Prediction of institutional sector development and analysis of enterprises active in agriculture

103 4, XXII, 2019 Business Adminis a ion and Managemen DOI: 10.15240/ ul/001/2019-4-007 In oduc ion Owing o a special ole o ag icul u e in he na ional economy, go e nmen s ha e become he main supplie s as well as he main use s o ag icul u al p edic ions ( o ecas s). They equi e in e nal o ecas s o implemen policies ha p o ide echnical and ma ke suppo o he ag icul u al sec o (Hed ich, Loy, & Muelle , 2012). Fo ecas s o ag icul u al p oduc ion and p ices ough o be help ul no only o he go e nmen s, bu p ima ily o a me s and he en i e ag icul u e indus y. Thus, ag icul u e is an a ea whe e poli icians, consume s, scien is s and en i onmen alis s encoun e (Ri e a-Fe e, 2008). Remeikiene, Rozsa, Gaspa eniene and Pěnčík (2018) s a e ha suppo i e poli ical a i udes owa ds he ag icul u al sec o along wi h employmen o p o ec ionis measu es de e mine igno ance o he ules o supply and demand, dis o ion o he condi ions o ee ma ke compe i ion, closeness o he ag icul u al sec o in compa ison o o he economic sec o s, incomple e in e na ional ag icul u al p ice ansmission, inconsis ence o long- e m ma ke p ices o ag icul u al commodi ies and exis ence o ag icul u e in dis a ou ed a eas. In i s Eu ope 2020 de elopmen s a egy, he Eu opean Commission has de i ned an ambi ious goal o imp o ing he le el o esou ce e i ciency (Li, 2018). Analyses o echnical e i ciency and o al ac o p oduc i i y (TFP) wi hin he Czech ag icul u e ha e been conduc ed by Čechu a (2012). His objec i e consis s in iden i ying key ac o s ha de e mine an e i ciency o inpu u iliza ion and TFP de elopmen . He uses a i xed con ol model o es ima e he echnical e i ciency and TFP design o he o e all s a us in ag icul u e and i s indi idual sec o s. The esul s show ha echnical ine i ciency is an impo an phenomenon o he Czech ag icul u e and i s indi idual sec o s. The mos impo an ac o s ha de e mine bo h he echnical e i ciency and he TFP a e ac o s associa ed wi h ins i u ional and economic changes, pa icula ly a d ama ic inc ease in mea impo s and an inc ease in subsidies (Čechu a, 2012). In addi ion, an analysis o ag icul u al p oduc i i y ends in he EU coun ies on he TFP basis was pe o med, o ins ance, by Dokic, Jo ano ic and Vujanic (2017). Thei esul s show ha he o e all EU ag icul u al p oduc i i y g ow h has slowed down in ecen yea s and has lagged behind leading global compe i o s, which is mainly due o dec easing numbe o employees in ag icul u e. Acco ding o he au ho s, ag icul u e de elopmen should be es ablished on scien i i c bases. Remeikiene e al. (2018) s a e ha his may be caused when he olumes o he in e na ional ade in ag icul u al p oduc s a e dec easing. One o he basic p inciples o sus ainable ag icul u e is he e o e o ecas ing i s u u e de elopmen . Fo ecas s a e p edominan ly made using con en ional econome ic me hods, wi h ime se ies app oaches ha ing smalle oles (Klieš ik, V bka, & Rowland, 2018). In ela ion o ce ain dominance o ag icul u al economis s, g ea emphasis has been pu on explana ions, ye he p edic i e powe o models has been o only small conce n. In ecen yea s, se e al ag icul u al economis s ha e been engaged in compa ing o ecas s wi h a ious o he me hods and hei conclusions gene ally co espond o commonly accep ed belie s. In he case o sho - e m o ecas ing, combina ions lead o be e and mo e accu a e p ognoses, e en be e han he ones p oduced by ec o au o- eg ession, which is su p isingly one o he bes me hods. Fo example, Klepáč and Hampel (2018) wan ed o i nd ou i i possible o p edic bank up cy 1–3 yea s ahead wi h he sound accu acy. Thei esul s show PREDICTION OF INSTITUTIONAL SECTOR DEVELOPMENT AND ANALYSIS OF ENTERPRISES ACTIVE IN AGRICULTURE Voj ěch S ehel, Jakub Ho ák, Ma ek Vochozka EM_4_2019.indd 103EM_4_2019.indd 103 25.11.2019 11:02:2525.11.2019 11:02:25 104 2019, XXII, 4 Business Adminis a ion and Managemen ha models a e able o no ice co ec labels o ac i e companies, howe e , he longe ime pe iod o bank up cy is, he wo se he esul s a e, o he esul s a e inconsis en o bank up ed companies which ha e been ac i e ill now. The au ho s o cou se also say ha bank up cy p edic ion is cha ac e ized by he ac ha he esul s a e ma kedly in l uenced by he da a used and pa ly by se ing o classi i ca ion models. Ka as and Režňáko á (2017) e i i ed whe he bank up cy p edic o s a e speci i c in e ms o indus y o ime and es ablished which p edic o s can signal an imminen bank up cy mo e han one pe iod be o e bank up cy occu s. The esea ch con i med hei p esump ion. Au ho s iden i i ed indica o s, which a e indus y-speci i c and a e equen ly associa ed wi h a speci i c pe iod o ime p eceding he bank up cy ( o manu ac u ing companies: he e u n on asse s and he ne wo king capi al o o al asse s; o cons uc ion companies: he ne wo king capi al o sales and he in e es co e age). Ka as and Režňáko á (2017) iden i i ed also indica o s whose inclusion in he model would p obably inc ease i s p edic ion capaci y – hese a e he indica o s e u n on asse s, in en o y u no e and asse composi ion. As demons a ed by Hed ich, Loy and Muelle (2012), i is also su p ising ha econome ic models as well as one-dimensional me hods a e poo ly compa able and nai e models. Howe e , i is inc easingly challenging o p edic bank up cy isk as co po a ions ha e become mo e global and mo e complex and as hey ha e de eloped sophis ica ed schemes o hide hei ac ual si ua ions unde he guise o op imiza ion o ax au ho i ies (Klieš ik e al., 2018). The p ice ola ili y o commodi ies has inc eased g ea ly in ecen yea s. In o ma ion abou he de elopmen o ag icul u al ma ke s has dissemina ed among ma ke pa icipan s o di e ing deg ees. This in o ma ion asymme y is he basis o ading p o i s on ma ke s. Di e en o ecas ing ools, pa icula ly s a is ical and econome ic me hods, we e de eloped in he ag icul u al sec o in he pas , bu hey did no achie e e y good o ecas ing accu acy, he eby esul ing in a conside able p ice isk. The addi ional in o ma ion held by ma ke pa icipan s and o he people in he ag icul u al sec o has been neglec ed by hese ools (Hed ich, Loy, & Muelle , 2012). Hed ich e al. (2012) u he add ha me hods using he “wisdom o c owds” e ec achie ed be e o equal o ecas ing accu acy in many o ecas ing applica ions han he s anda d app oaches applied. The “wisdom o c owds” e ec indica es ha g oups each be e esul s han indi iduals o expe s. P edic ion ma ke s a e seen by he same au ho s, i.e. Hed ich e al. (2012), as a new o ecas ing me hod. They use a ading mechanism simila o a s ock ma ke o achie e he “wisdom o c owds”. P edic ion ma ke s a e o ums o ading con ac s ha yield paymen s based on he ou come o unce