Beta katsayılarının ekonometrik analizi
Başlık çevirisi mevcut değil.
- Tez No: 55508
- Danışmanlar: DOÇ.DR. BURÇ ÜLENGİN
- Tez Türü: Yüksek Lisans
- Konular: Mühendislik Bilimleri, İşletme, Engineering Sciences, Business Administration
- Anahtar Kelimeler: Belirtilmemiş.
- Yıl: 1996
- Dil: Türkçe
- Üniversite: İstanbul Teknik Üniversitesi
- Enstitü: Fen Bilimleri Enstitüsü
- Ana Bilim Dalı: Belirtilmemiş.
- Bilim Dalı: Belirtilmemiş.
- Sayfa Sayısı: 240
Özet
ÖZET Kalkınmış veya kalkınma sürecinde olan tüm ülkelerin temel sorunu kaynak yaratmak, mevcut kaynakları etkin kullanmak ve kalkınmayı yeterli düzeyde finanse edebilmektir. Borsalar ekonomiye kaynak aktarılmasında ve sermayenin geniş kitlelere yayılmasında önemli işlevleri üstlenen kurumlardır. Ocak 1986'da faaliyetine başlayan İstanbul Menkul Kıymetler Borsası da sermaye piyasasının vazgeçilmez bir unsuru haline gelmiştir. İşte günlük hayatımızda önemli bir yeri olan îMKB'nda işlem gören 22 şirkete ait hisse senedinin getirilen ve riskleri arasındaki ilişki, yapılan çalışma esnasında incelenmiştir. Finasal Varlık Fiyatlandırma Modelinin (FVFM) temelini oluşturduğu çalışmamızda, yatırımcıların hisse senetleri üzerinde karar vermelerini kolaylaştıracak risk ölçütleri olan beta katsayıları hesaplanmış ve bu katsayıların güvenilirlikleri kontrol edilmiştir. Çalışma sırasında modelimizde kullandığımız veriler 1 Eylül 1990 - 31 Aralık 1994 tarihleri arasındaki 52 aylık dönemi kapsamaktadır. Bu veriler TSP adlı istatistik paket programına girilerek regresyon yapılmıştır. Çalışmada kullanılan veriler hisse senetlerinin aylık getiri oranları, IMKB bileşik endeksi ve aylık ortalama hazine bonosu faiz oranıdır. Birinci bölümde konuya giriş yapılmış, ikinci bölümde FVFM ile ilgili temel tanım ve kavramlara yer verilmiştir. Üçüncü bölümde risk kaynakları ve beta katsayıları hakkında bilgi verilmiş, dördüncü bölümde beta katsayılarının durağanlığı üzerine yapılan bir çalışmada izlenen yöntem ve bulunan sonuçlara kısaca değinilmiştir. Beşinci bölümde veriler testler hakkında bilgi verilmiş, altıncı bölümde araştırmada kullanılan testler kısaca açıklanmış, hisse senetlerine ait regresyon doğrusunun alfa ve beta katsayıları hesaplanmış ve test sonuçlan her senet için yorumlanmıştır. Bu arada katsayıların tahmin edilmesi aşamasında karşılaşılan sorunlara da değinilmiştir. Kurmuş olduğumuz modellerdeki değişkenlerle ilgili veriler Önder Menkul Değerler A.Ş. ve Akbank Hazine Bölümü Muhasebe ve Kontrol Müdürlüğü arşivinden yararlanılarak elde edilmiştir. vııı
Özet (Çeviri)
SUMMARY THE ANALYSIS OF STABILITY OF BETA COEFICIENTS OF 22 STOCKS TRANSACTED IN THE ISTANBUL STOCK EXCHANGE MARKET In applied statistical analysis, the necessity arises to examine conditions of stationarity in individual sample records of random data. Stationary data represent any class of data whose statistical properties do not change with time. In most applications of time series analysis, spectral analysis, and random process theory, examination of the stationary data is required before meaningful investigation can be performed. If the sample records of random data are nonstationary and cannot be transformed to a stationary form, or if suitable measures of statistical properties cannot be found, useful descriptions of the underlying process become impossible. The purpose of this study is to demonstrate how a relatively simple test for stationarity can be applied to a given set of data. The specific data considered in this application are beta coefficients. One of the most sought-after possessions for a typical private investor or securities market analyst is a reliable equation predicting the return on alternative securities. A first step in developing and empirically implementing such an equation involves gaining an understanding of why a particular stock has a low or high rate of return. Capital Asset Pricing Model (CAPM) is a model that helps in developing such understanding. A remarkable feature of CAPM is that its most important parameters can be estimated by using very simplest of econometric techniques, namely, a linear model in which a dependent variable is regressed on a constant and a single independent variable. The simple CAPM therefore provides a useful introduction to empirical econometrics. To derive the CAPM we must first list a few assumptions concerning the investors and the securities market: 1. Investors are risk-averse. 