Range Of Quantiles Research Articles

Specification Choices in Quantile Regression for Empirical Macroeconomics

ABSTRACTQuantile regression has become widely used in empirical macroeconomics, in particular for estimating and forecasting tail risks. This paper examines various choices in the specification of quantile regressions for macro applications, including how and to what extent to include shrinkage and whether to apply shrinkage in a classical or Bayesian framework. We focus on forecasting accuracy, measured with quantile scores and quantile‐weighted continuous ranked probability scores at a range of quantiles from the left to right tail. Across applications, we find that shrinkage is generally helpful to quantile forecast accuracy, with Bayesian quantile regression dominating frequentist quantile regression.JEL Classification: C53, E17, E37, F47

Uncertainty impacts on China’s agricultural commodity futures: a quantile perspective

ABSTRACT This study employs rolling quantile regression and non-parametric estimation methods, specifically Quantile-on-Quantile (QQ) and Causality-in-Quantiles (QC), to investigate the impact of uncertainty indices (Geopolitical Uncertainty (GPR), Climate Policy Uncertainty (CPU), and Economic Policy Uncertainty (EPU)) on China’s agricultural commodity futures. The rolling window quantile regression coefficients vary over time, with CPU having the most significant impact on agricultural commodity futures. The empirical results from the non-parametric estimation QQ method reveal when agricultural commodities are in a prosperous market, CPU and EPU exhibit relatively strong correlation, while GPR shows weaker or insignificant correlation. According to the non-parametric estimation QC method, the evidence suggests that the second-order (variance) causality relationship is relatively stronger than the first-order (mean) causality relationship, and the first-order mean causal relationship between CPU and corn and between EPU and wheat is not significant across the entire range of quantiles. The results on Hedging Effectiveness reveal that the efficiency of uncertainty hedging is substantiated. The hedging effectiveness of EPU exhibits improvement post-COVID-19, whereas the effectiveness of GPR and CPU is more pronounced in the pre-COVID-19 period.

Re‐estimating the pollution haven–halo hypotheses for Brazil via a machine learning procedure

AbstractIn this study, we re‐examined the pollution haven and halo hypotheses in Brazil for approximately five decades (1970–2019) while controlling for the effects of income, renewable energy and natural resource depletion. For clearer insights, the study employed both the conventional autoregressive distributed lag (ARDL) and the enhanced kernel regularized least squares (KRLS) techniques. Notably, the KRLS is a flexible machine learning nonlinear analytical technique that explains the interactions of the regressand and the regressors both at the average and across a range of quantiles. After ascertaining cointegration through the bounds tests and the Bayer–Hanck procedures, the following empirical outcomes emerged: The ARDL result suggests the acceptance of the pollution haven hypothesis in Brazil in both the short and long runs. However, the KRLS technique reveals that foreign direct investment (FDI) could enhance environmental quality (pollution halo) within the 25th quantile of the distributions of CO2 emissions. However, at the 50th and 70th quantiles, the pollution haven hypothesis is rectified. This suggests the adoption of varying policy options to ensure continuous inflows of FDI without compromising environmental quality. Additionally, among the control variables, a U‐shaped environmental Kuznets curve (EKC) structure is revealed from the influence of gross domestic product (GDP); renewable energy ensures a clean environment at all times, while resource rent ensures a clean environment only at the 25th and 50th quantiles of the distributions. Policies that could lead to clean environments in Brazil have been provided.

Расчет климатических характеристик осадков над Черным морем по данным региональных климатических моделей

The ability of 16 regional climate models (RCMs) from the CORDEX project to adequately simulate the average values and long-term trends in precipitation over the Black Sea is assessed. The results of the simulations performed using the RCMs are compared with the data of the ERA5 reanalysis for the control period (1959–2005). Average annual precipitation and average seasonal precipitation for winter and summer, as well as long-term trends in precipitation for all months in a wide range of quantiles, are analyzed. It was found that average seasonal precipitation over the Black Sea is satisfactorily simulated by most RCMs in winter and by six RCMs in summer. At the same time, the monthly precipitation trends estimated on the basis of forecast calculations are close to those based on the ERA5 data only for one RCM. Keywords: precipitation, quantile regression, Black Sea, CORDEX

