我国CEO报酬影响因素的实证研究

本文以2003—2005年间深沪市场上的CEO报酬数据为样本．采用因子分析和多元回归方法进行分析，发现CEO总报酬和CEO个人特征、公司规模、公司盈利及发展能力、董事会、现金流情况显著正相关．和股权结构及环境风险显著负相关。我国CEO报酬和行业盈利能力及发展能力、市场业绩之间不存在显著相关关系。CEO年薪除了上述影响因素外．还和监事会情况显著负相关，和主营业务情况显著正相关，但是相关系数的绝对值都较小，表明监事会情况和主营业务情况对CEO年薪的影响较小．而CEO持股价值和CEO个人特征及公司盈利能力显著正相关，和股权结构及环境风险显著负相关。     

基于修剪平均的神经网络集成时序预测方法

由于复杂时序存在结构性断点和异常值等问题，往往导致预测模型训练效果不佳，并可能出现极端预测值的情况。为此，本文提出了基于修剪平均的神经网络集成预测方法。该方法首先从训练数据中生成多组训练集，然后分别训练多个神经网络预测模型，最后将多个神经网络的预测结果使用修剪平均策略进行集成。相较于简单平均策略而言，修剪平均策略不容易受到极值的影响，能够使集成模型获得鲁棒性强的预测效果。在实证研究中，本文构造了两种神经网络集成预测模型，分别为基于修剪平均的自举神经网络集成模型(Trimmed Average based Bootstrap Neural Network Ensemble, TA-BNNE)和基于修剪平均的蒙特卡洛神经网络集成模型(Trimmed Average based Monte Carlo Neural Network Ensemble, TA-MCNNE)，并采用这两种模型对NN3竞赛数据集进行预测，结果表明在常规和复杂数据集上，修剪平均策略比简单平均策略具有更好的预测精度。此外，本文将所提出的集成模型与NN3的前十名模型进行比较，发现两种模型在全部数据集上均超过了第6名，在复杂数据集上的表现均超过了第1名，进一步验证本文所提方法的有效性。     

董事激励与公司业绩--实验的证据

我们在实验框架下检验了董事会成员的激励对股东财富和CEO报酬的影响,以及该报酬与公司业绩之间的敏感度.文中提出了两种任命董事的方法,一种由CEO任命,另一种由最大的股东自动担任董事.由董事会决定CEO的报酬,而CEO负责企业的生产、投资和分红决策.投资者根据接收到的每个公司的分红、资本收益这些信息,通过买卖这些企业的股票来调整他们的资产组合. 我们发现,薪酬与业绩之间的敏感度随董事持股比例的增加而上升.此外,当大股东作为董事会成员时,经济体所产生的财富(股东财富是其中的一部分)更大;而当CEO选择董事时,结果是缺乏效率的.本文讨论了关于董事职能和相关代理成本这一研究结果,以及标准会计框架下,限制执行人员报酬和报表标准化要求.     

基于PSO-SVR模型的能源需求预测

本文将支持向量回归机(support vector regression,SVR)与粒子群算法(Particle Swarm Optimizat ion,PSO)相结合,选取1 985~2008年的能源需求量及其影响因素作为学习样本,利用粒子群算法对支持向量机参数进行优化,通过训练、测试得到具有良好学习与推广能力的PSO-SVR能源需求预测模型。建立BP神经网络模型,并将两者的预测值进行对比,结果表明,PSO-SVR模型预测精度优于BP神经网络模型。     

基于优化回声状态网络的股价预测研究

股票市场是一个充斥着各种噪声的动态非线性系统,能够精确地对其进行预测是一项具有挑战性的任务。本文构建的BCC-ESN模型,是运用细菌群体趋药性算法(BCC)来优化回声状态网络(ESN)的权值结构,在继承ESN优良性质的同时,具有更高的模型预测能力。实验证明,BCC-ESN模型比前馈神经网络具有更好的学习和预测能力。经对上证指数进行短期价格预测,结果与BP网络、Elman网络和ESN网络进行比较,BCC-ESN模型精度明显优于其他三种网络预测。同时,在运算效率上,BCC-ESN模型继承了ESN的运算优势,明显优于其他神经网络预测模型,是一种切实可行、高效的预测算法,尤其在股票时间序列预测中有广泛的实用价值。针对BCC-ESN模型在训练预测中遇到的问题,比如耗费时间过长和过度拟合问题,本文亦提供了简单易行的思路和方法。     

广州软件产业社会网络与竞争关系复杂网络的分析与比较

通过对广州软件产业的问卷调查,建立了61个公司的CEO自我中心社会网络、2个公司的衍生网络以及578个公司的基于产品的竞争关系网络模型,并分别采用社会网络分析方法、复杂网络分析方法对上述模型进行了分析与比较。发现了广州软件产业CEO自我中心社会网络具有较强的封闭性,竞争关系复杂网络具有幂律分布、小世界性等特点,以及该产业的公司衍生网络与竞争关系复杂网络之间的内在联系。     