2. The CAPM is a single-period model because it is assumed that investors maximize the utility of their end-of-peri od wealth. 3. All investors have the same efficient frontier; that is, they have homogeneous expectations concerning asset returns and risk. 4. Portfolios and securities are characterized by their means and variances. 5. There is a risk-free asset with an interest rate Rf. This rate is IXavailable to all investors to either borrow or lend. The borrowing rate is equal to the lending rate. 6. All assets are marketable and perfectly divisible. 7. There are no transaction costs. 8. Investors have all information available to them at no cost. 9. There are no taxes and regulations associated with trading. Given these assumptions the CAPM is defined as E(Rit is the random error term. This regression equation describes the risky asset's return relationship with the market portfolio return. It is also called the characteristic line and is a statistical tool employed to measure systematic and unsystematic risk. The term at, is called the alphacoefficient for security i. This is the intercept point where the characteristic line intercepts the y-axis. Since beta is the slope coefficient for this regression (market model), it demonstrates how responsive returns for the individual assets are the market portfolio. Alternatively, a market model can be defined as follows: Ri.t - R-p.t = ffii + Bi (Rm,t - Rf.t) + 6i,t This is the risk premium form of the market model. To estimate the B parameter based on time series data of individual companies, it must be assumed that, for a particular company B is relatively stable over time. Econometric studies based on monthly data have found that in many studies B has tended to be relatively stable over a five-year (60-month) time span. There are cases, however, in which the conditions in an industry or a firm abrubtly change, implying that the relevant B might also vary. Oil company stocks for example had a beta below unity before the 1973 OPEC oil ambargo, but this soon changed after 1973, and since then oil company betas have typically increased. Similarly, when the airline industry in the United States was deregulated in 1978, the betas of most major U.S. airline companies rose; analogous changes in betas occured for electric utilities, particularly for those with substantial nuclear generating capacity, after the Chernobyl nuclear power accident in 1986. In this thesis the stationarity of beta coefficients of 22 stocks of largest companies in the Istanbul Stock Exchange Market are examined. These companies are selected from different industries as much as possible to determine if industry effects exist. The companies and the industry they belonged are shown below: Name of The Company 1. Adana (A) 2. Alarko 3. Aksa 4. Arçelik 5. Çukurova 6. Deva Holding 7. Eczacı başı 8. Ege Brewery 9. Erdemi r 10. Kartonsan 11. Kepez 12. Koç Holding 13. Kordsa 14. Mardin 15. Migros 16. Petrol Ofisi 17. Pınar Flour 18. Sarkuysan 19. Siemens 20. Tire Kutsan 21. T. İş Bank (C) 22. Yapı Kredi Bank Industry Cement Metal components, machine manufacturing Textile Metal components, machine manufacturing Electric utilities Chemistry Chemi stry Food and beverages Iron and steel Forest products Electric utiliti es Holding Textile Cement Foods and beverages Petroleum Foods and beverages Iron and steel Metal components, machine manufacturing Forest products Banking Banking XIThe time period covered is from September 1990 to December 1994. The betas for the individual securities are estimated by using the risk premium form of the market model. The data are the daily prices of securities, monthly rate of return for treasury bonds and daily IMKB compound index. For each of the companies the following hypotheses are tested: 1. H0 : ae = 0 Hi : a * 0 2. H0 : B = 0 Hi : B * 0 3. H0 : a = 0, B = 1 Hi : one of them is different 4. H0 : The equation is meaningful Hi : The equation is not meaningful 5. H0 : There is serial correlation Hx : There is not serial correlation 6. H0 : There is autocorrelation Hi : There is not autocorrelation 7. H0 : The distribution of residuals is normal Hx : The distribution of residuals is not normal 8. H0 : There is heteroscedasticity Hi : There is not heteroscedasticity 9. H0 : The function used in estimating the equation is true Hi : The function used in estimating the equation is false 10. H0 : The model is reliable in estimation Hi : The model is not reliable in estimation 11. H0 : The coefficient vector is stable Hx : The coefficient vector is not stable Output of each test is shown in the attachments. Among these tests the most important are those for autocorrelation and heteroscedasticity. In the presence of autocorrelation, OLS estimates and forecasts based on them are still unbiased and consisted, but they are not blue and hence are inefficient. In the presence of heteroscedasticity some of the properties of the estimators are altered. OLS estimates are still unbiased and consisted. Forecasts based on them are also unbiased and consisted. Estimates and forecasts, however, are inefficient and hence are no longer blue. Because the estimated variances and covariances of the estimates are biased and inconsistent, tests of hypotheses are not valid anymore. xnAmong the equations we have estimated for 22 companies, 13 of them involved neither autocorrelation nor heteroscedasticity, but 9 of them has come out with problems. Five of these equations had autocorrelation, three of them had heteroscedasticity, five of them had both of these problems. To correct for autocorrelation, ARIMA model and dummy variables are used. For heteroscedasticity, variables are redefined and all tests are applied to the reestimated models. For four equations with autocorrelation problem, it is accepted With % 95 reliability that beta coefficient is equal to 1. Only for the equation of Çukurova it is accepted that beta is = 0. But after autocorrelation is corrected, the result for Çukurova has changed and it is accepted that beta is between 0 and 1. Among the equations with heteroscedasticity problem, only for Ege, beta is accepted to be equal to unity. For four equations with both problems, it is accepted that beta is equal to 1 and after correction this result has not changed. Only for Kordsa beta is accepted to be equal to 0 before correction. But after correction Kordsa is accepted to have a beta coefficient between 0 and 1. The companies are classified according to the problems faced in their equations as follows: Autocorrelation and Autocorrelation Heteroscedast i ci ty Heteroscedast i ci ty Aksa Arçel i k Kordsa Alarko Ege Mardin Çukurova Kartonsan Petrol Ofisi Eczacıbaşı T.îş.Bank Ereğli Yapı Kredi Bank xı n
Benzer Tezler
- Takipteki kredilerin makroekonomik belirleyicileri: Moğolistan bankacılık sektörü üzerine bir uygulama
Macroeconomic determinants of non-performing loans: A case of Mongolia banking sector
BAYARMAA BATJARGAL
Yüksek Lisans
Türkçe
2019
BankacılıkAnkara Hacı Bayram Veli ÜniversitesiBankacılık Ana Bilim Dalı
PROF. DR. AFŞİN ŞAHİN
- Time varying beta estimation for Turkish real estate investment trusts: An analysis of alternative modeling techniques
Türk gayrimenkul yatırım ortaklıklarında değişken beta tahmini: Alternatif modelleme tekniklerinin analizi
GÖZDE ALTINSOY
- Beta and beta trend estimates in Istanbul Stock Exchange
İstanbul Menkul Kıymetler Borsası'nda beta ve beta yönseme kestirmeleri
ELİF SAÇBAĞLI
- Zamanla değişen beta katsayısında uzun bellek davranışı: BİST üzerine ampirik bir uygulama
The long memory behavior in time-varying beta coefficient: an empirical application on BIST
HÜSEYİN KESKİN
Yüksek Lisans
Türkçe
2019
BankacılıkBurdur Mehmet Akif Ersoy Üniversitesiİşletme Ana Bilim Dalı
DOÇ. DR. İSMAİL ÇELİK
- Portföy yönetiminde sistematik riskin ölçülmesi ve İ.M.K.B. için bir uygulama
Calculation of systematic risk in portfolio management and a case study in İstanbul Stock Exchange
UMUT CEMİL AKINCI
