عنوان مقاله [English]
نویسندگان [English]چکیده [English]
Arbitrage pricing theory presented by Ross is based on theory of the absence of arbitrage opportunities in financial market and its main condition is the existence of a linear relationship between the actual return and a set of common factors, In this model, asset pricing is based on risk, although the risk source is not just one factor and its not only the market portfolio, But several factors affect the assets which they are called risk factors. In most APT studies, the researchers test the model through using two scales of beta coefficient and variance. But, the experimental evidences indicate the inefficiency of mean-variance framework, which means that stock returns can't be described well by the mean and variance. In this study, in order to identify the inefficiency of variance (standard deviation), for the first time the new standards of semi-variance and downside Beta in form of APT Model called downside arbitrage pricing theory (D-APT) were used.
This is an experimental quantitative research based on the regression of composite panels with an efficiency of 97 traded stocks in Tehran Stock Exchange as the dependent variable, and six macro-economic variables as the independent variables from 1384 to 1393; so that these companies are among the major stock industries that their activities have not been interrupted during this period and their fiscal year ends on 29/12 of every year.
Materials and methods
The dependent variable in the D-APT model is the annual stock return which is calculated in in the form of [min (R_i-R_F, 0)]. To calculate the stock returns derived from RAHAVARD NOVIN 3 software, the required adjustments for dividend, stock awards, priority, and stock splits were conducted. Independent variables are predetermined macro-economic factors involving: exchange rates, stock index, OPEC oil price, the consumer inflation rate and the deposit rate, which are calculated in this way [min (r_i ^ f-R_F, 0)]. To estimate the model, EVIEWS 9 software is used.
Results and discussion
In the present study, at first, the stability of the variables was examined which the results represented that all variables are stable. Based on the F-LIMER statistic, the model was judged whether it is pool or panel which the findings confirmed the existence of fixed effects in comparison to the least square approximation (to put it simply the findings approved panel data in comparison to the pool ones). After confirming the panel model, the Hausman test was used to determine whether the fixed effects model is more efficient or not. The Hausman test results also indicated that there is no need to use random effects model and fixed effects model should be used. Therefore, the model was estimated based on fixed effects and the final results reflected the negative correlation between the exchange rate and deposit rate variables and the return rate of stock companies, which means that the increase of exchange rate and deposit rate, decreases the return rate and vice versa, as well as the positive relationship between stock index, consumer inflation rate and oil price variables and the rate of return, meaning that with increase of the stock index, consumer inflation rate and oil price, the return rate increases and vice versa.
Finally, by estimation of D-APT model, it can be concluded that a significance of more than one factor confirms the APT model with a downside risk approach, and this means that more than one factor can explain changes in stock returns. In addition, 27.17 percent of the changes in stock returns is explained by these factors, while 72.83 percent of return changes can indicate the idiosyncratic risk that this amount is the volatility of stock returns in the portfolio of firm that is the specific variance to each firm which has the reasons unrelated to systemic factors, So we can conclude that a large amount of return variance of each portfolio firm has an uncertain cause that for this type of risk, the market does not pay any compensation to investment. Results of determination coefficient indicates lack of adequate performance of APT model based on the Downside Risk in Stock Market. Therefore, it is suggested to the investment companies, investment funds, institutional and real investors to mind both the upside and downside risk dimensions in their analyzes for buying and selling stocks, and also pay attention to the effects of counterproductive economic activities based on downside risk on stock returns.