CEO网络嵌入性与企业价值

以社会网络理论为理论依据,本文构建了CEO网络嵌入性与企业价值关系的概念模型,提出CEO网络嵌入性与企业价值变化正相关的假设。运用结构方程方法的模型发展策略,选取2009年A股非金融保险业上市公司为样本,对概念模型与实证数据进行契合,得到一个在统计上与样本数据能够良好适配的理论模型。通过对模型进行实证检验,证实了本文所提出的理论假设和概念模型的合理性。     

基于两阶段优化算法的神经网络预测模型

采用基于两阶段优化算法(multi-stage optimization approach,MSOA)的GA人工神经网络,将训练集分为两部分,在前一训练集训练后获得的网络基础上使用后一训练集进行进一步的训练获得更为优化的网络结构.针对复杂系统建模输入节点难以确定的问题,提出将其与自组织数据挖掘算法相结合,利用GMDH算法获得神经网络的初始化节点,使用训练好的神经网络模型进行预测.将由此建立的预测模型应用于粮食价格的预测,并进一步探讨了MSOA算法的收敛性问题.结果表明基于GMDH和MSOA的神经网络组合预测模型能较大提高神经网络的全局收敛能力和收敛速度,提高预测精度.     

基于BP神经网络的电子企业业绩评价

本文针对电子行业的特点建立了评价电子企业业绩的指标体系,然后应用BP网络方法对网络模型进行训练,并选取了10家上市的电子企业应用训练好的神经网络进行评价,最后得到了总体的排名结果。本文研究的目的在于让企业能够看到自身的不足并加以改善,使企业得到更好的发展。     

考虑定性指标及误判损失的企业违约判别神经网络模型

识别和度量企业的违约风险是银行风险管理中很重要的一项工作.目前企业违约判别模型离实际应用还具有一定差距,表现在:1)模型所使用的样本基本都是配对模式,不能代表整体样本;2)很少直接引入影响违约的定性指标,如行业,地区和规模;3)没有考虑到误判损失的非对称性.针对上述问题,本文应用前向BP网络针对某国有商业银行的2003年全部有效的短期贷款企业的财务数据,引入了定性指标,采用全样本进行训练,最后确定使误判损失最小的切割点,这样就得到优化的神经网络模型.     

Company Performance,Corporate Governance,and CEO Compensation in Norway and Sweden

This paper examines the relationshipamong company performance, corporategovernance arrangements, and CEO compensationwithin the Scandinavian countries of Norway andSweden. Our sample consists of 224 tradedcompanies, 120 of which are from Norway and 104from Sweden. The empirical evidence fromboth Norway and Sweden reveals significantpositive relationships among board size andCEO compensation, foreign board membership andCEO compensation, and market capitalization andCEO compensation. A significant negativerelationship is found between CEO ownership andCEO compensation. In addition, no significantrelationship is found between companyperformance and CEO compensation or CEO tenureand CEO compensation, except in the case ofNorwegian firms when a change in market-to-bookperformance measure is used. Despite itslimited geographical scope, the study adds toour general understanding of internationalcorporate governance issues.     

Executive Discretion and Corporate Performance as Determinants of CEO Compensation,Contingent on External Monitoring Activities

In this paper we discuss the potential impactof managerial discretion and firm performanceon CEO compensation, contingent on the extentof monitoring activities. We argue that CEOcompensation may be positively related toexecutive discretion and corporate performancefor firms that have vigilant external monitors. We alternatively contend that CEO compensationmay neither be associated with discretion norperformance for firms with passive externalmonitors. The empirical results are supportiveof our contentions.     

管理决断权、高管薪酬与企业绩效的实证研究

本文从企业层面上研究了管理决断权对高管薪酬的影响及两者与企业绩效之间的权变关系。论文选取2002-2007年深沪两市上市公司作为研究样本,构建了管理决断权的多指标衡量体系,证明管理决断权对高管薪酬和企业绩效具有重要的决定作用,企业绩效好的情况下高管薪酬与管理决断权的联系更为紧密。同样,在高管理决断权情况下高管薪酬与企业绩效的挂钩更加紧密,但同时企业会面临监管难题。     

Fresh in the saddle: The influence of a new CEO's vision and origin,and CEO succession type on market actors' reactions

CEO succession is a critical event in the life of a company. How external stakeholders respond to it, can affect the company's valuation. This study investigates how securities analysts' and investors' reactions to CEO succession are affected by the interplay between the charisma of the new CEO's vision, the new CEO's origin (whether an insider or outsider), and the type of CEO succession (whether routine, dismissal or interim). Drawing on the literature on signaling, we suggest that because a charismatic vision emits a positive signal about the company's future performance, it will affect market actors' reactions by either weakening or strengthening the influence of the signals emitted by other succession context contingencies, namely, CEO origin and succession type. To test our predictions regarding analysts' and investors' reactions, we respectively analyze panel data and conduct an event study. The results support most of our predictions. We discuss the study's contributions and implications.     

