Risks and Uncertainties in Agriculture

Transcription

1 Volume, Issue () ISS: -89 Rsks and Uncertantes n Agrculture Isabella SIMA Camela MARI Constantn Brancoveanu Unversty Faculty of Management Marketng n Economc Affars Ptest, Romana _onescu@yahoo.com Constantn Brancoveanu Unversty Faculty of Management Marketng n Economc Affars Ptest, Romana camelamarn8@yahoo.com ABSTRACT KEY WORDS JEL CODES Decsons optmzaton n agrculture assumed, n general, choosng the most approprate way to act n stuatons where management unt had no relable data on varable factors, or n other words, when decsons must be taken under rsk and uncertantes. In some agrcultural holdngs, a sgnfcant number of decsons were taken on emprcal nterpretaton of the nformaton crculatng when desgnng decson, about the state of nature, whch often led to a poorqualty decsons, whch could not generate optmal results n any case. Decson-makng models dd not create new nformaton but had the great mert of transposng n an approprate manner the exstng one, systematcally presentng a certan stuaton. Wth the help of some mathematcal methods, based on ths nformaton, there could be defned the types of natonal behavor (for example, to mnmze the largest loss that could occur f nature would manfest n the most hostle way), so that to choose the most approprate strategy. In the present paper there was dscussed the case n whch the decsonal matrx had an equlbrum pont, the optonal choce beng made usng crtera based on the crcumstances and the level of herarchy of the decson maker. condtons, decsons, rsk, uncertanty D8, D8. Introducton Tradtonal Economcs descrbes a statc unverse, but new realtes and developments requre the explanaton of some changng stuatons. A socal system that produces economc values s characterzed prmarly by uncertanty. Moreover, uncertanty and rsk do not consttute an optonal factor, but they are a defnng aspect of the human condton. The purpose of ratonalty conssts not only n avodng the rsk and elmnatng uncertanty, as to adapt to rsk an uncertanty, acceptng stuatons as data. In the economc unts and n the agrcultural ones n partcular (egură, Găburc and Dmulescu, 97), a great number of decsons are taken under uncertanty. If the feasbltes of

2 Volume, Issue () ISS: -89 the state of nature are not known, decson makng process occurs under uncertanty and the best verson of mathematcal hope can not be used as gudance n choosng. To ensure the development of vable decson versons and the proper evaluaton of each one, decson makers must show a hgh degree of competence, so that the results obtaned from the applcaton of a verson n practce (real consequences) to not dffer from those evaluated n accepted proportons. The bgger the dfference between real results and the assessed ones, the more t s requred that wthn the shortest possble tme to be ntated correctve decsons. To solve an approprate model, two stages are requred: frst, the examnaton of decson matrx (to see f t has an equlbrum pont or not), and then choosng the solvng crtera of decson matrx. The study presented n the paper analyzes the decson that s requred to be taken regardng the ssue of feed seedng n dual crop at a company wthn a complex agrcultural holdng.. Model form In model developng, the frst phase s to buld the matrx and the second phase s represented by the use of a table covered by certan methods of calculaton, n order to reach optmal strategy. Model s elements are: Avalable strateges, the alternatves of the decson center, whch must be dstnctly denoted by: V,V,...,V,...,V m,, m ; States of nature, denoted by,,...,,..., j n, j, n, whch can be favorable or unfavorable, n a greater or smaller measure Placng n a table the m strateges on lnes and the n states of nature on columns, and at the ntersecton box of rows wth the columns the expected results, t s obtaned the matrx. Applyng the approprate decsons to allow the use of matrx for choosng the optmal strategy depends on the n formaton we have regardng the probablty of achevng the states of nature. The ablty to predct the occurrence of state depends on the state of knowledge of the j system s uncontrollable factors, and also depends on a large number of precedents and heavy volume of data. We consder that there are three possble (V) strateges: V - sowng the corn feed n double crop of an area of ha; V - sowng the corn feed n double crop of an area of ha; V - no double-seeded crop. These strateges of the decson maker (head of the farm) can brng dfferent economc results dependng on the states of nature (). In ths case, we consder as states of nature the weather forecast, gven the amount of precptaton: - There s a certan water supply n the sol and suffcent ran falls durng the optmal sowng tme; There s no water supply n the sol, but t rans enough durng the optmal sowng tme; - There s no water supply n the sol, and does not ran durng the optmal sowng tme or after (drought);

