The Financial Accounting and Auditing Researches
Year:2017 | Volume:8 | Issue:32 |Papers Count:8

Present study investigates the usefulness of ensemble regression and feature selection methods (including correlation-based feature selection and Relief) in predicting stock returns of companies listed on Tehran Stock Exchange. For performance evaluation of ensemble regression, evaluation criteria (including mean absolute percentage error, root mean squared error and coefficient of determination) of this method compared with linear regression and artificial neural networks. Also, for performance evaluation of feature selection methods, evaluation criteria of these methods compared with using all variables. The experimental results of investigating 101 companies listed in Tehran Stock Exchange in 2004-2013 indicate that ensemble regression outperforms the linear regression and artificial neural networks. Furthermore, the results show that selected variables with correlation-based feature selection and Relief result in better prediction in compare with using all variables.

The purpose of this study is to evaluate the relationship between principal suppliers and customers for goods and services with the firms dividend policy. In this paper we will describe how major suppliers and customers affect corporate dividend policy. The present research based on objective, Practical, and based on method is descriptive and correlation. The test of research hypotheses have been done with multiple regression analysis by using the method of panel least square (PLS). Statistical sample considered in this research consists of 99 companies of among the list of Accepted Companies in Tehran Stock Exchange. The Research findings reveals that if company has bulk buying, in other words, purchase the own raw materials in bulk, has the ability to distribute lower cash dividends and conversely, if has major customers, and carry out the own products sales in bulk, can be basis continuous and annual to the distribution of profits to pay and satisfy the own shareholders' need.

Increases in intangible investments reduce earnings quality for several reasons. One reason is high volatility in revenues and cash flows that carry higher uncertainty about future benefits than do tangible investments. In this paper earnings quality dimensions compare according to two kind of industries that include knowledge-based industries (industries that involve higher intangible costs as R & D) and another industries. In this research we use two scale of earnings quality that include matching and earnings volatility. Therefore we consider 342 firm-year observations from the years 2009 through 2014. Consequences indicate that earnings quality of knowledge-based industries is lower than other industries. Likewise market-to-book ration that intend as intangible intensity scale had significant and very high correlation with selling, general and administrative expense volatility. Main contribution this research is consideration of earnings quality on base knowledge-based industries and other industries that such division didn’t implement in prior studies.

The purpose of this article is to measure and compare the future value of maintenance portfolio in short periods of time due to maximum efficiency and minimum risk is that Portfolio until the investors and Stock brokers with respect to the anticipated evaluate your decisions, So to calculate and evaluate the default investment fund portfolio, the value at risk analysis from GARCH and ARCH model and Monte Carlo simulation technique was used. For this purpose, based on information collected from the portfolio fund investments Tehran Stock Exchange, initially portfolio returns using Monte Carlo simulation techniques was simulated in Crystal ball software, and then considering the 95% confidence level, the VaR of a portfolio of 15 investment fund is calculated. Then, using GARCH and ARCH models and compare the results also showed that not only the value of the Fund's portfolio exposed to default risk by extrapolating these models reliable calculations using parametric and nonparametric techniques can be purely historical but a combination of both methods can be closer to reality. Because where both pessimistic and conservative approach models and simulations can be used both optimistic approach.

This paper is about identifying and diagnosing earnings management by using changes in assets turnover (ATO) and profit margin (PM) in Tehran Stock Exchange. To perform the research, first the required data have been calculated using cross-sectional multi-variable linear regression method. Then, to get the final models and results hypothesis have been tested. The financial information of 142 companies is used. The results show that decrease in PM and contemporaneously increase in ATO is an indication of downward earnings management but, increase in PM and contemporaneously decrease in ATO doesn't confirm upward earnings management

The way in which risk is measured and involved poses a challenge to stock valuation models. The present paper aimed at comparing the effectiveness of two different models for measuring the risks. In the former, based on the Accounting Measures of Risk Model, covariance of fundamental characteristics of the firm including the accounting earnings, the excess return on equity together with relevant market factors were introduced as risk adjustment factors into the valuation model and, then combined with the risk-free present value. In the latter model, by means of the reward beta approach the risk was measured on the basis of return variation. The results from comparing the forecasted returns with the ex-post returns were indicative of the efficiency of both models in Tehran Stock Exchange. Other results, however, pointed to more precision of fundamental risk factors model than the reward beta approach in dividend return forecasting.

This paper investigates the effect of audit quality characteristics on stock price crash risk on accepted firms Tehran Stock Exchange. Information required for the study of the financial statements of 81 firms in the period 1389-1393 was collected. In this study audit quality through criteria auditor industry specialization, auditor tenure, auditor independence and audit firm size is measured. These results of the test hypothesis study showed that the relationship audit quality characteristics and stock price crash risk are negative significant. So that Increase audit quality will lead to the loss stock price crash risk. In other words, better oversight and work to prevent opportunistic by management and the accumulation of bad news, leading to increased shareholder value by timely pricing.

Agency theory suggests that agency costs are resulting from conflicts of interest between managers and shareholders that might cause consequences such deviations from the optimal investment behavior. Thus, it is likely that managers have different investment behavior where the agency costs exist. Hence the present study examines the impact of agency costs on investment behavior in Tehran Stock Exchange. Thus, sampling deletion, numbers of 110 companies was chosen. We use efficiency ratios for measuring agency costs and to estimate the investment behavior (over and under investment) the Biddle et al (2009) model have been used. The panel data model was used to test the hypothesis. The results show that agency costs are causing deviations from optimal investment behavior. Thereby increasing agency costs, related problems with over and under investment will intensify and investment decisions will affect companies. Also, the results showed that the impact of agency costs on the investment behavior in over and under investment firms is different.