我国沿海城市海洋经济效率的测度及其影响因素分析——基于两阶段双重自助抽样DEA方法

由于传统 DEA方法在对决策单元的效率评价及效率影响因素的分析上存在缺陷,本文运用基于自助抽样的 DEA方法,利用我国 12个主要沿海城市的相关数据,对它们的海洋经济效率值进行了测算,同时分析了影响沿海城市海洋经济效率的因素。研究发现,经济愈发达的沿海城市其海洋经济效率未必更高,但海洋经济发展水平对沿海城市海洋经济效率具有正向影响。同时,外贸依存度高的沿海城市海洋经济效率实际更低,表明其海洋资源可能存在过度开发的问题。此外,加强沿海城市的环境保护力度尧改善环境治理有助于海洋经济效率的提升。     

农户收入差距的微观基础:物质资本还是人力资本?

本文基于中国8个省份、1320个农户、跨度15年的微观面板数据,使用非参数回归方法对影响农户收入差距的原因加以分析,利用广义差分方法去除农户异质性以后,估计的结果发现:(1)教育和在职培训体现出的人力资本是拉大农户收入差距的主要原因;(2)物质资本,包括土地,对于农户收入的差距没有显著影响;(3)在不同的收入组别上,人力资本的回报都显著高于物质资本的回报。     

似无关动态协整模型中的水平扭曲与校正

本文在解析似无关动态协整模型及其动态最小二乘估计的基础上,从理论上揭示了关于协整参数的假设检验存在严重的水平扭曲,即对协整参数约束的Wald检验统计量的渐近卡方分布存在严重的有限样本扭曲。进一步,本文应用自举抽样技术对水平扭曲进行了有效校正。基于本文的发现,我们建议在对似无关动态协整模型中的参数进行假设检验时,为保证结论的准确性,应使用自举抽样推断技术产生统计量值并由此来形成检验结论。     

Blockwise bootstrap wavelet in nonparametric regression model with weakly dependent processes

In this paper, we suggest a blockwise bootstrap wavelet to estimate the regression function in the nonparametric regression models with weakly dependent processes for both designs of fixed and random. We obtain the asymptotic orders of the biases and variances of the estimators and establish the asymptotic normality for a modified version of the estimators. We also introduce a principle to select the length of data block. These results show that the blockwise bootstrap wavelet is valid for general weakly dependent processes such as α-mixing, φ-mixing and ρ-mixing random variables.     

政策支持、企业投资与旅游业效率

本文对30家旅游上市公司2009—2018年的数据包络分析效率值进行纠偏修正,并构建政策支持与企业投资对于旅游企业运营效率的影响因素模型,运用Bootstrap截断回归法进行实证分析。研究结果表明,旅游上市公司整体上处于非有效状态;政府补贴与金融支持会对企业运营效率产生抑制作用;企业投资对效率的提升作用是决定性的,远远大于政府补贴和金融支持的负面作用。     

基于粗糙集的产业集群竞争力评价研究——以绍兴纺织产业集群为例

为了弥补产业集群传统评价的不足,改进评价结果精度,笔者以绍兴纺织产业集群为例,构建了基于粗糙集的产业集群竞争力评价模型,并利用Bootstrap方法进行仿真处理.结果表明,基于粗糙集的产业集群竞争力评价方法为评价演算过程的客观性提供了良好的保证,Bootstrap方法的仿真结果比小样本下的评价精度有明显的改进,更符合绍兴纺织产业集群的实际发晨情况.     

明星基金溢价效应:“高技术”还是“好运气”?

本文研究我国股票型、偏股型开放式基金的评级与未来业绩的关系。根据晨星公司的基金评级数据,每月构造1星级至5星级的基金投资组合,并采用Carhart四因子模型对组合收益进行风险调整,研究发现5星级基金每年能获得2%的超额收益率,5星级与1星级基金的套利组合年超额收益为6%。此外,采用自助法对超额收益率的进一步检验,本文发现,明星基金溢价与基金经理的选股能力无关,与基金经理的好运气有关。     

My Chancy Life as a Statistician

In this short article, I will attempt to provide some highlights of my chancy life as a statistician in chronological order spanning over 60 years, 1954 to present.     

On the efficacy of stop-loss rules in the presence of overnight gaps

A stop-loss rule is a risk management tool whereby the investor predefines some condition that, upon being triggered by market dynamics, implies the liquidation of her outstanding position. Such a tool is widely used by practitioners in financial markets with the hope of improving their investment performance by cutting losses and consolidating gains. We analyze in this work the performance of four popular implementations of stop-loss rules applied to asset prices whose returns are modeled with consideration of overnight gaps, that is, jumps from the closing price of one day to the opening price of the next trading day. In addition, our models include acute momentary price drops (flash crashes), which are often believed to erode the performance gains that might be derived from stop-loss rules. For this analysis we consider different models of asset returns: random walk, autoregressive and regime-switching models. In addition, we test the performance of the considered stop-loss rules in a non-parametric, data-driven framework based on the stationary bootstrap. As a general conclusion we find that, even when including overnight gaps and flash crashes in our price models, in rising markets stop-loss rules improve the expected risk-adjusted return according to most metrics, while improving absolute expected return in falling markets. Furthermore, we find that in general the simple fixed percentage stop-loss rule may be, in risk-adjusted terms, the most powerful among the popular rules that this work considers.