QUID 2017, pp , Special Issue N 1- ISSN: X, Medellín-Colombia

Transcription

1 QUID 07, pp , Specal Issue N - ISSN: X, Medellín-Colomba SCENARIO FORECASTING OF TENDENCIES OF DEVELOPMENT OF MACROECONOMIC INDICATORS OF THE REGION ON THE BASIS OF MODELS OF THE MULTIPLE CHOICE (ON THE EXAMPLE OF THE REPUBLIC OF TATARSTAN) (Recbdo el Aprobado el ) ula V. Hayrudnova Kazan naonal research echnologcal unversy, he asssan o deparmen of busness sascs and mahemacal mehods n economy, jula-84@mal.ru Anna V. Asyanova Kazan naonal research echnologcal unversy, professor of deparmen of busness sascs and mahemacal mehods n economy, ayanova@rambler.ru Mara R. Safulln Kazan Federal Unversy, he vce recor of he Kazan federal unversy concernng economc and sraegc developmen, Leond.Elshn@aar.ru Leond A. Elshn Kazan Federal Unversy, Cener of sraegc esmaes and forecass of Insue of managemen, economy and fnance; Sae Budgeary Insuon Cener of Perspecve Economc Researches of Academy of Scences of he Republc of Taarsan Absrac. The wor s performed accordng o he Russan Governmen Program of Compeve Growh of Kazan Federal Unversy (Number for publcaons: /8.9). The purpose of he sudy s o develop an algorhm for scenaro forecasng of macroeconomc ndcaors when decdng on he developmen sraegy of he regon usng he economc and mahemacal mehods of mulple choce. Based on he macrosrucural modelng mehod, we developed a model for forecasng he developmen of he man macroeconomc ndcaors. We mplemened an approach o forecasng he developmen rends of macroeconomc ndcaors n he regon based on he bnary choce model. Usng he bnary choce model allowed developng he mos probable drecons of he economc suaon developmen - opmsc and realsc forecas scenaro. Key words: regonal economy, regonal developmen sraegy, forecasng, mulple choce models, macrosrucural modelng. Car, eslo APA: Hayrudnova,., Asyanova, A., Safulln. M. & Elshn, L. (07). Scenaro forecasng of endences of developmen of macroeconomc ndcaors of he regon on he bass of models of he mulple choce (on he eample of he republc of Taarsan). Revsa QUID (Specal Issue)