ain e en s. The e is moun ing e idence ha such ma ke s can help o p oduce o ecas s o e en ou comes wi h a lowe p edic ion e o han con en ional o ecas ing me hods (Valáško á, Klieš ik, Š ábo á, & Adamko, 2018). Fo example, p edic ion ma ke p ices can be used o inc ease he accu acy o poll- based o ecas s o elec ion ou comes (A ow e al., 2008). Business analysis is used o collec ing necessa y da a and hei iden i i ca ion, including o mula ing he need o change(s) in unc ioning o o ganiza ions and acili a ing he change(s). In e ms o e i cien use o business analyses, an in i ni e amoun o da a may be p o ided (Vochozka, Rowland, & V bka, 2016). Andekina and Rakhme o a (2013) desc ibes business analysis as a disciplined app oach o implemen ing and managing changes wi hin o ganiza ions, whe he hey a e bene i cial businesses, go e nmen s o non-p o i o ganiza ions. A p esen , a i i cial in elligence can be de i ni ely ecognized as a use ul ool o business analyses and o ecas ing. This esea ch may be da ed back o he 1940s, and o e he las h ee decades, i has been ex ended up o he poin o sol ing ag icul u al p oblems as well (Rowland, & V bka, 2016). A i i cial neu al ne wo ks a e l exible in hei own use, and a e able o analyse highly complex pa e ns e y quickly (San in, 2008). These ne wo ks can be used o classi y, app oxima e unc ions, and p edominan ly o p edic ime se ies (Al un, Bilgil, & Fidan, 2007). E en in he ag icul u al sec o , a i i cial in elligence has helped o de elop machine lea ning algo i hms om a i i cial neu ons and a i i cial neu al ne wo ks ha mimic he human b ain wi h syn he ic neu ons (Vochozka, 2017). The disad an age o hese ne wo ks is he need o la ge sample da a, since many es obse a ions a e needed o p oduce such da a, howe e his is e y EM_4_2019.indd 104EM_4_2019.indd 104 25.11.2019 11:02:2525.11.2019 11:02:25 105 4, XXII, 2019 Business Adminis a ion and Managemen complica ed o use s (Hossain, Chao, Ismail, No oozi, & Khoo, 2017). Fo da a p ocessing and business analyses, he so-called Kohonen ne wo k can be applied. I consis s o an inpu laye ha is comple ely in e connec ed wi h an ou pu laye and is sel - o ganizing, i.e. has he abili y o lea n wi hou “a eache ”. This ne wo k has a e y wide ange o uses, since i is an al e na i e ne wo k applicable o mos neu al ne wo k calcula ions (Vochozka & Macho á, 2018). I is used mainly o audio edi ing, speech p ocessing, pho os, ideos, secu i y applica ions, and allows o he p ojec ion o high dimensional da a in o lowe dimension da a (Konečný & T enz, 2010). When using his a ac i e model o g oup da a se s in o di e en g oups, he da a a e g ouped by a sys em whe e eco ds wi hin a g oup end o be simila o each o he and eco ds in di e en g oups a e di e en . Many expe imen al esul s show ha Kohonen ne wo ks a e e y e ec i e no only o assessing ag icul u al companies (Han & Wang, 2008). The objec i e o he con ibu ion is an analysis o companies ac i e in ag icul u e o he Czech Republic using Kohonen ne wo k and he subsequen p edic ion o hei de elopmen . 