Search results

The paper introduced the use of a hybrid system of neural networks support vector machines (NNSVMs) consisting of artificial neural networks (ANNs) and support vector machines (SVMs) to price single-family properties.

The mechanism of the hybrid system is such that its output is given by the SVMs which utilise the results of the ANNs as their input. The results are compared to other property pricing modelling techniques including the standalone ANNs, SVMs, geographically weighted regression (GWR), spatial error model (SEM), spatial lag model (SLM) and the ordinary least squares (OLS). The techniques were applied to a dataset of 3,225 properties sold during the period, January 2012 to May 2014 in Cape Town, South Africa.

The results demonstrate that the hybrid system performed better than ANNs, SVMs and the OLS. However, in comparison to the spatial models (GWR, SEM and SLM) the hybrid system performed abysmally under with SEM favoured as the best pricing technique.

The findings extend the debate in the body of knowledge that the results of the OLS can significantly be improved through the use of spatial models that correct bias estimates and vary prices across the different property locations. Additionally, utilising the result of the hybrid system is thus affected by the black-box nature of the ANNs and SVMs limiting its use to purposes of checks on estimates predicted by the regression-based models.

This study uses a simultaneous equations approach to examine the price‐earnings relationship of non‐U.S. firms that directly list their securities in U.S. capital markets or trade as American Depository Receipts (ADRs). The Hausman test shows that price changes and earnings changes are endogenously determined, thus the simultaneous equations approach is used to estimate the earnings response coefficient (ERC) and the returns response coefficient (RRC). Under the ordinary least squares (OLS) estimation, the parameter estimates are biased downward because the OLS fails to correct for endogeneity. In general, our results show that the joint estimation procedure mitigates some of the single‐equation bias. The estimated ERC and the RRC are higher under the three stage least regression (3SLS) than under the OLS regression. In addition, the product of the ERC and the RRC coefficients approaches its theoretical value of one when using the 3SLS estimation. The evidence also shows that institutional factors affect the way the market value information for these firms. We find that the ERC and RRC are insignificant for the common law non‐ADR firms and significantly positive for common law ADR firms.

Citizen trust in police is important in terms of citizen consent to government policies and of police achieving their organizational goals. In the previous study, improvements in police policy, organizational operation and policing activities were developed to clarify which factors influence trust in police and how trust can be improved. This research raises the question, would changes in trust in police have an impact on trust in government? In this paper, this research question is discussed theoretically and the causal relationship analyzed empirically by applying OLS, ordered logistic, 2SLS and logistic regressions.

The basic analysis methods are to apply the OLS and the ordered logistic regression. OLS regression analysis is an analytical method that minimizes an error range of a regression line. The assumptions for OLS are: linearity, independence, equilibrium, extrapolation and multicollinearity issues. These problems were statistically verified and analyzed, in order to confirm the robustness of the analysis results by comparing the results of the ordered logistic regression because of the sequence characteristic of the dependent variable. The data to be used in this study is the Asia Barometer Survey in 2013.

Trust in police and citizen perception of safety are analyzed as important factors to increase trust in the government. The effects of trust in police are more significant than the effects of control variables, and the direction and strength of the results are stable. The effect of trust in police on trust in government is strengthened by the perception of safety (IV). In addition, OLS, ordered logistic regression analysis, which analyzed trust in central government and local government, and logistic regression analysis categorized by trust and distrust show the stability.

This paper has implications in terms of theoretical and empirical analysis of the relationship between trust in police and trust in government. In addition, the impact of perception of safety on trust in police can be provided to police officers, policymakers and governors who are seeking to increase trust in government. This paper is also meaningful in that it is the microscopic research based on the citizens' survey. One of the limitations of macroscopic research is that it does not consider the individual perceptions of citizens.

The results of this paper can confirm the relationship of the virtuous cycle, which is perception of safety – trust in police – trust in government. The police will need to provide security services to improve citizens' perception of safety and make great efforts to create safer communities and society. Trust in police formed through this process can be an important component of trust in government. By making citizens feel safer and achieving trust in police, ultimately, trust in government will be improved.

