ESTIMATING QUANTILE REGRESSION FOR DYNAMIC PANEL DATA USING TWO-STAGE APPROACH

2018 ◽  
Vol 103 (2) ◽  
pp. 451-465
Author(s):  
A. R. Elhoussainy ◽  
A. E. Ahmed ◽  
M. A. Hossameldin
Keyword(s):  
Panel Data ◽  
Quantile Regression ◽  
Dynamic Panel Data ◽  
Two Stage ◽  
Dynamic Panel

scholarly journals Dynamic Quantile Panel Data Analysis of Stock Returns Predictability

2016 ◽  
Vol 8 (2) ◽  
pp. 115 ◽  
Author(s):  
Bülent Guloglu ◽  
Sinem Guler Kangalli Uyar ◽  
Umut Uyar
Keyword(s):  
Panel Data ◽  
Quantile Regression ◽  
Stock Returns ◽  
Conditional Distribution ◽  
Extreme Values ◽  
Dynamic Panel Data ◽  
Regression Technique ◽  
Heterogeneous Structure ◽  
Dynamic Panel ◽  
Different Parts

<p>This paper analyses the effect of financial ratios on stock returns using quantile regression for dynamic panel data with fixed effects. Eighty three firms of manufacturing industry, which were traded on the Borsa Istanbul for 2000-2014 period, are covered in the study. The most of financial variables have heterogeneous structure so they generally include extreme values. Thus, panel quantile regression technique, suggested by Koenker (2004), is used. Since the technique yields robust estimator in the case of extreme values the Gaussian estimators will be biased and not efficient. The sensitivity of relationship, on the other hand, can be studied for different parts of the stock returns’ conditional distribution by using quantile regression technique. However, because of that the lagged of dependent variable is used as an explanatory variable in dynamic panel models, fixed effect estimators will be biased. Thereby, in this study the instrumental variable approach suggested by Chernozhukov and Hansen (2006) is used to produce unbiased and consistent estimators.</p>The results show that the stock returns respond to the changes on the financial leverage ratio, the dividend yield, the market-to-book value ratio, financial beta and the total active profitability variables differently for the different parts of the stock returns’ conditional distribution. They also indicate that, at high quantiles, return fluctuations in the current period will be more effective for investors’ transaction attitudes on stocks for the next period.

Penalized quantile regression for dynamic panel data

2010 ◽  
Vol 140 (11) ◽  
pp. 3476-3497 ◽  
Author(s):  
Antonio F. Galvao ◽  
Gabriel V. Montes-Rojas
Keyword(s):  
Panel Data ◽  
Quantile Regression ◽  
Dynamic Panel Data ◽  
Dynamic Panel
Sign in / Sign up

Export Citation Format

Share Document