Mixture of MLP-experts for trend forecasting of time series: A case study of the Tehran stock exchange

A new method for forecasting the trend of time series, based on mixture of MLP experts, is presented. In this paper, three neural network combining methods and an Adaptive Network-Based Fuzzy Inference System (ANFIS) are applied to trend forecasting in the Tehran stock exchange. There are two experiments in this study. In experiment I, the time series data are the Kharg petrochemical company’s daily closing prices on the Tehran stock exchange. In this case study, which considers different schemes for forecasting the trend of the time series, the recognition rates are 75.97%, 77.13% and 81.64% for stacked generalization, modified stacked generalization and ANFIS, respectively. Using the mixture of MLP experts (ME) scheme, the recognition rate is strongly increased to 86.35%. A gain and loss analysis is also used, showing the relative forecasting success of the ME method with and without rejection criteria, compared to a simple buy and hold approach. In experiment II, the time series data are the daily closing prices of 37 companies on the Tehran stock exchange. This experiment is conducted to verify the results of experiment I and to show the efficiency of the ME method compared to stacked generalization, modified stacked generalization and ANFIS.     

Bayesian inference in cointegrated systems

In this paper I deal with Bayesian methods for conducting inference on important features of (potentially) cointegrated VAR models involving I(1) variables. Firstly, (informal) inference is made on the cointegrating rank of the system. Secondly, posterior analysis is used to verify the validity of over-identifying restrictions on the cointegration parameters. Thirdly, posterior distributions are obtained for impulse response functions and predictive densities at different horizons. The relevant posterior distributions are obtained by means of Monte Carlo integration. The analysis is based on the use of simple weakly informative priors. Two applications on simulated data and on the Danish money demand data are presented.     

Asymptotic consequences of neglected dynamics in individual effects models*

We assess the asymptotic consequences of estimating static models based on cross-section or panel data, when in reality the data are generated by a dynamic relationship, involving lagged dependent and current and lagged exogenous variables as well as individual effects. If the exogenous variable follows a stationary process, then the static estimators usually underestimate its long-run effect. This inconsistency is less severe, the higher the autocorrelation of the exogenous variable. If the exogenous variable follows a random walk with or without individual-specific drift, then the estimators are found to be consistent for the long-run effect.     

Bayesian inference in time series models using kernel quasi likelihoods

The paper takes up Bayesian inference in time series models when essentially nothing is known about the distribution of the dependent variable given past realizations or other covariates. It proposes the use of kernel quasi likelihoods upon which formal inference can be based. Gibbs sampling with data augmentation is used to perform the computations related to numerical Bayesian analysis of the model. The method is illustrated with artificial and real data sets.     

Inference Methods for Discretely Observed Continuous-Time Stochastic Volatility Models: A Commented Overview

In this paper an overview of inference methods for continuous-time stochastic volatility models observed at discrete times is presented. It includes estimation methods for both parametric and nonparametric models that are completely or partially observed in a variety of situations where the data might be nonlinear functions of the components of the model and/or contaminated with observation noise. In each case, the main reported methods are presented, making emphasis on underlying ideas, theoretical properties of the estimators (bias, consistency, efficient, etc.), and the viability of their implementation to solve actual problems in finance.     

The endo–exo problem in high frequency financial price fluctuations and rejecting criticality

The endo–exo problem lies at the heart of statistical identification in many fields of science, and is often plagued by spurious strong-and-long memory due to improper treatment of trends, shocks and shifts in the data. A class of models that has shown to be useful in discerning exogenous and endogenous activity is the Hawkes process. This class of point processes has enjoyed great recent popularity and rapid development within the quantitative finance literature, with particular focus on the study of market microstructure and high frequency price fluctuations. We show that there are important lessons from older fields like time series and econometrics that should also be applied in financial point process modelling. In particular, we emphasize the importance of appropriately treating trends and shocks for the identification of the strength and length of memory in the system. We exploit the powerful Expectation Maximization algorithm and objective statistical criteria (BIC) to select the flexibility of the deterministic background intensity. With these methods, we strongly reject the hypothesis that the considered financial markets are critical at univariate and bivariate microstructural levels.     

CAUSAL INFERENCE ON EDUCATION POLICIES: A SURVEY OF EMPIRICAL STUDIES USING PISA,TIMSS AND PIRLS

The identification of the causal effects of educational policies is the top priority in recent education economics literature. As a result, a shift can be observed in the strategies of empirical studies. They have moved from the use of standard multivariate statistical methods, which identify correlations or associations between variables only, to more complex econometric strategies, which can help to identify causal relationships. However, exogenous variations in databases have to be identified in order to apply causal inference techniques. This is a far from straightforward task. For this reason, this paper provides an extensive and comprehensive overview of the literature using quasi‐experimental techniques applied to three well‐known international large‐scale comparative assessments, such as PISA, PIRLS or TIMSS, over the period 2004–2016. In particular, we review empirical studies employing instrumental variables, regression discontinuity designs, difference in differences and propensity score matching to the above databases. Additionally, we provide a detailed summary of estimation strategies, issues treated and profitability in terms of the quality of publications to encourage further potential evaluations. The paper concludes with some operational recommendations for prospective researchers in the field.     

Heteroskedasticity-consistent covariance matrix estimators for spatial autoregressive models

In the presence of heteroskedasticity, conventional test statistics based on the ordinary least squares (OLS) estimator lead to incorrect inference results for the linear regression model. Given that heteroskedasticity is common in cross-sectional data, the test statistics based on various forms of heteroskedasticity-consistent covariance matrices (HCCMs) have been developed in the literature. In contrast to the standard linear regression model, heteroskedasticity is a more serious problem for spatial econometric models, generally causing inconsistent extremum estimators of model coefficients. This paper investigates the finite sample properties of the heteroskedasticity-robust generalized method of moments estimator (RGMME) for a spatial econometric model with an unknown form of heteroskedasticity. In particular, it develops various HCCM-type corrections to improve the finite sample properties of the RGMME and the conventional Wald test. The Monte Carlo results indicate that the HCCM-type corrections can produce more accurate results for inference on model parameters and the impact effects estimates in small samples.     

Opinion leadership vs. para-social relationship: Key factors in influencer marketing

The literature on influencer marketing has identified opinion leadership of influencers and parasocial relationship with influencers as two focal constructs affecting followers' purchase intention; yet they are only studied in isolation. This research aims to provide a comprehensive understanding of the effects of the two key constructs by combining and comparing them in a model with the moderators of post characteristics, namely post type and correspondent inference. Empirical results from 409 online followers of two Instagram accounts confirm the complementary effects of opinion leadership and parasocial relationship in influencer marketing and reveal the more prominent role of parasocial relationship over opinion leadership in affecting followers’ purchase intention. These effects are moderated by post characteristics. Storytelling posts intensify both effects. Correspondent inference enhances the effect of parasocial relationship, but not that of opinion leadership. These findings point to the importance of the social aspect of influencer marketing and inform the influencer marketing research and practice on “who says what” for improving communication effectiveness.     

TOOLS FOR RE‐RANDOMIZATION

With the rising popularity of field experiments in economics, re‐randomization schemes have emerged as tools to induce balance in observable variables across treatment groups. However, re‐randomization is not fully understood and the methodologies to estimate its effects on the distribution of parameters are still under‐developed. This paper helps to close that gap by suggesting an asymptotically normal re‐randomization scheme and bootstrapping procedure to carry inference under a wide range of estimators and re‐randomization schemes.