The use of bivariate cointegrated vector autoregressive models (CVAR) and Baba-Engle-Kraft-Koroner (BEKK) models is suggested for the selection of a stock portfolio (Markovitz type portfolio) based on estimates of average returns on shares and volatility of share prices. In other words, we adderss the problem of esitimating average returns and the associated risk on the basis of the prices of a certain number of shares over time.The estimate is then used to identify the assets offering the best performance and hence costituting the best investments. The use of VAR(1) model is a common practice; we suggest instead VEC models, which make it possible to take into account any cointegration between the series employed and the market trend as maesured by means of Equity Italy Index. The model put forward is applied to a series of data regarding the prices of 150 best capitalized shares traded on the Italian stock market (BIT) between 1 January 1975 and 31 August 2011; it takes into account the intrinsic value of the stocks to select the best performing ones. Moreover, we find the efficient portfolio minimizing the risk, that is to say analyzing the efficient frontier.
Naccarato, A., Pierini, A., Reale, M. (2012). The combined use of CVAR and BEKK models for portfolio selection of Italian stock-market. In Book of abstract CFE 2012 6th CSDA International Conference on Computational and Financial Econometrics (pp.51). Oviedo : Department of Statistics, OR and DM, University of Oviedo, Spain.
The combined use of CVAR and BEKK models for portfolio selection of Italian stock-market
NACCARATO, ALESSIA;
2012-01-01
Abstract
The use of bivariate cointegrated vector autoregressive models (CVAR) and Baba-Engle-Kraft-Koroner (BEKK) models is suggested for the selection of a stock portfolio (Markovitz type portfolio) based on estimates of average returns on shares and volatility of share prices. In other words, we adderss the problem of esitimating average returns and the associated risk on the basis of the prices of a certain number of shares over time.The estimate is then used to identify the assets offering the best performance and hence costituting the best investments. The use of VAR(1) model is a common practice; we suggest instead VEC models, which make it possible to take into account any cointegration between the series employed and the market trend as maesured by means of Equity Italy Index. The model put forward is applied to a series of data regarding the prices of 150 best capitalized shares traded on the Italian stock market (BIT) between 1 January 1975 and 31 August 2011; it takes into account the intrinsic value of the stocks to select the best performing ones. Moreover, we find the efficient portfolio minimizing the risk, that is to say analyzing the efficient frontier.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