Расчет климатических характеристик осадков над Черным морем по данным региональных климатических моделей

The ability of 16 regional climate models (RCMs) from the CORDEX project to adequately simulate the average values and long-term trends in precipitation over the Black Sea is assessed. The results of the simulations performed using the RCMs are compared with the data of the ERA5 reanalysis for the control period (1959–2005). Average annual precipitation and average seasonal precipitation for winter and summer, as well as long-term trends in precipitation for all months in a wide range of quantiles, are analyzed. It was found that average seasonal precipitation over the Black Sea is satisfactorily simulated by most RCMs in winter and by six RCMs in summer. At the same time, the monthly precipitation trends estimated on the basis of forecast calculations are close to those based on the ERA5 data only for one RCM. Keywords: precipitation, quantile regression, Black Sea, CORDEX

Does globalization escalate the carbon emissions? Empirical evidence from selected next-11 countries

Adverse consequences are observed in developing countries due to the impact of the globalization process. Therefore, our study aims to empirically verify whether globalization escalates carbon dioxide emissions in selected N-11 (next-11) countries between 1990 and 2019. The study also analyzes how per capita GDP, per capita GDP2, population growth, and renewable energy consumption affect carbon emissions. For this reason, the researchers used several econometric methods, including the slope homogeneity test, the cross-sectional dependency test, the panel unit root test, the panel cointegration test, the method of moment’s panel quantile regression analysis, and the Wald test. The estimated results of panel quantile regression show how carbon emissions change across a range of quantiles (0.1 to 0.9). The findings show that per capita GDP significantly impacts the overabundance of carbon emissions in N-11 countries. Over time, the study found that the positive coefficient value of per capita GDP decreased from the first to the last (7.41 to 5.87), leading to the validation of the EKC hypothesis. The adverse correlation between per capita GDP2 and environmental contamination confirms that the Environmental Kuznets Curve hypothesis is valid for selected N-11 countries. Globalization deteriorates the environment by directly affecting CO2 emissions. It increases monotonically from the lower quantile to the upper quantile (0.972 to 1.002). At the quantile level of 0.1 to 0.9, population growth and renewable energy consumption increase impede carbon dioxide emissions in these selected countries. Coefficient values in the quantile 0.1 to 0.9 (-0.35 to -0.53) suggest that governments can reduce carbon emissions more due to renewable energy consumption over time. But the negative coefficient values of the population (-0.97, -0.93, -0.90, -0.88, -0.86, -0.85, -0.83, -0.81, and -0.77) decrease from the lower quantile to the upper quantile. The Wald test supports the asymmetric effects of different quantiles. As a robustness check of estimators, the study used FMOLS, DOLS, and CCR, which show the variables’ long-run elasticity. The research developed targeted policy recommendations for sustainably mitigating carbon emissions based on the above results.

Is art market efficient? Evidence from non-linear quantile unit-root tests

ABSTRACT In this paper, we use quarterly stock price index data to test the weak form of the efficient market hypothesis for 13 art market indices. The fact that hypotheses can be efficiently tested by performing unit root tests is well known. The conventional unit root test results show that most art indices have unit roots, which means that these art markets are efficient. However, the results of the non-linear quantile unit root tests we conduct over a range of quantiles lead us to reject the null of the unit root test for several art markets. To acquire more information, we perform the test over sub-quantile intervals, finding asymmetric market efficiency between the low and high quantile intervals: we cannot reject the unit root null in upper quantile intervals for most art price indices.