The police perform one of the essential roles of government and are one of the major components of trust in government, but the police sector has been neglected compared to the roles of the economic and political sectors. These influences of macro factors are too abstract to allow specific policy directions to be suggested. If we consider trust in police, and factors that can improve trust in government, we can suggest practical policy alternatives.

Data envelopment analysis (DEA) is based on the production possibility set that involves the process of converting resources or inputs to outputs. Accordingly, most DEA models include endogenous variables and need an additional step to find the influence of exogenous variables on the process. The purpose of this paper is to examine the relationship between the efficiency scores of DEA and the exogenous variables using truncated regression analysis with double bootstrapping along with two additional methods.

First, the authors employ DEA for benchmarking the comparative efficiency of the health care institutes. Next, the authors run and compare truncated, ordinary least square (OLS) and Tobit regression analysis using the double bootstrapping algorithm for finding the influence of exogenous variables on the efficiency of the health care institutes.

The authors confirmed the amount of bias for the Tobit and OLS regression models, which was caused by serially correlated errors. Accordingly, the authors chose results from the truncated regression model with double bootstrapping for examining the influence of exogenous or environment variables on the efficiency scores.

The study includes cross-sectional data on health care institutes in the state of Washington, USA. Collecting data in various states or regions over time is left for future studies.

In this study, three exogenous variables such as Medicaid revenues, locations of health care institutes and ownership types are significant for explaining the relationship between the efficiency scores and a group of the exogenous variables. Managers and policy makers need to pay attention to these variables along with endogenous variables for promoting the sustainability of the health care institutes.

The study demonstrates the usefulness of the truncated regression analysis with double bootstrapping for confirming the relationship between the efficiency scores of DEA and a group of exogenous variables, which is rare in the DEA literature.

The purpose of this paper is to find out the factors affecting rentals of residential apartments in Pune, India.

Four regression models are developed, i.e. basic ordinary least square (OLS) regression model, OLS regression model with robust estimates, OLS regression model with clustered robust estimates and generalized least square (GLS) regression model with maximum likelihood (ML) robust estimates. Based on the Akaike information criterion and Bayesian information criterion criteria, OLS regression model with clustered robust estimates and GLS regression model with robust estimates are best fit. The data are tested for multicollinearity and the models are tested for heteroscedasticity. The study uses the expected rent value data collected from Web portals and the data on factors affecting the rental value of residential property are collected through the study of land use maps, Google earth software and field visits.

Total floor area and number of rooms are structure related factors that positively affect the rental value, i.e. more the area and number of rooms, higher the rental value. The distances from the nearest police station and fire station are security and safety factors. The results suggest that higher distance from these factors leads to lower rental values, as safety and security is the top priority of residents seeking residential property on rental basis. The distance from employment zones, distance from nearest school/college and the distance from the nearest public transport terminal are convenience related factors that negatively affect the rental value, as greater the distance, lesser the rental value and vice versa. The distance from Central Business District and hospitals has a positive effect on the rental values of a residential property implying that higher distances from these places command higher rental value.

The study relies on rental data that owner is expecting for a particular property, it is not certain that the property would be actually rented for the same value. Second, researchers had to drop certain important drivers of rental value because of the issue of multicollinearity.

This is one of the rare studies conducted in Indian context, and the findings of the study are useful from the owner, tenants, urban bodies and developers’ point of view. Knowing that India is one of the fastest growing markets and need for housing is increasing day by day (including housing facility on rental basis), the stakeholders need to take care of the factors that affect the rental values of a residential property.

The authors suggest the governments and the municipal bodies in India to come up with a public rental housing policy that separately caters to the needs of the lower income group, middle and upper income group in at least metros, tier I and tier II cities that are witnessing unprecedented growth in job seeking immigrants, who are seeking properties on rental basis. While developing a public rental policy, they must keep in mind the factors that are driving the rental values, such as proximity to employment zones, proximity to proper school and college, efficient public transport system as well as all safety and security measures. Creation of such a public rental policy is a win–win situation for immigrants, property owners and government/urban development bodies.

This paper is the first empirical study about the factors affecting rental values in Pune, India. The study will help property owners, immigrant and local tenants, government and urban development bodies to develop an understanding about the important factors affecting rental value and come up with their respective plans. Advanced econometric regression models are used based on the data that is collected through actual field visits, study of maps and secondary information rather than use of survey method or creation of dummy variables.