1. Da a and Me hods The esea ch ques ion was se : Is i possible o i nd ou he s a us o companies ope a ing in ag icul u e in he Czech Republic on he basis o clus e analysis using a i i cial in elligence? Fo he pu poses o his con ibu ion, a da a se will be c ea ed, which will include comple e da a om i nancial s a emen s o 4,201 companies ac i e in ag icul u e o he Czech Republic in 2016. These a e he en i ies whose co e ac i i ies a e classi i ed in Sec ion A o he economic ac i i ies classi i ca ion CZ-NACE. The se o companies will be gene a ed om he Bisnode Albe ina da abase. The da a will be lis ed in an Excel able. Each line will con ain da a om i nancial s a emen o one conc e e company, which will be iden i i ed by i s name and iden i i ca ion numbe . The da a se will no con ain he da a o he en e p ises ha did no pe o m hei co e ac i i ies in he en i e moni o ed pe iod, ha is he companies ha we e closed in his pe iod (and he e o did no ha e any signi i can in l uence on he di ec ion o na ional economy), and companies ha s a ed hei ac i i ies (and did no ha e a signi i can in l uence on he di ec ion o he ag icul u e sec o in he CR ei he ). The companies ha s a ed hei ac i i ies on 1 Janua y 2016 o hose ha i nished hei business on 31 Decembe 2016 could ha e been conside ed; howe e , his would no ha e big in l uence on achie ing he objec i e o he con ibu ion. Mo eo e , he columns o da a showing no a iance we e also excluded. The da a se will be subsequen ly subjec ed o clus e analysis using Kohonen ne wo k. Fo clus e analysis, Dell’s S a is ica so wa e, e sion 12 will be used. The e will be used he Da a mining module and as a speci i c ool, neu al ne wo ks. Neu al ne wo k wi hou a eache (Kohonen ne wo k) will be used. The da a o analysis will be chosen – Excel able wi h da a se . In all cases hese a e con inuous p edic o s. The se will be di ided in o h ee pa s: 1. T aining da a se : ep esen s 70% o he en e p ises om he da a se . Fo his se , Kohonen ne wo k will be c ea ed. 2. Tes ing da a se : i includes 15% o companies om he o iginal da a se . This se will be used o e i ying he pa ame e s o c ea ed Kohonen ne wo k. 3. Valida ion da a se : his se will also con ain 15% o he companies om he da a se . This da a se will be es ed he applicabili y o inapplicabili y o he c ea ed Kohonen ne wo k. Topological leng h and wid h o Kohonen ne wo k will be se a 10. The numbe o i e a ions will be se a 10,000. Howe e , i shall be no ed ha e o le el is a decisi e ac o . I he Kohonen ne wo k pa ame e s a e no imp o ed wi h each ei e a ion, he aining will be i nished be o e he 10,000 h i e a ion is comple ed. In case ha he ne wo k pa ame e s do no imp o e wi h he 10,000 h ei e a ions, he en i e p ocess mus be epea ed, and a highe numbe o equi ed i e a ions shall be se in o de o ge he bes esul possible. The lea ning speed will be se a 0.1 a he beginning and 0.02 a he end. The esul s, i.e. he di ision o he indi idual companies in o 100 clus e s will be en e ed in Excel able. Subsequen ly, he indi idual clus e s will be subjec ed o analysis o absolu e and selec ed indica o s (o mo e p ecisely, hei mean alues – a i hme ic a e age) and he esul s will be in e p e ed. EM_4_2019.indd 105EM_4_2019.indd 105 25.11.2019 11:02:2625.11.2019 11:02:26 106 2019, XXII, 4 Business Adminis a ion and Managemen 2. Resul s Based on he me hods applied, clus e s we e c ea ed. The numbe o companies in each clus e o he Kohonen ne wo k is shown in Fig. 1. Fig. 1 shows a 3-D ep esen a ion o he Kohonen ne wo k c ea ed and he numbe o companies in he indi idual clus e s. The i gu e shows ha he highes ep esen a ion o companies is in clus e (1, 8), and clus e (3, 9). The hi d place