多元化、公司业绩与总经理变更

总经理变更是公司内部治理的重要手段,其业绩敏感性受到多种因素的影响,以2002年～2007年沪深两市702家A股上市公司的数据为样本,探讨多元化经营对总经理变更绩效敏感性及其业绩改进效应的影响,为公司经营模式对公司治理效率的影响提供经验证据,利用传统的Logistic回归、边际效应分析以及面板数据随机效应回归对多元化和总经理变更的关系进行研究。实证结果表明,多元化降低了总经理变更的业绩敏感性,多元化公司总经理变更的概率与资产收益率代表的多元化公司业绩变化没有显著的关系,即使在公司业绩下降的条件下,股东替换总经理的动机也弱于集中化公司;多元化公司总经理变更后的资产重组行为规模显著弱于集中化公司,带来的业绩改进也小于集中化公司。     

衰退企业CEO变革的理论假说与实证检验

在企业扭转理论中,许多学者主张CEO变革是企业成功扭转的前提.这一主张基于3个基本的理论假设:①CEO变革可以提高企业运营绩效;②CEO变革可以促进企业战略的变革;③CEO变革可以维持企业合法性,即维持企业相关利益者对企业的继续支持.采用我国上市公司的相关数据,对这3个理论假说进行了实证检验.研究结果表明,CEO变革的绩效假说并没有得到支持;CEO变革的合法性假说得到了部分支持;CEO变革的战略变革假说得到了显著支持.因此,在衰退困境中,企业变革CEO的目的主要是希望能够更客观地认识企业衰退原因和现有战略的问题,从而制定出更为符合企业实际情境的扭转战略.由此推断,衰退企业变革CEO的本质是改变企业的战略认知框架.     

CEO self-discipline in power use: A key moderator for the effect of CEO power

The effect of CEO power depends ultimately on how it is used (i.e., a CEO’s approach to power use), which may vary from one CEO to another. Notwithstanding extensive research on the effect of CEO power on organizational outcomes, researchers have thus far paid very limited attention to how the effect depends on individual differences. In this study I propose a new construct of CEO self-discipline in power use—defined as a CEO’s appeared conformance to the prescribed leadership norms (in particular, norms regarding how leaders should use their power)—and examine how it moderates the effect of CEO power. With a longitudinal dataset from the U.S. computer hardware and software industries, I found that CEO self-discipline in power use weakened the positive effect of CEO power on performance extremeness and improved the effect of CEO power on firm performance.     

The influence of CEO equity incentives on licensing

In a study of life science firms, we find that, in accordance with predictions drawn from agency theory and behavioral agency theory, CEO stock ownership is negatively associated with licensing while CEO stock options are positively associated with licensing. Furthermore, by combining theoretical insights from the capabilities literature with both agency theory and behavioral agency theory, we predict that a key measure of capabilities in the licensing context—a firm's alliance experience—significantly influences the ways in which CEO equity incentives impact licensing. More specifically, we find that, in accordance with our theoretical predictions, alliance experience positively (negatively) moderates the relationship between CEO stock ownership (CEO stock options) and licensing. Our study contributes to the wider literature on the determinants of licensing by examining whether licensing is sensitive to CEO equity incentives. We also extend the capabilities literature on licensing by examining the contrasting influences of a firm's alliance experience on the relationship between CEO equity incentives and licensing. Our findings also inform behavioral agency-based research on the effects of equity incentives by highlighting the usefulness of a capabilities perspective in augmenting our understanding of the behavioral role of CEO equity incentives.     

Industrial Origin of CEOs in Outside Succession: Board Preference and Stockholder Reaction

Although scholars have studied outside executive succession for decades, there is still no clear definition of the term outsider. We focus on a new dimension of outsiderness, the industrial background of executives hired from outside the firm into a CEO position. This paper examines the antecedents of boards' decisions on the industry origin of outside successors as well as the short-term stock market response. We find that firms with more independent boards and with blockholders owning large amounts of stock are more likely to hire industry unrelated successors. However, the board's decision does not strictly follow the rhetoric of stockholder's interests. The stock market reacts more positively to outside CEO succession announcements when the CEO comes from an industry related firm. These findings support our theoretical arguments that the boards of directors may use a logic or rule of appropriateness in deciding the industrial origin of outside successors.