Exploring the influencing factors on urban rail transit (URT) ridership is vital for travel demand estimation and urban resources planning. Among various existing ridership modeling methods, direct demand model with ordinary least square (OLS) multiple regression as a representative has considerable advantages over the traditional four-step model. Nevertheless, OLS multiple regression neglects spatial instability and spatial heterogeneity from the magnitude of the coefficients across the urban area. This paper aims to focus on modeling and analyzing the factors influencing metro ridership at the station level.

This paper constructs two novel direct demand models based on geographically weighted regression (GWR) for modeling influencing factors on metro ridership from a local perspective. One is GWR with globally implemented LASSO for feature selection, and the other one is geographically weighted LASSO (GWL) model, which is GWR with locally implemented LASSO for feature selection.

The results of real-world case study of Shenzhen Metro show that the two local models presented perform better than the traditional global model (OLS) in terms of estimation error of ridership and goodness-of-fit. Additionally, the GWL model results in a better fit than GWR with global LASSO model, indicating that the locally implemented LASSO is more effective for the accurate estimation of Shenzhen metro ridership than global LASSO does. Moreover, the information provided by both two local models regarding the spatial varied elasticities demonstrates the strong spatial interpretability of models and potentials in transport planning.

The main contributions are threefold: the approach is based on spatial models considering spatial autocorrelation of variables, which outperform the traditional global regression model – OLS – in terms of model fitting and spatial explanatory power. GWR with global feature selection using LASSO and GWL is compared through a real-world case study on Shenzhen Metro, that is, the difference between global feature selection and local feature selection is discussed. Network structures as a type of factors are quantified with the measurements in the field of complex network.

The purpose of this paper is to track the financial performance of manufacturing firms at different levels of their conditional quantiles. It also analyzes the relevance of revenue and cost channels along with key firm-specific parameters that influence firm’s profitability.

The study analyses a sample of 1,000 manufacturing firms over a study period spanning from 2000 to 2016. It uses both quantile regression and panel ordinary linear square (OLS) models to analyze the financial performance of the firms.

The study finds large scale of heterogeneity among the firms under different quantiles of profitability. Export earnings, firm size, asset turnover and volatility of exchange rate are the decisive determinants of financial performance across all quantiles. Financing assets by current debt is negatively impacting return on assets and return on capital employed of firms from lower quantile whereas profitability is positively impacted if they are financed by long term debt. Debt financing of assets does not make any sense for firms with high quantile of profitability. The study also finds that quantile regression approach is a better method than panel OLS models in the presence of highly heterogeneous and non-normal distributions.

This study is limited to the financial performance of manufacturing firms and does not consider service sector which is also equally competitive. However, a sector wise analysis of firm’s profitability could be more meaningful than comparing all the firms in one basket of manufacturing domain.

The research findings have both practical as well as policy implications. Practically, the study helps the firm managers to identify critical success factors that significantly influence firm’s financial performance at different levels of profitability. It also helps the policy makers to align policy focus to stabilize firms at lower level of profitability and also to manage conducive business environment for all firms at different levels of their profitability.

The study provides a deep theoretical underpinning of literatures on firm’s financial performance and empirically investigates it using advanced methodology. The robust estimates of the study ensure to analyze financial performance under revenue and cost channels at diverse level of their profitability.

The purpose of this study is to compare the predictive performance of the hedonic multivariate regression model with the probabilistic neural network (PNN)-based general regression neural network (GRNN) model of housing prices in “Pune-India.”

Data on 211 properties across “Pune city-India” is collected. The price per square feet is considered as a dependent variable whereas distances from important landmarks such as railway station, fort, university, airport, hospital, temple, parks, solid waste site and stadium are considered as independent variables along with a dummy for amenities. The data is analyzed using a hedonic type multivariate regression model and GRNN. The GRNN divides the entire data set into two sets, namely, training set and testing set and establishes a functional relationship between the dependent and target variables based on the probability density function of the training data (Alomair and Garrouch, 2016).