is occupied by clus e (3, 10). Mo eo e , he clus e wi h a highe ep esen a ion o companies a e clus e s (4, 10) and (5, 10). Fo he emaining clus e s, he ep esen a ion o companies is signi i can ly lowe . I shall also be no ed ha h ee o he Kohonen ne wo k clus e s a e no occupied. Mo e pa icula ly, hese a e clus e s (2, 3), (3, 1) and (3, 2). To ge a de ailed iew on ep esen a ion o companies in he indi idual clus e s, Fig. 1 is complemen ed by conc e e alues shown in Tab. 1. Tab. 1 shows ha only he clus e s ma ked in g ey include mo e han 100 companies. O he clus e s include less han 100 companies, wi h on (2, 7) clus e includes 90 companies. The ollowing analysis deals only wi h he highligh ed clus e s. They con ain da a on 1,919 companies, which accoun s o 45% o all he companies in he se , al hough hey a e in 5 clus e s ou o he 100 exis ing (and 97 occupied ones). Mo eo e , one o he clus e s, speci i cally (1, 8), includes 990 companies, i.e. almos 24% o he companies in he se . I can he e o e be assumed, ha he analysis pe o med will ha e a su i cien explana o y powe and will be able o ep esen he po en ial o he en i e Czech ag icul u e. The mean alues o he se as a whole, and selec ed clus e s a e gi en in Tab. 2. Tab. 2 clea ly shows he di e ence o alues o ag icul u e as a whole and o indi idual clus e s. The mean alues o he whole ag icul u e sec o a e signi i can ly highe Fig. 1: Numbe o companies in each clus e o Kohonen ne wo k Sou ce: own EM_4_2019.indd 106EM_4_2019.indd 106 25.11.2019 11:02:2625.11.2019 11:02:26 107 4, XXII, 2019 Business Adminis a ion and Managemen Neu on 1 2 3 4 5 6 7 8 9 10 1 1 3 9 2 4 17 16 990 26 4 2 2 7 0 20 44 42 90 30 29 11 3 0 0181448328682427249 4 131315194645372264150 5 9 14 26 22 27 22 47 15 72 103 6 10212413193721395556 7 6 17 28 17 25 36 23 17 37 48 8 6 10 12 28 17 16 7 40 23 30 9 2 51931181928231818 10 1 2 12 24 31 26 21 34 14 26 Sou ce: own I em In o al (1, 8) (3, 9) (3, 10) (4, 10) (5, 10) To al asse s 74,413.5 1,297.1 2,914.8 5,281.9 7,782.5 13,113.9 Fixed asse s 50,237.8 694.7 1,164.8 2,651.2 4,589.1 7,737.8 In angible i xed asse s 139.5 5.1 12.0 4.4 8.3 82.0 Tangible i xed asse s 46,922.1 524.8 1,123.8 2,340.4 4,170.6 6,410.0 Land 20,010.9 57.3 56.9 189.6 468.8 1,010.0 Cons uc ions 15,733.1 110.5 171.5 285.6 1,182.2 1,424.7 Sepa a e mo able hings and se s o mo able hings 6,748.5 53.2 201.0 366.4 693.5 1,112.6 Long- e m i nancial asse s 3,176.2 164.8 29.0 306.7 410.2 1,245.8 Cu en asse s 23,623.8 589.0 1,705.8 2,555.4 3,035.1 4,961.8 In en o ies 9,576.0 68.4 284.4 643.5 802.6 1,580.9 Ma e ial 954.6 6.7 23.4 38.5 66.5 41.1 Un i nished and semi- i nished p oduc s 1,985.7 6.9 31.6 57.5 82.5 308.5 Long e m ecei ables 487.9 19.1 35.0 17.3 169.2 101.0 Sho e m ecei ables 7,784.0 182.1 661.8 1,130.4 1,354.3 2,221.3 T ade ecei ables (sho - e m) 4,260.9 43.0 198.8 354.8 487.2 700.4 S a – ax asse s 691.7 10.2 31.6 68.6 116.6 279.6 O he ad ances p o ided 165.1 4.0 9.9 21.3 80.9 32.9 Sho - e m i nancial asse s 5,789.6 320.6 724.7 764.8 712.3 1,058.6 Cash 141.8 49.4 54.8 109.0 94.4 185.1 Bank accoun s 4,466.3 96.9 236.0 302.8 279.9 384.6 Acc uals 513.9 10.5 42.5 73.5 150.2 389.3 Sou ce: own No e: The amoun s a e gi en in housands CZK. Tab. 1: Numbe o companies in indi idual clus e s o Kohonen ne wo k Tab. 2: Mean alues o selec ed a iables o mos ep esen ed clus e s EM_4_2019.indd 