Publication: Nitel bağımlı değişkenli panel veri modelleri ile İMKB’ de işlem gören hisse senetlerinin başarısının tahmini
Abstract
Ekonometrik araştırmalarda genellikle zaman boyutu veya kesit boyutu ile ilgilenilmektedir. Zaman boyutu ile ilgilenildiğinde birimsel farklılıklar göz ardı edilmekte, kesit boyutu ile ilgilenildiğinde ise zamana göre oluşan farklılıklar göz ardı edilmektedir. Yatay kesit verisiyle zaman serisi verisinin birleşiminden oluşan panel verilerde ise hem zaman ve hem de yatay kesit boyutu birlikte ele alınabilmektedir. Bu durumda, farklı modeller oluşturulabilmekte ve bunlar panel veri modeli türlerine göre farklılıklar gösterebilmektedir. Bu çalışmanın temel amacı, nitel bağımlı değişkenli panel veri modellerinin teorik incelemesinden sonra İstanbul Menkul Kıymetler Borsası'nda hisse senetleri işlem gören imalat sanayi sektöründe yer alan işletmelerden alınan bir örnek ile gelecek devrelerde hisse senedi getirisinin enflasyon oranından fazla olup olmayacağının tahminini yapmaktır. Çalışmada, farklı devreler için farklı panel veri modelleri denenerek en uygun model belirlenmeye çalışılmıştır. Panel veriler, panel doğrusal olasılık modeli, panel logit modeli, panel probit modeli, iki uçlu panel veriler, sabit etkiler, rassal etkiler.
In econometric researchs are generally dealt with time dimension or cross-section dimension. When in time dimension is omitted individual differences, in cross-section dimension is omitted the differences occured through time. As for panel data being with pooling of cross section and time series data, both time and cross-section dimension may be tackled together. In this case, different models could be developed and these migth show differences according to panel data kinds. Main goal of this study is to estimate that whether or not gains of sharies are more than inflation ratio in future periods after theoretical examination of qualitative dependent variable panel data models. The sample has been taken from the İstanbul Stock Exchange firm list among the firms operating in manufacturing industry. In this study, it has been worked on the determination of the most appropriate model by trying different panel data models for different periods. Panel data, panel linear probability model, panel logit model, panel probit model, binary panel data, fixed effects, random effects.
In econometric researchs are generally dealt with time dimension or cross-section dimension. When in time dimension is omitted individual differences, in cross-section dimension is omitted the differences occured through time. As for panel data being with pooling of cross section and time series data, both time and cross-section dimension may be tackled together. In this case, different models could be developed and these migth show differences according to panel data kinds. Main goal of this study is to estimate that whether or not gains of sharies are more than inflation ratio in future periods after theoretical examination of qualitative dependent variable panel data models. The sample has been taken from the İstanbul Stock Exchange firm list among the firms operating in manufacturing industry. In this study, it has been worked on the determination of the most appropriate model by trying different panel data models for different periods. Panel data, panel linear probability model, panel logit model, panel probit model, binary panel data, fixed effects, random effects.