Shapour Mohammadi - Academia.edu (original) (raw)

Papers by Shapour Mohammadi

SSRN Electronic Journal, 2015

The objective of this research is to apply the recently developed fractional cointegration vector... more The objective of this research is to apply the recently developed fractional cointegration vector autoregressive (FCVAR) model in developing pairs trading model and compare the results to that of the cointegration approach. To the best of our knowledge, there are no papers on fractional cointegration approach to date. Hence, this paper would be the first practical research based on fractional cointegration on futures market.As the focus of this paper in on short-term trading positions, minute intraday interval prices are used to estimate the model and execute the trading strategies. First, we form two portfolios, “FCI” and “CI”. FCI is created using fractional cointegration and CI using cointegration approach. Then, using an estimation window of specific length, fractional cointegration test and cointegration test are preformed to verify the existence of fractional cointegration and cointegration relationship between the selected pair respectively. At last, the long-term equilibrium is estimated using Fractional Vector Error Correction Model (FVECM) for FCI and Vector Error Correction Model (VECM) for CI and based on this estimation the trading strategy is implemented. The results show that the fractional cointegration approach yields statistically significant higher returns than the cointegration approach. The number of trades is also lower for fractional cointegration approach, which further contributes to the performance of the strategy. Finally, on a risk-adjusted level, the fractional cointegration approach has a higher Sortino ratio compared to the cointegration approach. Furthermore, window sizes are optimized and the best and worst window sizes for both the cointegration and fractional cointegration estimation window and trading window is reported.

Statistical Software Components, 2020

This code tests the statistical significance of inputs by using an artificial neural network with... more This code tests the statistical significance of inputs by using an artificial neural network with a flexible structure. The code tries to find the best number of neurons for the neural network for testing the significance of inputs.

Taḥqīqāt-i mālī, 2013

Probability of Private information Based Trade (PIN) has introduced as information risk measure. ... more Probability of Private information Based Trade (PIN) has introduced as information risk measure. This paper is going to estimate probability of private information based trade (PIN) in Tehran Stock Exchange using microstructure models. Our results show that PIN is significantly different from zero for Tehran Stock Exchange.

Statistical Software Components, 2009

Рассматриваются содержание и объективные предпосылки применения контроллинга в управлении затрата... more Рассматриваются содержание и объективные предпосылки применения контроллинга в управлении затратами на качество на предприятии. Определены категории общих затрат на качество на начальной стадии внедрения контроллинга, а также необходимость в разработке формы отчётности о расходах на качество, отвечающих требованиям предприятий различных отраслей экономики.

This function tests harmful multicollinearity based on the difference between p-value of estimate... more This function tests harmful multicollinearity based on the difference between p-value of estimated coefficients of ordinary least squares and generalized ridge regression estimators.

International Journal of Accounting Information Systems, 2017

This paper concentrates on the effectiveness of using a hybrid intelligent system that combines m... more This paper concentrates on the effectiveness of using a hybrid intelligent system that combines multilayer perceptron (MLP) neural network, support vector machine (SVM), and logistic regression (LR) classification models with harmony search (HS) optimization algorithm to detect corporate tax evasion for the Iranian National Tax Administration (INTA). In this research, the role of optimization algorithm is to search and find the optimal classification model parameters and financial variables combination. Our proposed system finds optimal structure of the classification model based on the characteristics of the imported dataset. This system has been tested on the data from the food and textile sectors using an iterative structure of 10-fold cross-validation involving 2451 and 2053 test set samples from the tax returns of a two-year period and 1118 and 906 samples as out-of-sample using the tax returns of the consequent year. The results from out-of-sample data show that MLP neural network in combination with HS optimization algorithm outperforms other combinations with 90.07% and 82.45% accuracy, 85.48% and 84.85% sensitivity, and 90.34% and 82.26% specificity, respectively in the food and textile sectors. In addition, there is also a difference between the selected models and obtained accuracies based on the test data and out-of-sample data in both sectors and selected financial variables of every sector.