2 . INTRODUCTION The modern mechansms of regulaon and developmen of he economy do no f no he norms of auomac and renewable developmen processes and are characerzed by he frequen volaons of some paerns esablshed n he pas. The economc developmen of naonal and regonal sysems can manfes self n a hghly unpredcable manner (Safulln, Elshn & Prygunova, 06) (Fredmann, 966). The dynamcs of economc growh can be sharply replaced by negave values and vce versa, he economc declne can drascally shf no he phase of progressve growh bypassng he processes of a smooh (phase) ranson from one sae o anoher n accordance wh he heory of cyclc developmen. The volaly of he observed developmen rajecores of soco-economc sysems acualzes he need for mulvarae calculaons ha deermne he forecas rajecores of he socoeconomc shfs (Izard, 966) (Granberg, 006). In oher words, he mos popular mechansms for plannng and sae regulaon of he economy are he mehods of scenaro modelng ha allow assessng he effec of he change n he dfferenaon of values of varous facors on he aggregae ndcaor under sudy. A he same me, should be noed ha hs ssue s gven a lo of aenon n he scenfc and research and professonal envronmena presen, as well as has been developed an eensve se of mehods and approaches. In our opnon, he mos promsng, among he consdered se of mehods for scenaro modelng of economc sysems, are mehods based on he use of bnary choce models. They allow us developng he mos probable drecons for he developmen of economc suaon, relyng, among oher hngs, on he probablsc esmaon mehods (Nazareno & Zvyagnseva, 0) (Asyanova & Kharudnova, 0). In hs regard, we made he bnary predcve esmaes of he mos probable scenaros of he developmen based on he adapaon of he V.V. Davns algorhm (Davns & Tnyaova, 005) o he facors and analyzed phenomena nheren n a sngle regon of he Russan Federaon, - he Republc of Taarsan - n hs sudy.. MATERIALS AND METHODS To consruc a combned scenaro forecasng model, we used a modfed mechansm for combnng sochasc forecasng mehods wh he eper esmaon mehods proposed by V.V. Davns (005), ncludng applcaon of he bnary choce model. A he nal sage, we mplemened he developed mechansm for consrucng a macrosrucural model for forecasng he macroeconomc ndcaors n he regon. The model s a sysem of economerc equaons, conssng of s nerrelaed equaons (Davns & Tnyaova, 005). A he same me, by projecng hs model o he regon under sudy (he Republc of Taarsan), we oo no accoun he specfc parameers of s developmen for he purpose of he hghly adaped V.V. Davns model o he socal and economc parameers and peculares of hs regon. Ths allowed obanng a sysem of qualave forecass, whch was confrmed by he daa on he sascal sgnfcance of he equaon raos. As a resul of analyzng he man rends n he developmen of he economy of he Republc of Taarsan, we denfed he mos sgnfcan facors of managemen and eernal envronmen ha can nfluence he adjusmen of economc developmen. The volume of he gross regonal produc s aen as he man ndcaor of he developmen rajecory of he economc sysem. Gven ha he qualy and speed of he economc growh of he republc are largely deermned by he radonally hgh level of nvesmen acvy, characerzed by he hghes rangs n Russa as a whole, he volume of nvesmen n he fed asses s chosen as he endogenous varable of he second equaon of he macrosrucural forecasng model. The ne sgnfcan ndcaor of he developmen and well-beng of he czens of he regon s he real rade urnover, whch ndcaes he level of effecve demand of he populaon. In erms of real rade urnover, he Republc rans 9h among he subjecs of he Russan Federaon. The Republc of Taarsan s an acve parcpan n he foregn economc acvy of he Russan Federaon, occupyng he s place n he Volga Federal Dsrc n erms of foregn rade urnover. Therefore, he epor volume ndcaor was also ncluded n he model as an endogenous varable. An mporan place n he regonal economy s occuped by a revenues (ncomes), whch are he bass for he formaon of he budge revenue. The endogenous varable of he sh equaon s he volume of shpped goods of own producon, wors performed and servces rendered by own srengh n he ndusry.

3 The mahemacal model conans egh eogenous varables. Durng he model consrucon, we eamned he cause-effec relaonshps beween he endogenous and eogenous varables on he qualave level, ang no accoun he muual nfluence of facors on each oher. To evaluae he relaonshps, we used he mehods of correlaon and regresson analyss, an unconvenonal mehodology of he Durbn-Wason creron, suggesng he mplemenaon of a seres of procedures amed a deermnng he crcal value of hs creron for s dfference from zero. Ths allows rejecng or, conversely, accepng he mos relable hypoheses ha form he bass for he evaluaons and analyss mplemened. We also used he ess of he cause-effec relaonshps of Granger. We checed he qualy of each equaon of a sysem of nerrelaed economerc equaons, whch made possble o conclude ha he model corresponded o he descrbed process. Thus, a macrosrucural model was he bass for he mehodology approbaon а b a Х 8 a3 X а b _ a X 3 а3 a3 X 3 b3 4 а4 b4 а4 X 5 а5 b5 a5 X 6 а6 a6 X 6 _ GRP, mln.roubles (n comparable prces); X volume of ol producon, hs. ons; nvesmens n fed asses; 3 urnover of real rade, mln.roubles; Х growh rae of nvesmens n fed asses n he Russan Federaon, n %; X 3 money supply (un M); 4 epor volume, mln. US dollars; 5 a revenues, mln.roubles;, where X 4 raes of ncrease n ol prces, USD/barrel; () 6 volume of shpped goods of own producon, wors performed and servces rendered by own srengh n he ndusry, mln. roubles; Х 8 average annual number of persons employed n he economy, hs.people; Х 9 agrculural producs of all caegores, mln.roubles) (Davns & Tnyaova, 005). When developng he regonal forecass, he man dea was o use a mul-rend model ha provded an adequae descrpon of he processes wh several ways of developmen. The essence of he proposed mehod s as follows. Inally, we creaed an auoregressve model for each eogenous varable of he macrosrucural model. As s nown, he radonal auoregressve model s a me seres model n whch he values of me seres are lnearly dependen on he prevous values of he same seres a a gven momen: ~ a 0 a a a m m () As a rule, s no always possble o oban a successful model denfcaon ha mees all he qualy crera. Therefore, almos always here s a need o use dummy varables, whch are nroduced by calculang he devaon of calculaed values from acual ones, m, T (3) The dummy varables allow us breang he nvesgaed se of observaons no wo componens, n accordance wh he followng rule: 0, 0, I, 0 I. (4) The resul of hs dvson s a vecor f (called a dummy varable) conssng of 0 and. Addng an addonal (new) varable o he equaon leads o he appearance of he followng equaon:

4 ~ a0 a a amm d f, (5) m, T, f n whch d - rao value a a dummy varable; - rao value a a dummy varable a he momen (Asyanova & Kharudnova, 0). In hs sudy, he goal was neher o acheve a hgh qualy of he consran equaon, bu o oban a "splng" ool,.e. denfcaon of he dummy varable sae wh dfferen rajecores of developmen (opmsc and pessmsc rajecores). A model wh one dummy varable provdes wo forecas rajecores, he dfferenaon of whch s deermned by he consdered effecs. The resulng model s called mul-rend. There s a need o oban probablsc esmaes of he plausbly of predcve rajecores, based on he approaches defned by he mulnomal log model: b j e j J e j0 J J e j0 bj jbj, j, J, (6) Accordng o he mehod of V.V. Davns (Nazareno & Zvyagnseva, 0), he eper esmaes are used as a vecor of ndependen varables, accordng o whch he alernave opons are dsngushed.in he sudy, he ndependen varable was represened a varable havng drec mpac on he endogenous varable. As a resul, s possble o buld a combned model for forecasng he endogenous ndcaors y along wo developmen rajecores a f and a f 0 aˆ aˆ aˆ 0 aˆ m m where ep bˆ K ep 0 bˆ, f (7) ˆ - -h forecas calculaon opon; d - a j delayed values of he dependen varable; - evaluaon of he j -h regresson rao; - realy probably of he -h forecas esmae opon; f - vecor of values ha oo he dummy varables n he -h opon; d - rao vecor for dummy varables; b - evaluaon of he parameer vecor of log model of mulple choce of he -h opon; - vecor of values descrbng he condons epeced n he ancpaory perod (Asyanova & Kharudnova, 0). 3. RESULTS AND DISCUSSION In hs sudy, he combned model s used o forecas he macroeconomc ndcaors of he soco-economc developmen of he regon (Republc of Taarsan). Accordng o he above algorhm, he frs-order auoregressve models were creaed for each eogenous change n he macrosrucural model () (Table ): Table. Resuls of creang he auoregressve models for eogenous varables of macrosrucural model of socoeconomc forecasng. Varable Х Х Х 3 Х 4 Х 5 Х 6 Auoregresson х 0.97х equaon R - deermnaon rao 54.6 х х х х5 0.6 х х.х х 0.95х х , х F- Fsher creron

5 An analyss of he presened sascal characerscs of he qualy of auoregressve models (Table ) made possble o reveal ha all he equaons obaned for eogenous varables were sascally sgnfcan, ecep for he auoregressve equaon for he varable х -he growh raes of nvesmen n fed asses n he Russan Federaon. The qualy of he auoregresson equaon for he facor х should o be ncreased by nroducng a dummy varable ha denfes he random effec. To oban he values f, we calculaed devaons of he calculaed values of dependence equaon from he acual values accordng o rule (4). If he devaon value was posve, hen he dummy varable f oo he value of, oherwse - he value of 0. Usng a dummy varable allowed obanng he followng regresson equaon wh a dummy and a lag varable: а0 df a, where 0, f 0, f, f 0 (8) при Fgure. Dynamcs of nvesmen n fed capal of he Russan Federaon on opmsc and pessmsc developmen rajecores. Tang no accoun he splng of he eogenous facor х no wo developmen rajecores, we obaned he followng predcve dynamcs for he endogenous varable (nvesmen n he fed capal of he RT) (Fgure ): a Afer denfcaon, he followng model was obaned: f 6, (7.6) (6.9) (9) All he raos of he resulng equaon are sgnfcan. оптимистичный прогноз пессимистичный прогноз opmsc forecas pessmsc forecas When obanng he forecas values for 06, he rajecory spl no possble opons: for f 0 and f. The drecon of he suaon developmen f s aen as an opmsc scenaro, and a f 0 a pessmsc scenaro. The correspondng forecas values of he growh raes of nvesmens n he fed capal of he Russan Federaon were 0.34% and 85.87% (Fgure ): Fgure. Dynamcs of nvesmen n fed capal of he RT on opmsc and pessmsc developmen rajecores. 4.CONCLUSIONS The resuls of he mplemened esmaes and calculaons demonsrae a sgnfcan level of nfluence of nvesmen acvy on he parameers of he socal and economc developmen of he