107EM_4_2019.indd 107 25.11.2019 11:02:2625.11.2019 11:02:26 108 2019, XXII, 4 Business Adminis a ion and Managemen han he mean alues o he clus e s wi h he highes numbe o companies. This could mean ha he selec ed clus e s conside ed also he size o companies as one o he pa ame e s o clus e ing ( his does no necessa ily mean i was a decisi e ac o ; i could ha e happened inad e en ly). The mos impo an alues o asse s a e gi en in g aphs below. Fig. 2 compa es he mean alues o asse s o ag icul u e in o al and o indi idual clus e s. Fo he sec o o ag icul u e, a e age balance shee , ha is, he alue o asse s is almos CZK 75 million pe company. Clus e (1, 8) achie es he a e age asse s alue o CZK 1.2 million. Only one clus e , speci i cally (5, 10) achie es he a e age alue o o al asse s o mo e han CZK 10 million. In his clus e wi h he alue o mo e han CZK 13 million he companies ha e mo e han CZK 5 million a hei disposal compa ed o he companies Fig. 2: Mean alues o asse s o ag icul u e and o selec ed clus e s Sou ce: own Fig. 3: Mean alues o i xed asse s o ag icul u e sec o and o selec ed clus e s Sou ce: own EM_4_2019.indd 108EM_4_2019.indd 108 25.11.2019 11:02:2625.11.2019 11:02:26 109 4, XXII, 2019 Business Adminis a ion and Managemen in he second mos success ul clus e (4, 10). These a e ollowed by clus e s (3, 10) and (3, 9) wi h he asse s a he amoun o almos CZK 5.3 million, o almos CZK 3 million. An indica o o equal impo ance, which di e s he companies in he indi idual clus e s is he amoun o i xed asse s (Fig. 3). Ag icul u e is a e y speci i c sec o . I is a s a egic ins i u ional sec o o he na ional economy. I s esul s a e highly dependen on wea he and ex e nal in l uence s. The ou pu s o ag icul u al p oduc ion a e seasonal in cha ac e . Due o his ac , i egula cash l ows a e ypical o companies ac i e in ag icul u e. In ag icul u e is also e o o highe le el o p oduc ion mechaniza ion and au oma ion. Fixed asse s hus indica e a possible ma ginal a e o eplacemen o capi al o wo k. The mo e machines he companies ha e, he ewe wo ke s a e needed. In ag icul u e, he amoun o i xed asse s pe one company is abo e CZK 50 million. The companies in clus e s examined use i xed asse s signi i can ly less. The bes posi ion has he clus e (5, 10), achie ing he alue o mo e han CZK 7.2 million. The clus e wi h he highes numbe o companies, clus e (1, 8), achie es he alue o almos CZK 0.7 million pe company. The second mos ep esen ed clus e (3, 9) achie es he amoun o CZK 1.16 million pe company. The alue o he wo emaining clus e s is CZK 2.6 million (3, 10), o 4.5 million o clus e (4, 10). Assuming ha using i xed asse s is g owing wi h he size o company, i may be concluded ha he clus e s examined do no include la ge o ex emely la ge companies ac i e in ag icul u e o he Czech Republic. Ano he a iable examined is cu en asse s. Fo mo e de ails, see Fig. 4. Cu en asse s (al hough no equal o wo king capi al) a e an impo an pa o each company, in pa icula o an ag icul u al company. I includes in en o ies, ecei ables and i nancial asse s. Small animals, eed and ools cos s a e ela i ely high in ag icul u e companies. Simila ly, he paymen cul u e in he Czech Republic is no op imal; he e o e he le el o ecei ables is a he high. The highes mean alues o cu en asse s a e in ag icul u e as a whole (mo e han CZK 23.6 million). The companies in he clus e (1, 8) ha e