Data stream fusion for accurate quantile tracking and analysis

UDDSketch is a recent algorithm for accurate tracking of quantiles in data streams, derived from the DDSketch algorithm. UDDSketch provides accuracy guarantees covering the full range of quantiles independently of the input distribution and greatly improves the accuracy with regard to DDSketch. In this paper we show how to compress and fuse two or more data streams (or datasets) by leveraging the mergeability of the UDDSketch data summaries. In general, two summaries on two data streams are said to be mergeable if there exists an algorithm that allows combining the two summaries into a single one related to the union of the two datasets, simultaneously preserving the error and size guarantees. The property of mergeability of a sketch enables the parallel and distributed processing of big volume data streams that can be compressed and fused by means of such mergeable data structures. Among the applications strictly related to accurate tracking of quantiles, requiring parallel and/or distributed processing we recall here estimating the latency of a web site, database query optimizers and the need of succinctly summarizing the distribution of values occurring over a sensor network. We prove that UDDSketch is fully mergeable and introduce PUDDSketch, a parallel version of UDDSketch suitable for message-passing based architectures. We formally prove its correctness and compare it to a parallel version of DDSketch, showing through extensive experimental results that our parallel algorithm almost always outperforms the parallel DDSketch algorithm with regard to the overall accuracy in determining the quantiles. Moreover, we also design and implement parallel versions of both the state of the art KLL and REQ sequential algorithms in order to compare and contrast PUDDSketch versus the corresponding parallel algorithms. Our experiments clearly show that PUDDSketch is faster or on par with regard to parallel running time, whilst providing simultaneously greater accuracy.

Hydrological response in a highly urbanized watershed in China

Abstract Urbanization has a strong signal on the hydrologic cycle, leading to reduced infiltration and faster and larger runoff. However, the detailed analysis of the contribution of urbanization to different quantiles of discharge is still lacking; particularly, less is known in watersheds that have been experiencing such large and rapid urbanization as those in China. Here, we focus on the Wenyu watershed, a rapidly urbanizing basin located in the Beijing metropolitan area. Using a statistical attribution framework, we examine the hydrological response to the increasing urbanization across a wide range of discharge quantiles, from low to high flows; moreover, we perform analyses at the seasonal scale to capture differences in the physical processes at play during the year. Results indicate that the selected GAMLSS (generalized additive models for location, scale, and shape) models can capture well the variability in streamflow in this highly urbanized basin, with the average Spearman correlation coefficients between observations and the median of the fitted models of 0.84, 0.79, and 0.81 in spring, summer, and winter, respectively. Overall, urbanization played a different role in the different seasons and discharge quantiles. More specifically, we find its strongest impact to be in winter and spring, and for low and median quantiles. The role of precipitation is the strongest in summer, and it increases as we move toward the upper tail of the discharge distribution, especially above the 55th percentile, for which precipitation is selected as the only important predictor. Recycled water, on the other hand, tends to play a more dominant role in winter and spring. HIGHLIGHTS Urbanization has a strong signal on the hydrologic cycle; however, less is known in watersheds that have been experiencing such large and rapid urbanization. In this research, we focus on a fast urbanizing basin, and the hydrological responses to the increasing urbanization have been examined. The quantification of the impacts of drivers for different discharge quantiles and seasons has been examined. Urbanization plays a decreasing role towards the upper tail of the distribution.

An analysis of OPEC oil production reaction to non-OPEC oil supply

This paper examines the impact of non-OPEC oil supply on OPEC oil production level at different quantiles of the OPEC production empirical distribution. It employs the Quantile Autoregressive Distributed Lags (QARDL) model that allows to probe simultaneously short-term connections and long-run cointegrating relationships across a range of quantiles. The analysis is undertaken using monthly data from January 1993 to March 2020.The main findings show that the influence of non-OPEC production on OPEC production is symmetric in the long-run but quantile-dependent in the short-run. In the short-run OPEC production decreases significantly following an increase of non-OPEC production. However, in the long run, the increase in non-OPEC production causes OPEC production to rise. Furthermore, the results show that oil prices increase OPEC production in the long- and short-run. The potential policy implications for OPEC and non-OPEC oil production are discussed.