Taḥqīqāt-i mālī, 2017

This paper analyses whether joint probability distribution function of losses due to different ex... more This paper analyses whether joint probability distribution function of losses due to different exposures covered under the same policy could be modeled in an appropriate manner via mixture distribution proposed and copula concept. <br />Special type of distribution which is a mixture of Generalized Hyperbolic Skew t distribution and Extreme Value theory (EVT) has been used for modeling marginal distributions of claims and copula function has been considered as a means of modeling dependency structure among claims. Most important copula including; Gaussian, t, Frank, Gumbel and Clayton was tested from goodness of fit point of view. <br />The data used in this study are the amount of property damage and bodily injury covered under automobile liability insurance. <br />Results reveal that joint probability distribution of claims could be effectively modeled by Clayton copula and proposed mixture distribution.

During the past few decades, there have been many evidences to believe that the stock markets aro... more During the past few decades, there have been many evidences to believe that the stock markets around the world follow cyclical trends. In this paper, we study the cyclical trends using wavelet function based on various time windows on some major stock market indices. We use two methods of Daubechies and reverse bi-orthogonal wavelet methods and determine the optimal values of both methods. The results are used for Tehran stock exchange using the most recent ten years daily information as an empirical study. The details of our analysis on TEDPIX index for the last decade indicate that there are, at least, four trends of weekly, monthly, quarterly and yearly and the cycles would be expected to be repeated in future.

SSRN Electronic Journal, 2004

ABSTRACT Testing fixed and random effects is one of peractical problems in panel estimations. Thi... more ABSTRACT Testing fixed and random effects is one of peractical problems in panel estimations. This program tests fixed and random effects for user defined models.

SSRN Electronic Journal, 2005

ABSTRACT This program helps to identification and aytomatic forecasting with ARMA models for fore... more ABSTRACT This program helps to identification and aytomatic forecasting with ARMA models for forecasters and analysts.This program is compatible with EViews 3,3.1,4,4.1. For running it in EViews 5 and 5.1 simply check the box Version 4 compatible variable substitution.

SSRN Electronic Journal, 2004

ABSTRACT This program tests multicolinearity with condition number and variance inflation factor ... more ABSTRACT This program tests multicolinearity with condition number and variance inflation factor (VIF) for given set of explanatory variables. It report high multicolinearity if condition number (cond) be greater than 20(or VIF greater than 10).

Business cycles analysis is one of the most important tasks for economists and statisticians. Var... more Business cycles analysis is one of the most important tasks for economists and statisticians. Various methods such as HP filter, BP filter, CF filter spectral analysis and Fourier transformation are used in time series analysis. All of these methods are sensitive to stationarity of time series. Also the traditional methods don't offer information in scale analysis. In this paper we introduce the Wavelet theory and its applications especially to economics .Also decomposition of seasonal GDP of Iran and analysis of business cycles are the other main propose. Results of GDP decomposition show 7 cycles with length of 16-32 quarters and 13 cycles with 8-16 quarters. Volatility analysis implies no change in variance of Wavelet coefficients in prewar and war periods .However volatility of GDP increased in the post war period. JEL Clarification: C14, C22, C19, E32. One way of solving welfare loss asymmetric information in insurance is to design informationally consistent contracts. When...

Department of Environmental Science, Science and Research Branch, Islamic Azad University, Tehran... more Department of Environmental Science, Science and Research Branch, Islamic Azad University, Tehran, Iran 2 Faculty of Environment, University of Tehran, P.O.Box 14155-6135, Tehran, Iran Faculty of Management, Tehran University School of Management, Tehran, Iran 4 Environmental Health Engineering Department, health faculty, Shahid Beheshti University of Medical Sciences, Tehran, Iran [E.Mail: Azadehnavazi@yahoo.com]

 Stock index as time series are non-stationary and highly noisy due to the fact that stock marke... more  Stock index as time series are non-stationary and highly noisy due to the fact that stock markets are affected by a variety of factors. It is regarded as one of the most challenging application of time series forecasting. Predicting stock index with the noisy data directly is usually subject to large errors. In this paper we compare forecasting the stock index via Wavelet De-noising-based Neural Network (WDNN) with forecasting stock index via single neural network. The daily Tehran Stock index from April 2006 to June 2013 are used to compare the application of the WDNN in predicting the stock index. Experimental results show that de-noising with wavelet transform outperforms the single neural network.

