JOINT EUROPEAN COMMISSION OECD WORKSHOP ON INTERNATIONAL DEVELOPMENT OF BUSINESS AND CONSUMER TENDENCY SURVEYS BRUSSELS 14 15 NOVEMBER 25 Business Survey and Short-Term Projection Edita Holickova Statistical Office of the Slovak Republic - November 25 -
Business Survey and Short-Term Projection Edita HOLICKOVA, Statistical Office of the Slovak Republic Results of business survey are being also taken as a base for short-term projection in the Slovak Republic. Following this aim the Statistical Office of the Slovak Republic uses capacities of its subordinated Institute of Statistics and Informatics - INFOSTAT that deals in the field of economic trends modelling and solves research and developing projects of the sector of statistics. Within the framework of one of completed projects in the year 24 comparison of selected production indicators trends with business tendency balances was made. For completing this task the EVIEWS software was applied. Concerning the industry sector, monthly time series of industrial production index (IPI) were compared with business tendency balance of expected output that was based on survey results from the previous period of 1998-24. The data was not seasonally adjusted; through the comparison higher tightness of dependence was found out by use of initial data compared to the use of seasonally adjusted ones. High tightness of dependence by selected indicators followed from. Comparison of IPI trends with business tendency balance of expected output of the period (t-1) 6 5 4 3 2 1-1 -2 1998 1999 2 21 22 23 24 D L O G _ I P P O c a k a v a n a p r o d u k c i a ( t - 1 ) 2 15 1 5-5 -1-15 -2 DLOG(IPP)*1=C(1)*Q5_OCAKQ(-1)+C(2)*@SEAS(1)+C(3) *@SEAS(3)+C(4)*@SEAS(4)+C(5)*@SEAS(7)+C(6)*@SEAS(1) +C(7)*@SEAS(12)+C(8)*REZ_IPP211(-2)+C(9)*DLOG(IPP(-1)) *1+C(1)*UM213 Coefficient Std. Error t-statistics Prob. C(1).7333.11775 6.22759. C(2) -7.449167 1.246667-5.975266. C(3) 6.9423 1.91327 6.359441. C(4) -7.97695 1.1644-6.852813. C(5) -8.47844.952959-8.44511. C(6) 6.949937 1.1326 6.15243. C(7) -11.83646.94778-12.49787. C(8) -.251511.83146-3.24937.35 C(9) -.218621.5689-3.897787.2 C(1) 8.629486 1.163421 7.417335. R-squared.878914 Mean dependent var.17637 Adjusted R-squared.863565 S.D. dependent var 6.698568 S.E. of regression 2.474254 Akaike info criterion 4.764899 Sum squared resid 434.6572 Schwarz criterion 5.651 Log likelihood -182.9784 Durbin-Watson stat 2.345176
Concerning the construction sector, the data on construction output were used for the comparison, depending on business tendency balance of expected construction activity in the previous month for the period of 1993-24, and the data on employment in comparison with business tendency balance of expected number of employees in the previous month for the period of 1999-24. Comparison of STAVPROD - construction output trends with business tendency balance of expected construction activity of the period (t-1) 2 16 12 8 4-4 -8-1 2-1 6 1994 1996 1998 2 22 24 DLOG_STAVPROD O cakavana stavebna aktivita (t-1) 3 2 1-1 -2-3 -4-5 -6 DLOG(STAVPROD)*1=C(1)*Q8_OCAKQ(-1)+C(2)*@SEAS(1)+C(3) *@SEAS(2)+C(4)*@SEAS(3)+C(5)*@SEAS(4)+C(6)*@SEAS(5) +C(7)*@SEAS(6)+C(8)*@SEAS(1)+C(9)*@SEAS(12)+C(1) *REZ_PROD211(-2)+C(11)*DLOG(STAVPROD(-1))*1+C(12) *UM213 C(1).34234.1968 3.12134.22 C(2) -42.54159 1.457197-29.19413. C(3) -5.75917 2.79633-2.122397.357 C(4) 19.97819 1.39249 15.25928. C(5) 14.34968 1.66696 8.931175. C(6) 13.72952 1.362656 1.7555. C(7) 1.585 1.375787 7.6958. C(8) 9.238127 1.1535 8.3298. C(9) -9.759181 1.28232-7.61674. C(1) -.191399.7432-2.575328.111 C(11) -.339111.59381-5.71756. C(12) 11.272 1.517299 7.251514. R-squared.93746 Mean dependent var.87946 Adjusted R-squared.93268 S.D. dependent var 14.88728 S.E. of regression 3.88178 Akaike info criterion 5.6394 Sum squared resid 1942.195 Schwarz criterion 5.881862 Log likelihood -384.9787 Durbin-Watson stat 1.918457 Results of both comparisons carried out in the sector of construction indicate the most significant and very high tightness of dependence by time series trends in the construction sector, and it is worth to state that the trend is significantly similar while expected employment rate in the years 22-23 has been more optimistic compared to the really recorded development. By the calculation primary time series were entered in. 3
