门限分位数自回归模型(threshold quantile autoregressive model, 简记为TQAR)是一种非线性分位数回归模型, 主要用于讨论系统中的门限效应. 在TQAR模型中, 自回归阶数与门限值的确定等, 都会影响模型分析效果. 为此,本文给出模型定阶、门限值估计及门限效应检验等方法. 数值模拟结果表明, TQAR模型在门限值估计、回归系数估计的有限样本表现方面都优于传统的门限均值自回归模型(threshold autoregressive model, 简记为TAR)及门限均值自回归条件异方差(TAR-GARCH)模型. 最后, 将TQAR模型应用于中国股市收益的自相关性研究, 实证结果证实收益序列的自相关性呈现出明显的门限效应和异质效应. 这一发现, 有助于准确刻画股市收益动态变化规律, 为重新认识金融市场运行机制提供了一个实证基础.
Abstract
The threshold quantile autoregressive (TQAR) model is a kind of nonlinear quantile regression model. It can be mainly used to test the threshold effects in a system. It is important to decide the optimal lag order of the autoregression and the threshold value that have huge impact on the performance of the TQAR model. To this end, we propose some methods for model selection, threshold parameter estimation, and threshold effect test. The numeric simulation results show that the TQAR model is superior to the TAR and TAR-GARCH model in terms of the accuracy of the threshold parameter estimation and regression coefficients estimation. Finally, we apply the TQAR model to reveal auto-correlation of stock returns in China. The empirical results indicate that there are threshold effects and heterogeneity effects in the auto-correlation. The findings are helpful to give a reasonable description of the dynamics in stock returns, which provides an empirical basis for re-understanding operating mechanism of financial market.
关键词
分位数回归 /
门限自回归 /
股市收益率 /
自相关性
{{custom_keyword}} /
Key words
quantile regression /
threshold autoregression /
stock returns /
auto-correlation
{{custom_keyword}} /
中图分类号:
F224.0
{{custom_clc.code}}
({{custom_clc.text}})
{{custom_sec.title}}
{{custom_sec.title}}
{{custom_sec.content}}
参考文献
[1] Fama E F. Efficient capital markets: A review of theory and empirical work[J]. The Journal of Finance, 1970, 25(2): 383-417.
[2] Veronesi P. Stock market overreactions to bad news in good times: A rational expectations equilibrium model[J]. Review of Financial Studies, 1999, 12(5): 975-1007.
[3] Lewellen J. Momentum and autocorrelation in stock returns[J]. Review of Financial Studies, 2002, 15(2): 533-564.
[4] 饶育蕾, 彭叠峰, 贾文静. 交叉持股是否导致收益的可预测性——基于有限注意的视角[J]. 系统工程理论与实践, 2013, 33(7): 1753-1761. Rao Yulei, Peng Diefeng, Jia Wenjing. Does cross-shareholding cause return predictability?——Based on limited attention perspective[J]. Systems Engineering——Theory & Practice, 2013, 33(7): 1753-1761.
[5] 高秋明, 胡聪慧, 燕翔. 中国A股市场动量效应的特征和形成机理研究[J]. 财经研究, 2014, 40(2): 97-107. Gao Qiuming, Hu Conghui, Yan Xiang. On characteristics and formation mechanisms of momentum effect in China's A-share market[J]. Journal of Finance and Economics, 2014, 40(2): 97-107.
[6] Tong H. On a threshold model[M]. Netherlands: Sijthoff & Noordhoff, 1978.
[7] Tong H. Threshold models in non-linear time series analysis[M]. New York: Springer-Verlag, 1983.
[8] Tsay R S. Testing and modeling threshold autoregressive processes[J]. Journal of the American Statistical Association, 1989, 84(405): 231-240.
[9] Chan K S, Petruccelli J D, Tong H, et al. Consistency and limiting distribution of the least squares estimator of a threshold autoregressive model[J]. The Annals of Statistics, 1993, 21(1): 520-533.
[10] Hansen B E. Sample splitting and threshold estimation[J]. Econometrica, 2000, 68(3): 575-603.
[11] Potter S M. A nonlinear approach to US GNP[J]. Journal of Applied Econometrics, 1995, 10(2): 109-125.