While comparing the performance of the hedonic multivariate regression model and PNN-based GRNN, the study finds that the output variable (i.e. price) has been accurately predicted by the GRNN model. All the 42 observations of the testing set are correctly classified giving an accuracy rate of 100%. According to Cortez (2015), a value close to 100% indicates that the model can correctly classify the test data set. Further, the root mean square error (RMSE) value for the final testing for the GRNN model is 0.089 compared to 0.146 for the hedonic multivariate regression model. A lesser value of RMSE indicates that the model contains smaller errors and is a better fit. Therefore, it is concluded that GRNN is a better model to predict the housing price functions. The distance from the solid waste site has the highest degree of variable senstivity impact on the housing prices (22.59%) followed by distance from university (17.78%) and fort (17.73%).

The study being a “case” is restricted to a particular geographic location hence, the findings of the study cannot be generalized. Further, as the objective of the study is restricted to just to compare the predictive performance of two models, it is felt appropriate to restrict the scope of work by focusing only on “location specific hedonic factors,” as determinants of housing prices.

The study opens up a new dimension for scholars working in the field of housing prices/valuation. Authors do not rule out the use of traditional statistical techniques such as ordinary least square regression but strongly recommend that it is high time scholars use advanced statistical methods to develop the domain. The application of GRNN, artificial intelligence or other techniques such as auto regressive integrated moving average and vector auto regression modeling helps analyze the data in a much more sophisticated manner and help come up with more robust and conclusive evidence.

To the best of the author’s knowledge, it is the first case study that compares the predictive performance of the hedonic multivariate regression model with the PNN-based GRNN model for housing prices in India.

The purpose of this paper is to examine the effect of bank capital on liquidity creation. Especially, the authors test two competing hypotheses: the “risk absorption” hypothesis and the “financial fragility-crowding out” hypothesis that describe such association in the context of UK and French banking industry.

The authors use data collected from Bankscope for commercial banks pertaining to the aforementioned countries. The sample period ranges from 2000 to 2014. Liquidity creation was measured using a novel approach proposed by Berger and Bouwman (2007). This study uses the quantile regression (QR) and the instrumental variables QR, along with classical ordinary least squares (OLS) and panel regression, to deal with the mixed results reported by previous papers.

Using OLS and panel regression, the authors first find that bank capital negatively affects liquidity creation which supports risk absorption hypothesis. Second, the result from QR confirms the negative association between the aforementioned variables and shows that the effect is homogenous across quantiles of liquidity creation distribution. The result remains unchanged when using the QR with instrumental variables to address the potential problem of endogeneity.

This paper sheds more lights on the relationship between bank capital and liquidity creation by using a novel estimation approach based on the QR methodology.

The management of liquidity has always been seen as a critical but often ignored issue in finance. Despite the abundance of studies on liquidity management, these studies mainly focus on developed countries and on large firms. Liquidity is critical for the small firm but studies on liquidity management in small and medium enterprises (SMEs) are lacking. The purpose of this paper is to examine the firm-level determinants of liquidity of SMEs in Malaysia.

Data are collected for a total of 986 small firms in Malaysia from 2011 to 2014, resulting in a total of 2,683 observations. Firm-specific variables and the effect of the economy are considered as the possible determinants of liquidity. Ordinary least squares (OLS) regression analysis with standard errors adjusted for firm-level clustering and quantile regression analysis are used for this purpose.

Analysis using OLS regression technique indicates that a firm’s profitability, its growth, asset tangibility, size, age and firm status are significant factors in influencing its liquidity decision. Leverage and economic condition are not found to have any significant influence on liquidity. However, quantile regression analysis provides a different picture especially for SMEs with liquidity at the quantile levels of θ=0.10 and 0.90. At θ=0.10, only profitability, tangibility and firm status are significant, while at θ=0.90, tangibility, size, firm status and, to some extent, age are significant in influencing liquidity levels.

To the author’s knowledge, this is the first study analyzing the liquidity decision of SMEs in an emerging market such as Malaysia. Most studies on liquidity management of SMEs are focused on developed countries due to data availability but these studies are also only a handful. Additionally, this study uses quantile regression analysis which highlights the need to analyze financial decisions at different levels rather than at the aggregate level as done in OLS regression analysis.