Statistical Software Components, 2020

Output of the code includes following nonlinearity tests: Ramsey, Keenan, Terasvirta, Lin, and Gr... more Output of the code includes following nonlinearity tests: Ramsey, Keenan, Terasvirta, Lin, and Granger (1993), and Tsay. This Matlab code has been used in S. Mohammadi(ASA Data Sci Journal, 2019). McLeod-Li and Arch test for nonlinearity can be done by MATLAB built-in functions, namely lbqtest and archtest. Please pay attention to that in the nonlinearity test by lbtest you should use squares of residual series of the time series.

Journal of Money and Economy, 2014

This paper presents an optimal portfolio selection approach based on value at risk (VaR), conditi... more This paper presents an optimal portfolio selection approach based on value at risk (VaR), conditional value at risk (CVaR), worst-case value at risk (WVaR) and partitioned value at risk (PVaR) measures as well as calculating these risk measures. Mathematical solution methods for solving these optimization problems are inadequate and very complex for a portfolio with high number of assets. For these reasons, a combination of particle swarm optimization (PSO) and genetic algorithm (GA) is used to determine optimized weights of assets. Stocks' Optimized weight results show that proposed algorithm gives more accurate outcomes in comparison with GA algorithm. According to back-testing analysis, PVaR and WVaR overestimate risk value while VaR and CVaR give a rather accurate estimation. A set of companies in Tehran Stock Exchange are considered as a case study for empirical analysis.

This M-File forecasts univariate time series such as stock prices with a feedforward neural netwo... more This M-File forecasts univariate time series such as stock prices with a feedforward neural networks. It finds best (minimume RMSE) network automatically and uses early stopping method for solving overfitting problem.

SSRN Electronic Journal, 2015

The objective of this research is to apply the recently developed fractional cointegration vector... more The objective of this research is to apply the recently developed fractional cointegration vector autoregressive (FCVAR) model in developing pairs trading model and compare the results to that of the cointegration approach. To the best of our knowledge, there are no papers on fractional cointegration approach to date. Hence, this paper would be the first practical research based on fractional cointegration on futures market.As the focus of this paper in on short-term trading positions, minute intraday interval prices are used to estimate the model and execute the trading strategies. First, we form two portfolios, “FCI” and “CI”. FCI is created using fractional cointegration and CI using cointegration approach. Then, using an estimation window of specific length, fractional cointegration test and cointegration test are preformed to verify the existence of fractional cointegration and cointegration relationship between the selected pair respectively. At last, the long-term equilibrium is estimated using Fractional Vector Error Correction Model (FVECM) for FCI and Vector Error Correction Model (VECM) for CI and based on this estimation the trading strategy is implemented. The results show that the fractional cointegration approach yields statistically significant higher returns than the cointegration approach. The number of trades is also lower for fractional cointegration approach, which further contributes to the performance of the strategy. Finally, on a risk-adjusted level, the fractional cointegration approach has a higher Sortino ratio compared to the cointegration approach. Furthermore, window sizes are optimized and the best and worst window sizes for both the cointegration and fractional cointegration estimation window and trading window is reported.

Statistical Software Components, 2020

This code tests the statistical significance of inputs by using an artificial neural network with... more This code tests the statistical significance of inputs by using an artificial neural network with a flexible structure. The code tries to find the best number of neurons for the neural network for testing the significance of inputs.

Taḥqīqāt-i mālī, 2013

Probability of Private information Based Trade (PIN) has introduced as information risk measure. ... more Probability of Private information Based Trade (PIN) has introduced as information risk measure. This paper is going to estimate probability of private information based trade (PIN) in Tehran Stock Exchange using microstructure models. Our results show that PIN is significantly different from zero for Tehran Stock Exchange.