he a e age cu en asse s alue o CZK 0.6 million. The alue o cu en asse s o clus e (3, 9) is less han CZK 2 million. The emaining h ee clus e s (3, 10), (4, 10) and (5, 10) include he companies wi h cu en asse s mean alues highe han CZK 2 million. Fig. 5 shows he amoun s o cash in he con ex . Fig. 5 shows ha an a e age ag icul u al company has cash a he amoun abo e CZK 141 housand. All companies in he examined clus e s ha e a ela i e highe amoun o money. Fig. 4: Mean alues o cu en asse s o ag icul u e and o selec ed clus e s Sou ce: own EM_4_2019.indd 109EM_4_2019.indd 109 25.11.2019 11:02:2725.11.2019 11:02:27 110 2019, XXII, 4 Business Adminis a ion and Managemen In clus e (5, 10), he amoun o cash is abo e he a e age (mo e han CZK 185 housand). No su p isingly, equally in e es ing a e he alues o unding sou ces – capi al. The selec ed liabili ies i ems and compa ison o hei mean alues a e shown in Tab. 3. A en ion should be also paid o he s uc u e o i nancing ag icul u e companies. The e o e, Fig. 6 shows he mean alues o equi y and bo owed capi al o each clus e and o ag icul u e. Fig. 6 shows ha in he ag icul u al sec o , he sha e o bo owed capi al is 33% on a e age. In Fig. 5: Mean alues o cash in ag icul u e and in selec ed clus e s Sou ce: own Fig. 6: Mean alues o equi y and bo owed capi al o ag icul u e and o selec ed clus e s Sou ce: own EM_4_2019.indd 110EM_4_2019.indd 110 25.11.2019 11:02:2725.11.2019 11:02:27 111 4, XXII, 2019 Business Adminis a ion and Managemen he case o he en i e se , he amoun o equi y and bo owed capi al is signi i can ly highe han CZK 70 million. The highes capi alized clus e in he examined se is clus e (5, 10). The e is a sligh p edominance o bo owed capi al. Bo h unding componen achie e he amoun highe han CZK 12 million, whe e he equi y accoun s o 25%. In he emaining clus e s, he e is p edominan ly bo owed capi al. The only excep ion is clus e (1, 8). An in e es ing compa ison can be seen in Tab. 4, which shows selec ed i ems o p o i and loss accoun . Ag icul u e as an ins i u ional sec o o na ional economy iden i i es a ce ain deg ee o sel -su i ciency and independence o na ional economy on o he economies. The ag icul u e pe o mance is less sensi i e o l uc ua ions in economic cycle. Howe e , i is s ill possible o judge he le el o economy acco ding o he ag icul u e pe o mance sha e o he o al GDP. I he absolu e alue is high and he ela i e alue is low, he economy is e alua ed as de eloped. This means ha no only he sales olume bu also pe o mance and alue added o ag icul u e companies is conside ed. I em In o al (1, 8) (3, 9) (3, 10) (4, 10) (5, 10) To al liabili ies 74,432.0 1,297.2 2,914.8 5,281.9 7,782.5 13,113.9 Equi y 51,225.2 1,062.6 1,213.0 1,420.2 1,334.9 3,311.0 Regis e ed capi al 15,768.1 746.6 522.5 588.5 1,545.3 3,201.6 Regis e ed capi al 2 15,226.3 455.1 167.6 266.3 1,321.8 2,913.8 Rese e unds, indi isible und and o he e enue ese es 8,460.5 106.2 102.8 268.9 229.9 496.5 Legal ese e / Indi isible und 3,898.4 32.5 19.0 177.2 128.3 143.9 S a u o y and o he unds 4,296.3 9.7 53.0 77.9 90.8 119.2 P o i / loss om p e ious yea s 8,590.4 -46.3 161.3 136.5 -1,029.5 -1,019.5 Re ained ea nings om p e ious yea s 10,195.7 179.1 339.5 494.9 446.0 1,951.1 Loss o he p e ious yea s be o e co e age -2,130.4 -311.6 -148.0 -464.4 -1,182.4 -2,832.6 P o i / loss o he cu en i nancial yea 3,340.1 10.5 