3 Volume, Issue () ISS: -89 There s a water supply n the sol whch s consdered to be suffcent for the obtanng 4 of crop even wthout any further precptaton after the sowng tme (ths requres the applcaton of specal works to mantan water n the sol). There s no water supply n the sol, t doesn t ran durng the sowng tme, but t rans n about - weeks of sowng. By estmatng corn feed producton n double crop whch can be obtaned for each of the three strateges and fve states of nature, t s made the consequences matrx. The consequences are expressed n thousands le (RO) benefts or losses, and recorded for each case, at the ntersecton between the lne whch represents the strategy and the column whch represents the state of nature. Snce we don t know whch wll be the states of nature at the moment of decson makng regardng the sowng of the double crop, means that we have to take a decson under uncertanty. Table. The values correspond to those strateges and to those states of nature 4 V [ ha] V [ ha] -4 V [ ha] We wll use the fve rules of decson (Albc, Teselos, Tenovc, ): a) Prudence crteron (the rule of Abraham Wald). Applyng ths crteron to the gven matrx, we obtan: maxmn,8, 7,6,,mn,, 4,,,mn 4,,,, max 7, 4, 4 4 V V b) Optmstc crteron: It s preferred the verson whch leads to the hghest payment throughout the matrx, regardless of negatve consequences that mght occur. Accordng to the formula we get: max,, V V We notce that the V verson allows the obtanng of a maxmum beneft of thousand le for the state of nature, but may be resultng n the greatest loss (- 7 thousand le) f the state of nature wll occur (there wll be drought). c) Weghted optmsm crteron: Consderng α, as a socal medum value of the optmstc degree results: max,, 7 ;,, 4 ;,, 4 max ; 7,, ;, max;6;, V V d) Mnmum regret crteron (Savage): The Savage Method requres the calculaton of regrets matrx:

4 Volume, Issue () ISS: -89 R j And wll mnmze the largest antcpated regret: mnmax,,4,, ;max 4,,8,, ;max4,,,6,7 mn4;;4 4 V V e) Laplace crteron. We obtan: max ; ; 9 max ; ; V V We notce that the V verson s not approved by any crtera and thus s out of the queston. It remans to be decded between the alternatves V and V. Whle the majorty rule may not be relevant n such matters, the V verson can stll be proposed for mplementaton. In case n whch analyzng statstcal data of past perods nvolves determnng probablty of states,,,, as follows: 4, p, ; p ; p, p, ; 4 p,, Choosng the optmal verson wll be made by usng mathematcal expectancy. Mathematcal expectancy of some economc actvtes results s the average weghted sze of the actvty results, the weghts beng equal to the events probabltes or states of nature. In case of a V verson of solvng a problem, characterzed by random events, these are featured by x, the probablty of ther producton p x j j and the obtaned results r, mathematcal j expectancy of results n V verson denoted S V, wll be gven by the formula: S V r px,, m p x,, j j j j px. j j If there are economc effects expressed by rj results, then the optmal value wll be the one that satsfes the condton: max S V V, 4

5 Volume, Issue () ISS: -89 Thus: S V,,8,7,6, 6, 9,,7 6,7 Thousands le S V,,, 4,,,7 66,4 97, Thousands le S V, 4,,,,, 6,6,7 6,6 Thousands le It s therefore recommended to be chosen the V verson. Ths corresponds to the applance of the condton of the three mathematcal expectaton calculated above n order to be maxmum.. Conclusons The essental purpose of solvng problems must ensure consstency between performances levels and the possbltes of superor captalzaton of the avalable resources. Optmzaton actvty of relatons between decson maker s objectves and the exstng resources must be addressed sequentally, startng wth the branches of producton, crop structure and lvestock and endng wth the sale of agrcultural products. Economc-mathematcal models as knowledge tools (Raţu-Sucu, 99), used n the preparaton of decsonal versons, wll have to capture the whole complex of factors and the nteractons between them, so that, by applyng the decson verson, the actual results to not dffer from the evaluated ones. Gven the large number of decson problems that are addressed n terms of rsk and uncertanty, the use of specfc methods of substantatng decsons wll contrbute to the mprovement of decson process n economc unts. References. Albc, M., Teselos, D., Tenovc, C. (). Decsons n Rsk and Uncertanty Condtons, The 6th Internatonal Conference The Knowledge Based Organzaton, colae Bălcescu Land Forces Academy Publshng House, Sbu, Conference Proceedngs, pp Albc M., Belu., Ţenovc C. (9). Economc decsons n uncertanty condtons, Unversty Lbrary of Munch, Germany, Belu., Albc M. (9). Economc decsons n rsk condtons, Unversty Lbrary of Munch, Germany, 4. Coşea M., astovc L. (997). Rsk assessment. Methods and technques of analyss at mcro-and macro-economc, Lux Lbrs Publshng House, Brasov.. Mărăcne V. (998). Manageral decsons. Improvng company's decson-makng performances, Economca Publshng House, Bucharest. 6. egură I., Găburc A., Dmulescu S. (97). Operatonal research n agrculture, Ceres Publshng House, Bucharest. 7. Raţu-Sucu C. (99). Modelng and smulaton of economc processes, Ddactcal and Pedagogcal Publshng House, Bucharest. 8. Raţu Sucu C. (99). Economc modelng, Sylar Publshng House, Bucharest.