[12] Narayan P K. The behaviour of US stock prices: Evidence from a threshold autoregressive model[J]. Mathematics and Computers in Simulation, 2006, 71(2): 103-108.
[13] Kapetanios G, Shin Y. Unit root tests in three-regime SETAR models[J]. The Econometrics Journal, 2006, 9(2): 252-278.
[14] Hansen B E. Threshold autoregression in economics[J]. Statistics and Its Interface, 2011, 4(2): 123-127.
[15] 靳晓婷, 张晓峒, 栾惠德. 汇改后人民币汇率波动的非线性特征研究——基于门限自回归TAR模型[J]. 财经研究, 2008, 34(9): 48-57. Jin Xiaoting, Zhang Xiaotong, Luan Huide. Modeling nonlinearities in RMB nominal exchange rate fluctuations after RMB exchange rate reform based TAR model[J]. Journal of Finance and Economics, 2008, 34(9): 48-57.
[16] 吴武清, 李东, 潘松, 等. 三阶段均值回复、TAR及其应用[J]. 系统工程理论与实践, 2013, 33(4): 901-909. Wu Wuqing, Li Dong, Pan Song, et al. Three-regime mean reversion, TAR and its application[J]. Systems Engineering——Theory & Practice, 2013, 33(4): 901-909.
[17] 许启发, 蒋翠侠. 分位数局部调整模型及应用[J]. 数量经济技术经济研究, 2011, 28(8): 115-133. Xu Qifa, Jiang Cuixia. Quantile partial adjustment model and its application[J]. The Journal of Quantitative & Technical Economics, 2011, 28(8): 115-133.
[18] Koenker R, Bassett Jr G. Regression quantiles[J]. Econometrica, 1978, 41(1): 33-50.
[19] Koenker R, Xiao Z. Quantile autoregression[J]. Journal of the American Statistical Association, 2006, 101(475): 980-990.
[20] Baur D G, Dimpfl T, Jung R C. Stock return autocorrelations revisited: A quantile regression approach[J]. Journal of Empirical Finance, 2012, 19(2): 254-265.
[21] Caner M. A note on least absolute deviation estimation of a threshold model[J]. Econometric Theory, 2002, 18(3): 800-814.
[22] Cai Y, Stander J. Quantile self-exciting threshold autoregressive time series models[J]. Journal of Time Series Analysis, 2008, 29(1): 186-202.
[23] Galvao A F, Montes-Rojas G, Jose O. Threshold quantile autoregressive models[J]. Journal of Time Series Analysis, 2011, 32(3): 253-267.
[24] Hansen B E. Inference when a nuisance parameter is not identified under the null hypothesis[J]. Econometrica, 1996, 64(2): 413-430.
[25] Ju H, Su L, Xu P. Pricing for goodwill: A threshold quantile regression approach[R]. Shanghai: Shanghai University of Finance and Economics, 2013.
[26] Tversky A, Kahneman D. The framing of decisions and the psychology of choice[J]. Science, 1981, 211(4481): 453-458.
[27] Barberis N, Shleifer A, Vishny R. A model of investor sentiment[J]. Journal of Financial Economics, 1998, 49(3): 307-343.
[28] Kahneman D. Maps of bounded rationality: Psychology for behavioral economics[J]. American Economic Review, 2003, 93(5): 1449-1475.
[29] 陈蓉, 陈焕华, 郑振龙. 动量效应的行为金融学解释[J]. 系统工程理论与实践, 2014, 34(3): 613-622. Chen Rong, Chen Huanhua, Zheng Zhenlong. Behavioral finance interpretation of momentum effect[J]. Systems Engineering——Theory & Practice, 2014, 34(3): 613-622.
[30] Koul H L, Qian L, Surgailis D. Asymptotics of M-estimators in two-phase linear regression models[J]. Stochastic Processes and Their Applications, 2003, 103(1): 123-154.
{{custom_fnGroup.title_cn}}
脚注
{{custom_fn.content}}
基金
国家自然科学基金(71071087); 教育部人文社会科学研究规划基金项目(14YJA790015); 安徽省哲 学社会科学规划基金项目(AHSKY2014D103)
{{custom_fund}}