Statistical Software Components, 2009

Рассматриваются содержание и объективные предпосылки применения контроллинга в управлении затрата... more Рассматриваются содержание и объективные предпосылки применения контроллинга в управлении затратами на качество на предприятии. Определены категории общих затрат на качество на начальной стадии внедрения контроллинга, а также необходимость в разработке формы отчётности о расходах на качество, отвечающих требованиям предприятий различных отраслей экономики.

This function tests harmful multicollinearity based on the difference between p-value of estimate... more This function tests harmful multicollinearity based on the difference between p-value of estimated coefficients of ordinary least squares and generalized ridge regression estimators.

International Journal of Accounting Information Systems, 2017

This paper concentrates on the effectiveness of using a hybrid intelligent system that combines m... more This paper concentrates on the effectiveness of using a hybrid intelligent system that combines multilayer perceptron (MLP) neural network, support vector machine (SVM), and logistic regression (LR) classification models with harmony search (HS) optimization algorithm to detect corporate tax evasion for the Iranian National Tax Administration (INTA). In this research, the role of optimization algorithm is to search and find the optimal classification model parameters and financial variables combination. Our proposed system finds optimal structure of the classification model based on the characteristics of the imported dataset. This system has been tested on the data from the food and textile sectors using an iterative structure of 10-fold cross-validation involving 2451 and 2053 test set samples from the tax returns of a two-year period and 1118 and 906 samples as out-of-sample using the tax returns of the consequent year. The results from out-of-sample data show that MLP neural network in combination with HS optimization algorithm outperforms other combinations with 90.07% and 82.45% accuracy, 85.48% and 84.85% sensitivity, and 90.34% and 82.26% specificity, respectively in the food and textile sectors. In addition, there is also a difference between the selected models and obtained accuracies based on the test data and out-of-sample data in both sectors and selected financial variables of every sector.

Taḥqīqāt-i mālī, 2017

This paper analyses whether joint probability distribution function of losses due to different ex... more This paper analyses whether joint probability distribution function of losses due to different exposures covered under the same policy could be modeled in an appropriate manner via mixture distribution proposed and copula concept. <br />Special type of distribution which is a mixture of Generalized Hyperbolic Skew t distribution and Extreme Value theory (EVT) has been used for modeling marginal distributions of claims and copula function has been considered as a means of modeling dependency structure among claims. Most important copula including; Gaussian, t, Frank, Gumbel and Clayton was tested from goodness of fit point of view. <br />The data used in this study are the amount of property damage and bodily injury covered under automobile liability insurance. <br />Results reveal that joint probability distribution of claims could be effectively modeled by Clayton copula and proposed mixture distribution.

During the past few decades, there have been many evidences to believe that the stock markets aro... more During the past few decades, there have been many evidences to believe that the stock markets around the world follow cyclical trends. In this paper, we study the cyclical trends using wavelet function based on various time windows on some major stock market indices. We use two methods of Daubechies and reverse bi-orthogonal wavelet methods and determine the optimal values of both methods. The results are used for Tehran stock exchange using the most recent ten years daily information as an empirical study. The details of our analysis on TEDPIX index for the last decade indicate that there are, at least, four trends of weekly, monthly, quarterly and yearly and the cycles would be expected to be repeated in future.

SSRN Electronic Journal, 2004

ABSTRACT Testing fixed and random effects is one of peractical problems in panel estimations. Thi... more ABSTRACT Testing fixed and random effects is one of peractical problems in panel estimations. This program tests fixed and random effects for user defined models.

SSRN Electronic Journal, 2005

ABSTRACT This program helps to identification and aytomatic forecasting with ARMA models for fore... more ABSTRACT This program helps to identification and aytomatic forecasting with ARMA models for forecasters and analysts.This program is compatible with EViews 3,3.1,4,4.1. For running it in EViews 5 and 5.1 simply check the box Version 4 compatible variable substitution.