Understanding the dynamics of the resource curse and financial development in China? A novel evidence based on QARDL model

This paper aims to examine the quantile behavior of the relationship between natural resource rent and financial development. Furthermore, gross domestic product, gross fixed capital formation, and trade openness are also under observation as explanatory variables for China's economy during 1981–2018. This research employs a recently developed quantile autoregressive distributed lag (QARDL) model of Cho et al.'s (2015) research. The present study contribution can be examined by applying QARDL as it depicts the causality patterns through a range of quantiles of natural resource rent and financial development along with other explanatory variables. The study findings confirm that the variable's relationship is quantile-dependent, highlighting the shortcomings of findings in earlier studies and providing meaningful contributions. Furthermore, the findings under Wald-test reject the parameter constancy for the Chinese economy. Additionally, our findings specified a resource curse hypothesis that shows that financial development is negatively and significantly affected by China's natural resource rent level. However, gross domestic product, gross capital formation, and trade openness are observed as positive financial development determinants during a similar study period. The study findings recommend that the Chinese government and policymakers lower the negative linkage between resource rent and financial development, specifically in the long-run.

Hypothesis testing of varying coefficients for regional quantiles

Testing the behavior of varying coefficients (VC) over a range of quantiles is important in the field of regression analysis. This study tests whether coefficient functions in varying quantile regression share common structural information across a certain range of quantile levels, even when linear combinations of covariates are unspecified in the null hypothesis. Our approach allows varying the coefficients, β(τ,t), as a function of the quantile level, τ∈Δ, and a random variable, t∈T, where Δ is the quantile region of interest and T is a domain of t. By incorporating an interval of quantiles in the inference, the proposed method can test whether the model can be justified by using a VC quantile model against other important reduced models, such as the linear quantile model, varying coefficient linear model, or linear model with homogeneous errors. We use bivariate B-splines to approximate the varying quantile functions, β(τ,t), and utilize composite quantile regression to estimate the parameters. Furthermore, we develop constrained composite quantile regression to provide a more efficient estimate in case the null hypothesis is not rejected. We show that the proposed test admits normal approximations. Using simulation and real data analysis, we demonstrate the superiority of the proposed test over other tests designed for a finite number of quantile levels.

Distributional changes in oil price and inflation in the net oil exporting countries in Africa

The debate on oil price changes and inflation has attracted considerable empirical investigations among policy makers over the years. Most of the empirical literature so far studies disregard the likelihood that oil price changes may influence inflation along various consumer price index distribution. The paper employs a quantile regression model piloted by Koenker and Xiao (2004) to examine the distributional changes in oil price and their impact on inflation of the net oil exporting countries in Africa. By analyzing the behaviour of inflation in a wide range of quantiles, the technique allows us to capture possible variation and inflation distribution towards to its long-run equilibrium. Our results reveal that negative and positive changes in oil price impacts inflation negatively for majority of the quantiles. The paper also reveals significant disparity in magnitude and sign of the parameters. The result of the quantile regression shows an insignificant negative relationship between positive oil price and inflation across the quantiles at 5% level of significance. However, the opposite is the case when we modeled the negative oil price. It turns out to be a significantly negative coefficient at the lower quantile of 0.50 to higher levels of 0.90. The paper suggests that in order to stabilize the price, stakeholders must continue to monitor impending inflation volatility and take counteractive actions to sustain inflation potentials. The finding also implies that monetary policy makers have the ability to suppress the effects of hostile oil price changes on inflation

Stochastic frontiers or regression quantiles for estimating the self-thinning surface in higher dimensions?

Stochastic frontier analysis and quantile regression are the two econometric approaches that have been commonly adopted in the determination of the self-thinning boundary line or surface in two and higher dimensions since their introduction to the field some 20 years ago. However, the rational for using one method over the other has, in most cases, not been clearly explained perhaps due to a lack of adequate appreciation of differences between the two approaches for delineating the self-thinning surface. Without an adequate understanding of such differences, the most informative analysis may become a missed opportunity, leading to an inefficient use of data, weak statistical inferences and a failure to gain greater insight into the dynamics of plant populations and forest stands that would otherwise be obtained. Using data from 170 plot measurements in even-aged Larix olgensis (A. Henry) plantations across a wide range of site qualities and with different abundances of woody weeds, i.e. naturally regenerated non-crop species, in northeast China, this study compared the two methods in determining the self-thinning surface across eight sample sizes from 30 to 170 with an even interval of 20 observations and also over a range of quantiles through repeated random sampling and estimation. Across all sample sizes and over the quantile range of 0.90 ≤ τ ≤ 0.99, the normal-half normal stochastic frontier estimation proved to be superior to quantile regression in statistical efficiency. Its parameter estimates had lower degrees of variability and correspondingly narrower confidence intervals. This greater efficiency would naturally be conducive to making statistical inferences. The estimated self-thinning surface using all 170 observations enveloped about 96.5% of the data points, a degree of envelopment equivalent to a regression quantile estimation with a τ of 0.965. The stochastic frontier estimation was also more objective because it did not involve the subjective selection of a particular value of τ for the favoured self-thinning surface from several mutually intersecting surfaces as in quantile regression. However, quantile regression could still provide a valuable complement to stochastic frontier analysis in the estimation of the self-thinning surface as it allows the examination of the impact of variables other than stand density on different quantiles of stand biomass.

