivery date. Cred. finan Quotas- oth. Rest Foreign conj. which factor(s) might limit your export, rank according to degree of importance 27. Over the next 12 months, New investment what is the main reason for To increase capacity the planned expenditures on To increase productivity building, plant or equipment, Renovation rank according to degree No spending planned of importance 28. Which factor(s) might restrict Cost of financing the realization of these Shortage of capital expenditures, rank according to Shortage of external resources degree of importance Insufficient demand Insufficient net proceeds Cost of labor force 29. Over the next three months, Up Same Down what is your expectation for inflation (producer prices) rate 3. Over the next twelve months, what is your expectation for inflation (producer prices) rate 31. What is your expectation for year-end inflation (producer prices) rate % % 32. Over the next three months, Up Same Down what is your expectation for short term Turkish Lira credit interest rate 33. Over the next three months, what is your expectation for short term Turkish Lira credit interest rate (Determine as an annualized rate) 34. Over the next twelve months, what is your expectation for short term Turkish Lira credit interest rate (Determine as an annualized rate) % % 16

17 Appendix C. Data Description Questions Realizations General economic situation Industrial Production Index, (1997=1) (TURKSTAT) (Quarterly), (Private) Manufacturing Export prospects GNP at fixed prices (1987=1) (TURKSTAT) (Quarterly), (Expenditures) Exports (Goods and Services) Investment expenditures GNP at fixed prices (1987=1) (TURKSTAT) (Quarterly), (Expenditures) Machinery (Private) Total Industrial Production Index, (1997=1) (TURKSTAT) (Quarterly), (Private) Manufacturing Monthly stocks of finished goods GNP at fixed prices (1987=1) (TURKSTAT) (Quarterly),(Expenditures) Changes in Stocks Employment Index of Production Workers in Manufacturing Industrial Establishments, (1997=1) (TURKSTAT) (Quarterly), (Private) Manufacturing Volume of output Industrial Production Index, (1997=1) (TURKSTAT) (Quarterly), (Private) Total Industry Goods sold in domestic market Raw material stocks GNP at fixed prices (1987=1) (TURKSTAT) (Quarterly), (Expenditures) Consumption (Private) Foreign Trade Import Volume Index by Classification of BEC, (1994=1) (TURKSTAT) (Monthly), Import- (Volume) Intermediate Goods 17