Model selection for nonlinear time series

We investigate the finite-sample performance of model selection criteria for local linear regression by simulation. Similarly to linear regression, the penalization term depends on the number of parameters of the model. In the context of nonparametric regression, we use a suitable quantity to account for the Equivalent Number of Parameters as previously suggested in the literature. We consider the following criteria: Rice T, FPE, AIC, Corrected AIC and GCV. To make results comparable with other data-driven selection criteria we consider also Leave-Out CV. We show that the properties of the penalization schemes are very different for some linear and nonlinear models. Finally, we set up a goodness-of-fit test for linearity based on bootstrap methods. The test has correct size and very high power against the alternatives investigated. Application of the methods proposed to macroeconomic and financial time series shows that there is evidence of nonlinearity.First version received: September 2002/Final version received : October 2003I would like to thank Cees Diks, Cars Hommes and an anonymous referee for useful comments that significantly improved the paper.     

Stackelberg equilibrium with multiple firms and setup costs

I provide conditions that guarantee that a Stackelberg game with a setup cost and an integer number of identical leaders and followers has an equilibrium in pure strategies. The main feature of the game is that when the marginal follower leaves the market the price jumps up, so that a leader’s payoff is neither continuous nor quasiconcave. To show existence I check that a leader’s value function satisfies the following single crossing condition: When the other leaders produce more the leader never accommodates entry of more followers. If demand is strictly logconcave, and if marginal costs are both non decreasing and not flatter than average costs, then a Stackelberg equilibrium exists. Besides showing existence I characterize the equilibrium set and provide a number of results that contribute to the applied literature. As the number of leaders increases, leaders produce more and eventually they deter entry. Leaders produce more than the Cournot best reply, but they may underinvest in entry deterrence. As the number of followers increases, leaders become more aggressive. When this number is large, if leaders can produce the limit quantity and at the same time have market power, then they deter entry.     

Impact of Segment‐level Natural Resource Operational Risk Reporting on Earnings Predictions

Non‐financial reports alert investors to operational risks associated with issues such as insufficient access to natural resource inputs and related costly interruptions to production, while segment‐level reports alert investors to operational risk distribution across a firm. An important issue, to date unexplored, is how segment‐level non‐financial reporting has an impact on earnings predictions. We report the results of an experiment used to examine how mining company segment‐level water reports affect investors' earnings predictions, where water reports indicate whether the firm and its segments will have access to sufficient water to meet production needs. We find that investors do not change their earnings predictions when firm and segment‐level reports indicate low water risk but they do revise down their earnings predictions when firm and segment‐level water reports indicate high water risk. This is consistent with investors responding to the additional information provided in segment‐level reports confirming that water risk is high across the firm. Regardless of whether firm‐level water reports indicate high or low water risk, when segment‐level reports indicate that one segment is low water risk and another is high water risk, investors revise down their earnings predictions. This is consistent with investors recognizing that natural resource operational risk concentration in one segment can affect earnings more than evenly‐distributed risk. Overall, our findings suggest that belief‐adjustment theory explains how investors react to prospective operational risk information contained in segment‐level water reports according to the similarity of the segment‐level risks, and that this information is factored into earnings predictions.     

中国上市公司可转换债券非理性转股行为研究

以上海证券交易所上市的30只已进入转股期的可转换债券为样本研究,发现约58.5%的自愿转股行为违背了理性原则.用非参数回归分析法进行实证检验,发现市场流动性并不是影响非理性转股的因素,而隔夜风险、转股损失则是导致非理性转股行为发生的影响因素;投资者的心理因素对非理性转股具有非线性影响:历史收益率对投资者的影响取决于趋势信念和参考点两个因素.此外,非理性转股还受到框定依赖的影响,即当历史收益为正时,高的收益波动会增加非理性转股;而当历史收益为负时,高的收益波动会减少非理性转股.     

环境扰动对创新活动中非知识协同行为和知识协同行为的影响

知识协同行为和非知识协同行为是知识管理领域新的研究主题,探讨其内在影响机理具有重要意义。通过数据调查,利用结构方程模型深入分析和探讨了环境扰动对知识协同行为和非知识协同行为的影响机理。实证研究表明：环境扰动正向影响非知识协同行为的产生,而环境扰动对知识协同行为的影响并不显著;良性非知识协同行为与企业创新绩效显著正相关,而恶性非知识协同行为对企业创新绩效有一定反向影响,知识协同行为对组织创新绩效有着显著正向影响;良性非知识协同行为对知识协同行为有较强的正向影响,而恶性非知识协同行为反向影响良性非知识协同行为的产生,这意味着恶性非知识协同行为可以通过良性非知识协同行为的中介效应对知识协同行为、创新绩效起反向作用。     