U.S. stock prices and macroeconomic fundamentals: Fresh evidence using the quantile ARDL approach

AbstractThis paper explores the long‐run relationship and the associated short‐run dynamics between the U.S. stock market and three major macroeconomic fundamentals, namely the U.S. industrial production index, the U.S. 10‐year Treasury bond yield and the West Texas Intermediate oil price, for the time period covering 1985–2015. The quantile autoregressive distributed lag (QARDL) model presented by Cho et al. (2015) Journal of Econometrics, 188, 281–300, which combines the autoregressive distributed lag model of Pesaran and Shin (1998), Cambridge University Press, and Pesaran et al. (2001) Journal of Applied Econometrics, 16, 289–326, and the quantile regression methodology of Koenker and Bassett (1978), Econometrica, 46, 33–50, is used. This approach analyses the links among the variables over a range of quantiles which represent different states of the equity market. The findings show a significant quantile‐varying cointegration relationship between U.S. equity prices and the three selected macroeconomic factors. The U.S. industrial production index is found to have the largest long‐run influence on the equity market. These findings highlight the importance of market players as well as policymakers to develop a greater understanding and awareness of how macro variables impact stock prices under different stock market conditions, particularly in relation to optimal asset allocation and risk management strategies, among others.

Modelling the Relationship between Match Outcome and Match Performances during the 2019 FIBA Basketball World Cup: A Quantile Regression Analysis.

The FIBA Basketball World Cup is one of the most prominent sporting competitions for men’s basketball, with coaches interested in key performance indicators (KPIs) that give a better understanding of basketball competitions. The aims of the study were to (1) examine the relationship between match KPIs and outcome in elite men’s basketball; and (2) identify the most suitable analysis (multiple linear regression (MLR) vs. quantile regression (QR)) to model this relationship during the men’s basketball tournament. A total of 184 performance records from 92 games were selected and analyzed via MLR and QR, using 10th, 25th, 50th, 75th and 90th quantiles. Several offensive (Paint Score, Mid-Range Score, Three-Point Score, Offensive Rebounds and Turnovers) and defensive (Defensive Rebounds, Steals and Personal Fouls) KPIs were associated with match outcome. The QR model identified additional KPIs that influenced match outcome than the MLR model, with these being Mid-Range Score at the 10th quantile and Offensive Rebounds at the 90th quantile. In terms of contextual variables, the quality of opponent had no impact on match outcome across the entire range of quantiles. Our results highlight QR modelling as a potentially superior tool for performance analysts and coaches to design and monitor technical–tactical plans during match-play. Our study has identified the KPIs contributing to match success at the 2019 FIBA Basketball World Cup with QR modelling assisting with a more detailed performance analysis, to support coaches with the optimization of training and match-play styles.