6 Republc of Taarsan. Ths mpac s due, frs of all, o he level of he effecve nfrasrucure secury of economc processes, he degree of deprecaon of fed asses, ec. n he secoral cone esng n he regon. Ths, n urn, predeermnes a hgh level of nvesmen aracveness and, as a resul, he sgnfcan growh raes of nvesmens n fed asses and he economy growh as a whole. In hs regard, he degree of deprecaon of fed asses was aen as an ndcaor. By usng he "STATISTICA" pacage, we obaned he dsrbuon of probables of he onse of a forecas opmsc or realsc drecons and a model of he form: f 0 ep where degree of deprecaon of fed asses, %. The Wald sascs show he rao sgnfcance. Based on he resuls of he mulple choce model, he mos probable s he predced value of he ndcaor under sudy, calculaed on he bass of an opmsc developmen rajecory. Thus, hs model gves he forecased value of nvesmens n fed asses of he Republc of Taarsan - 667,35.6 bln. roubles. wh a probably of 66%. 4. SUMMAR In he conduced research, we appled he echnque o fnd he forecas values of macroeconomc ndcaors of he regon developmen n he form of alernave rajecores wh probablsc esmaes of he degree of her realy. The forecased values of nvesmens n fed asses of he Republc of Taarsan are calculaed accordng o he opmsc and pessmsc developmen rajecory, ang no accoun he nfluence of eernal facors. The obaned model allows fndng he forecas values for one me perod n advance. If s necessary o forecas for a more dsan fuure, hs procedure s repeaed, whle he forecas esmae obaned a he frs sep s used as he acual nal nformaon (Logunov, 06). The applcaon of he developed mehods and models s possble for mang predcve calculaons n he process of jusfyng managemen decsons, ang no accoun he wors-case scenaro n he fuure. ACKNOWLEDGEMENTS The wor s performed accordng o he Russan Governmen Program of Compeve Growh of Kazan Federal Unversy (Number for publcaons: /8.9). REFERENCES Asyanova, A.V. & Kharudnova, u.v. (0). Forecasng of he Soco-Economc Developmen Indcaors of he Regon // Bullen of Kazan Technologcal Unversy. No. 0. Davns, V.V. & Tnyaova, V.I. (005). Forecas Models of Eper Preferences: Monograph. Voronezh: Publshng House of Voronezh Sae Unversy, p. 48. Fredmann, J. (966) Regonal developmen polcy. Boson: Mass. Ins. Techn., p. 37. Gadelshna, G. A., Asyanova, A.V. (03) Prof Forecasng of an Enerprse Usng a Mul-Trend Model. Bullen of Kazan Technologcal Unversy. No.. Granberg, A.G. (006). Bases of Regonal Economy. М.: Publshng House of he Sae Unversy - Hgher School of Economcs. Izard, U. (966). Mehods of Regonal Analyss: Inroducon o he Scence of he Regons. М.: Progress. Logunov, V.N. (06). Facors of Invesmen Concenraon n Fed Asses. Rereved 8 July 07 from: hp:// = Naonal Rang of he Invesmen Clmae n he Subjecs of he Russan Federaon. Resuls of Rereved from: hps:// 6/05_ru.pdf Nazareno, A. & Zvyagnseva, O. (0). Scenaro Forecasng of he Developmen of Socal and Economc Sysems. Scenfc Journal of KubSAU. No. 84. Safulln, M. R., Elshn, L. A. & Prygunova, M.u. (06). Influence of he Epecaons of Economc Agens on he Economy Cyclcal Developmen Parameers under Varous Types of Srucural Dsequlbrum.News of he FEFU.Economcs and Managemen. 4. P.50 64