非CEO高管独立性的治理效应——基于审计定价的证据

从高管团队内部治理角度出发,当非CEO高管相对于CEO具有较强的独立性时,可以对CEO的不当行为发挥“自下而上”的制约与平衡作用。基于会计信息外部监督者的视角,通过考察会计师事务所对客户非CEO高管独立性的反应,来揭示企业非CEO高管独立性的治理效应。研究发现：非CEO高管独立性越高,则会计师事务所的审计定价越低。机制检验表明,上述结果主要因企业财务报表错报风险和经营风险下降导致。此外,当非CEO高管约束能力更强时,非CEO高管独立性能够更加有效地发挥治理效应,进而降低事务所的审计定价;且这一关系在董事会监督效率较低、事务所类型为“非十大”的企业当中尤为明显。进一步研究表明,非CEO高管独立性提高了企业收到标准无保留意见的概率;而且即使审计定价降低,审计质量也并未下降。研究结论为明晰非CEO高管独立性的治理机制提供了重要依据,对于上市公司优化管理层结构以及事务所进行审计决策具有一定的借鉴意义。     

Using Non Market Valuation to Inform the Choice Between Permits and Fees in Environmental Regulation

The purpose of this short note is to open an exploration regarding the use of non market valuation to help guide the selection of economically efficient pollution control instruments. As long as non market valuation techniques can correctly estimate the slope of the marginal benefit of abatement curve, this information along with engineering cost estimates of the unit costs or slope of the marginal abatement cost will provide useful information to policy makers in choosing between fees and permits. An illustrative review of the literature suggests that both stated and revealed preference methods have estimated slopes of marginal benefit functions for reducing several pollutants. To investigate the efficiency of permits versus fees, an illustrative review of corresponding marginal abatement costs is also made. For air pollutants affecting visibility, the slope of the marginal benefit curve is far greater than the slope of the marginal abatement costs, suggesting permits as the efficient instrument. For nitrates in groundwater used for drinking, the marginal benefit curve is flatter than the rather steep marginal abatement cost, suggesting fees/taxes would be a more efficient economic instrument. We hope this note stimulates more emphasis in non market valuation on estimating the slope of the marginal benefit function to enhance environmental economists ability to make policy recommendations regarding the choice of pollution instruments for specific pollutants.      

非常规性工作对员工创造性的影响:一个被调节的中介模型

以工作特性理论为基础,运用多元回归模型探索非常规性工作对个体创造性的影响机制,检验自我效能感的中介作用及授权型领导的调节作用。运用Mplus7.0、SPSS19.0及PROCESS软件对258份员工及主管的有效问卷进行数据分析,结果显示,非常规性工作与员工创造性显著正相关;自我效能感在非常规性工作对员工创造性的影响过程中起完全中介作用,授权型领导在非常规性工作对自我效能感的影响过程中起调节作用。进一步地,授权型领导显著正向调节自我效能感在非常规性工作与创造性之间的中介作用,即在高授权型领导情境下,上述中介作用增强,反之减弱。     

A Review on Dimension Reduction

Summarizing the effect of many covariates through a few linear combinations is an effective way of reducing covariate dimension and is the backbone of (sufficient) dimension reduction. Because the replacement of high‐dimensional covariates by low‐dimensional linear combinations is performed with a minimum assumption on the specific regression form, it enjoys attractive advantages as well as encounters unique challenges in comparison with the variable selection approach. We review the current literature of dimension reduction with an emphasis on the two most popular models, where the dimension reduction affects the conditional distribution and the conditional mean, respectively. We discuss various estimation and inference procedures in different levels of detail, with the intention of focusing on their underneath idea instead of technicalities. We also discuss some unsolved problems in this area for potential future research.     

NON‐LINEAR DYNAMICS,COMPLEXITY AND RANDOMNESS: ALGORITHMIC FOUNDATIONS

Abstract In this paper I make an attempt to describe, discuss and extend a few aspects of the rich mathematical tapestry that can be woven with rigorous notions of non‐linear dynamics, complexity and randomness, in terms of algorithmic mathematics. It is a tapestry that I try to weave with economic analysis, economic theory and economic modelling in mind. All three notions – that is, non‐linear dynamics, complexity and randomness – have a rich conceptual, modelling or analytic tradition in core areas of economic theory, both at the micro and macro levels. It is the algorithmic foundation I try to provide for them that could be considered the novel contribution in this paper. Once the algorithmic foundations are in place, it is, for example, almost natural to consider the famed difficulties of obtaining closed form solutions for non‐linear, complex or random dynamic models in economics almost a trivial vestige of a pre‐simulation era in mathematical modelling.