SSRN Electronic Journal, 2004

ABSTRACT This program tests multicolinearity with condition number and variance inflation factor ... more ABSTRACT This program tests multicolinearity with condition number and variance inflation factor (VIF) for given set of explanatory variables. It report high multicolinearity if condition number (cond) be greater than 20(or VIF greater than 10).

Business cycles analysis is one of the most important tasks for economists and statisticians. Var... more Business cycles analysis is one of the most important tasks for economists and statisticians. Various methods such as HP filter, BP filter, CF filter spectral analysis and Fourier transformation are used in time series analysis. All of these methods are sensitive to stationarity of time series. Also the traditional methods don't offer information in scale analysis. In this paper we introduce the Wavelet theory and its applications especially to economics .Also decomposition of seasonal GDP of Iran and analysis of business cycles are the other main propose. Results of GDP decomposition show 7 cycles with length of 16-32 quarters and 13 cycles with 8-16 quarters. Volatility analysis implies no change in variance of Wavelet coefficients in prewar and war periods .However volatility of GDP increased in the post war period. JEL Clarification: C14, C22, C19, E32. One way of solving welfare loss asymmetric information in insurance is to design informationally consistent contracts. When...

Department of Environmental Science, Science and Research Branch, Islamic Azad University, Tehran... more Department of Environmental Science, Science and Research Branch, Islamic Azad University, Tehran, Iran 2 Faculty of Environment, University of Tehran, P.O.Box 14155-6135, Tehran, Iran Faculty of Management, Tehran University School of Management, Tehran, Iran 4 Environmental Health Engineering Department, health faculty, Shahid Beheshti University of Medical Sciences, Tehran, Iran [E.Mail: Azadehnavazi@yahoo.com]

 Stock index as time series are non-stationary and highly noisy due to the fact that stock marke... more  Stock index as time series are non-stationary and highly noisy due to the fact that stock markets are affected by a variety of factors. It is regarded as one of the most challenging application of time series forecasting. Predicting stock index with the noisy data directly is usually subject to large errors. In this paper we compare forecasting the stock index via Wavelet De-noising-based Neural Network (WDNN) with forecasting stock index via single neural network. The daily Tehran Stock index from April 2006 to June 2013 are used to compare the application of the WDNN in predicting the stock index. Experimental results show that de-noising with wavelet transform outperforms the single neural network.

Statistical Software Components, 2020

Output of the code includes following nonlinearity tests: Ramsey, Keenan, Terasvirta, Lin, and Gr... more Output of the code includes following nonlinearity tests: Ramsey, Keenan, Terasvirta, Lin, and Granger (1993), and Tsay. This Matlab code has been used in S. Mohammadi(ASA Data Sci Journal, 2019). McLeod-Li and Arch test for nonlinearity can be done by MATLAB built-in functions, namely lbqtest and archtest. Please pay attention to that in the nonlinearity test by lbtest you should use squares of residual series of the time series.

Journal of Money and Economy, 2014

This paper presents an optimal portfolio selection approach based on value at risk (VaR), conditi... more This paper presents an optimal portfolio selection approach based on value at risk (VaR), conditional value at risk (CVaR), worst-case value at risk (WVaR) and partitioned value at risk (PVaR) measures as well as calculating these risk measures. Mathematical solution methods for solving these optimization problems are inadequate and very complex for a portfolio with high number of assets. For these reasons, a combination of particle swarm optimization (PSO) and genetic algorithm (GA) is used to determine optimized weights of assets. Stocks' Optimized weight results show that proposed algorithm gives more accurate outcomes in comparison with GA algorithm. According to back-testing analysis, PVaR and WVaR overestimate risk value while VaR and CVaR give a rather accurate estimation. A set of companies in Tehran Stock Exchange are considered as a case study for empirical analysis.

This M-File forecasts univariate time series such as stock prices with a feedforward neural netwo... more This M-File forecasts univariate time series such as stock prices with a feedforward neural networks. It finds best (minimume RMSE) network automatically and uses early stopping method for solving overfitting problem.