Comparison of ZAMSTAV - number of employees in construction trends with business tendency balance of expected number of employees of the period (t-1) 6 4 2-2 -4-6 -8-1 1999 2 21 22 23 24 DLOG_ZAMSTAV O cakavany pocet zam estnancov (t-1) 8 6 4 2-2 -4-6 -8 DLOG(ZAMSTAV)*1=C(1)+C(2)*Q4_OCAKZAM(-1)+C(3)*@SEAS(1) +C(4)*@SEAS(4)+C(5)*@SEAS(6)+C(6)*@SEAS(12)+C(7) *UM111 C(1).791553.9226 8.579575. C(2).27194.2678 1.15447. C(3) -3.2547.264567-12.11568. C(4).722612.242977 2.973995.41 C(5).496377.248935 1.9944.54 C(6) -1.73595.261758-6.631897. C(7) 1.914862.256461 7.466484. R-squared.91758 Mean dependent var -.114421 Adjusted R-squared.99972 S.D. dependent var 1.841451 S.E. of regression.552521 Akaike info criterion 1.743515 Sum squared resid 19.84317 Schwarz criterion 1.964858 Log likelihood -55.76656 Durbin-Watson stat 1.64374 By comparison of results in the retail trade sector retail trade receipts and balance of expected receipts of the previous month for the period of 1995-24 were used. High tightness of dependence followed from the comparison, but the expectations were rather higher compared to the really recorded development. At the same time also some extreme negative expectations by the respondents became evident in prevailing part of evaluated periods, for instance in the years 1995-1996, 1999-2, but also in 23-24. On the other hand, positive expectations by the respondents were not coming true in majority of cases. The results are to considerable extent influenced by rather high level of non-response in this sector, but also of significant changes that were experienced in recent 1 years (closing small businesses, income of big trade chains, establishing new hypermarkets). The data have not been seasonally adjusted. 4
Comparison of TRO - retail trade receipts trends with business tendency balance of expected receipts of the period (t-1) 1 8 6 4 2-2 -4-6 95 96 97 98 99 1 2 3 4 3 2 1-1 -2-3 -4-5 DLOG_TRO Ocakavane trzby (t-1) DLOG(TRO_MO)*1=C(1)*Q7_OCAKTRZBY(-1)+C(2)*@SEAS(1) +C(3)*@SEAS(3)+C(4)*@SEAS(8)+C(5)*@SEAS(1)+C(6) *@SEAS(12)+C(7)*REZ_TRO_111(-2)+C(8)*DLOG(TRO_MO(-1)) *1+C(9)*UM113 C(1).3416.5868 5.812398. C(2) -25.8329 1.673-25.67695. C(3) 6.689767 1.16256 6.582758. C(4) -2.38559.942877-2.5332.128 C(5) 4.319913.931629 4.636944. C(6) 1.11964.913131 11.8235. C(7) -.123985.69398-1.786577.768 C(8) -.153213.3479-5.26863. C(9) 9.7947 1.1872 9.53985. R-squared.916894 Mean dependent var 1.25852 Adjusted R-squared.91738 S.D. dependent var 9.46753 S.E. of regression 2.826558 Akaike info criterion 4.9898 Sum squared resid 862.8588 Schwarz criterion 5.22275 Log likelihood -282.933 Durbin-Watson stat 1.7757 Results of models applied and parameters of seasonality in partial calculation, however, indicate that the seasonality is significant, but for the purpose of making trend comparison the calculations have been made with the use of primary data, i.e. not seasonally adjusted. Finally calculated parameters, coefficients by individual variables and the t-statistics do confirm appropriateness of applied approach. The calculations were completed by INFOSTAT and presented hereby as confirmed by the authors: Messrs. Michal OLEXA, Ján HALUŠKA and Ms. Jana JURIOVÁ. 5