131.9 130.4 8.0 91.3 Bo owed capi al 22,970.2 227.4 1,656.2 3,829.2 6,420.4 9,661.8 Long- e m liabili ies 5,422.9 53.4 350.6 1,018.3 1,826.7 2,948.5 Sho - e m liabili ies 7,765.0 145.8 1,064.1 2,124.2 3,396.4 4,708.6 T ade liabili ies (sho - e m) 4,286.0 11.4 170.3 431.7 1,029.8 1,667.6 Liabili ies o pa ne s, membe s o assoc. and pa icipan s in associa ion (sho - e m) 505.7 21.3 132.5 369.4 363.1 696.6 Liabili ies o employees 339.5 4.8 18.7 21.2 23.0 42.6 Liabili ies om social secu i y and heal h insu ance 202.9 2.1 9.6 20.5 37.2 29.0 S a e – ax liabili ies and subsidies 184.2 6.4 22.6 -24.8 79.8 65.9 Bank loans and bailou s 8,785.4 25.2 222.5 654.5 1,194.4 1,992.4 Bank loans – long- e m 6,478.1 6.4 50.6 199.3 462.7 835.0 Bank loans – sho - e m 1,756.2 1.1 15.7 90.8 94.7 362.7 Sou ce: own No e: The alues in he able a e gi en in housands CZK. Tab. 3: Mean alues o liabili ies o se examined and selec ed clus e s EM_4_2019.indd 111EM_4_2019.indd 111 25.11.2019 11:02:2825.11.2019 11:02:28 118 2019, XXII, 4 Business Adminis a ion and Managemen Abs ac PREDICTION OF INSTITUTIONAL SECTOR DEVELOPMENT AND ANALYSIS OF ENTERPRISES ACTIVE IN AGRICULTURE Voj ěch S ehel, Jakub Ho ák, Ma ek Vochozka The o e all EU ag icul u al p oduc i i y g ow h has slowed down in ecen yea s and has lagged behind leading global compe i o s, which is mainly due o dec easing numbe o employees in ag icul u e. Technical ine i ciency is hen an impo an phenomenon o he Czech ag icul u e and i s indi idual sec o s. Ag icul u e de elopmen should be es ablished on scien i i c bases. One o he basic p inciples o sus ainable ag icul u e is he e o e o ecas ing i s u u e de elopmen . In ecen yea s, se e al ag icul u al economis s ha e been engaged in compa ing o ecas s wi h a ious o he me hods and hei conclusions gene ally co espond o commonly accep ed belie s. A p esen , a i i cial in elligence can be de i ni ely ecognized as a use ul ool o business analyses and o ecas ing. The objec i e o he con ibu ion is an analysis o companies ac i e in ag icul u e o he Czech Republic using Kohonen ne wo k and he subsequen p edic ion o hei de elopmen . A da a se is c ea ed, which includes comple e da a om i nancial s a emen s o 4,201 companies ac i e in ag icul u e o he Czech Republic in 2016. The se o companies is gene a ed om he Bisnode Albe ina da abase. The da a se is subsequen ly subjec ed o clus e analysis using Kohonen ne wo k. Fo clus e analysis, Dell´s S a is ica so wa e, e sion 12 is used. The se is di ided in o h ee pa s: aining da a se , es ing da a se , alida ion da a se . Topological leng h and wid h o Kohonen ne wo k a e se a 10. The numbe o i e a ions is se a 10 000. Subsequen ly, he indi idual clus e s a e subjec ed o analysis o absolu e and selec ed indica o s (o mo e p ecisely, hei mean alues – a i hme ic a e age) and he esul s a e in e p e ed. I can be s a ed ha he ag icul u e companies show e y a o able alues – op imal asse s le el, accep able i nancing s uc u e and adequa e economic esul . I can be e en s a ed ha he indica o s show abo e-a e age alues compa ed o o he in es men op ions. Keywo ds: Ag icul u e en e p ises, ins i u ional sec o de elopmen , p edic ion, a i i cial neu al ne wo ks, alue o he business. JEL Classi i ca ion: C45, O13. DOI: 10.15240/ ul/001/2019-4-007. EM_4_2019.indd 118EM_4_2019.indd 118 25.11.2019 11:02:3025.11.2019 11:02:30