The inequality of health-income effect in employed workers in China: a longitudinal study from China Family Panel Studies

BackgroundThe relationship between health and income is an essential part of human capital research. The majority of current analyses using classical regression models show that health has a significant impact on income after controlling for the endogeneity of health due to the measurement error and reverse causality. Currently, the Chinese government implements various policies including health related policies to fiercely fight for the domestic poverty issues, and thus only estimating the average effect of health on income could underestimate the impact for low income population and will make policy makers neglect or not pay enough attention to the significant role of health in poverty alleviation. To study the effect of health on income for workers at different income quantiles, we apply the quantile regression method to a panel data from a Chinese household survey. Furthermore, we test the heterogeneity of this health-income effect for different subgroups of workers characterized by sex, registered residence, and residential area. Lastly, we provide an explanation on the possible mechanism of the health-income effect.MethodsThis study uses data from four waves of the China Family Panel Studies (CPFS)- a biennial longitudinal study spanning from 2012 to 2018. The final data used in the regression analysis includes a balanced sample of 19,540 person-year observations aged between 18 to 70 years, with complete information of demographic and social economic status characteristics, job information, and health status of individuals. We use lagged self-reported health to control the potential endogeneity problem caused by reverse causality between health and income. Our identification on heterogenous treatment effects relies on panel quantile regressions, which generate more information than the commonly used mean regression method, and hopefully could reveal the effects of health on income for workers with income distributed at a wide range of quantiles. In addition, we compare the results derived from panel quantile regressions and mean regressions. Finally, we added interaction terms between health and other independent variables to recover the influence channel of health on income.ResultsThe regression estimates show that the effects of health on income are more pronounced for workers distributed on the lower ends of income spectrum, and the health-income effect decreases monotonically with the increase of income. The treatment effect is robust to alternative measures of health and seems to be more pronounced for females than males, for rural workers than their urban counterparts. Finally, we find that health not only directly affects worker’s income but also has different effects on income for different occupation cohorts.ConclusionsThis study provides a different perspective on the impact of individual health status on income, uncovering the heterogeneous effects of health deterioration on income reduction for workers with different incomes by using panel data and rather advanced statistical techniques- panel quantile regressions. At present, the Chinese government is making every effort to solve the problem of poverty and our findings suggest public policies on health and income protections should emphasize different needs of workers with different incomes and special attention should be paid to low-income workers who are much more financially fragile to health deterioration than other income groups.

Connectedness of financial institutions in Europe: A network approach across quantiles

We create cross-quantilogram networks capturing the dependence between stock market returns of 205 financial institutions in Europe and across a large range of quantiles. Based on network density, centrality and centralization, we find that financial institutions in Europe are strongly interconnected across all quantiles, particularly during tranquil periods. These results suggest that financial institutions in Europe are susceptible to spillover, contagion and increased systemic risk. We also find an asymmetric effect: connectedness during crisis periods is higher than that during bullish periods. The most interconnected financial institutions are found to be insurance companies in Switzerland. Relying on an exponential random graph modeling (ERGM) approach, we find that across all quantiles, the connectedness is larger between (i) financial institutions in the same country, (ii) insurance companies, and (iii) financial institutions that operate in countries that adopted the common currency.

UDDSketch: Accurate Tracking of Quantiles in Data Streams

We present UDDSketch (Uniform DDSketch), a novel sketch for fast and accurate tracking of quantiles in data streams. This sketch is heavily inspired by the recently introduced DDSketch, and is based on a novel bucket collapsing procedure that allows overcoming the intrinsic limits of the corresponding DDSketch procedures. Indeed, the DDSketch bucket collapsing procedure does not allow the derivation of formal guarantees on the accuracy of quantile estimation for data which does not follow a sub-exponential distribution. On the contrary, UDDSketch is designed so that accuracy guarantees can be given over the full range of quantiles and for arbitrary distribution in input. Moreover, our algorithm fully exploits the budgeted memory adaptively in order to guarantee the best possible accuracy over the full range of quantiles. Extensive experimental results on both synthetic and real datasets confirm the validity of our approach.

Quantile dependence of tourism activity between Southern European countries

ABSTRACTUsing monthly data on arrivals at tourist accommodation establishments, we estimate the quantile dependence of tourism activity between four Southern European countries: Greece, Italy, Spain and Portugal. We find evidence of strong positive dependence between tourism activities across a wide range of quantiles. Most of the dependence is driven by right-tail events; i.e. tourism sectors seem to grow together. Our results, therefore, suggest that the efforts of a joint European tourism policy appear to have some merit, as tourism activity already behaves similarly across countries, and therefore, pursuing joint policies should be less risky and less difficult to implement.