Determinants of Current Account Imbalances in 16 OECD Countries: An Out-Of-Sample Perspective

We employ panel data as well as country-specific models to uncover empirically the determinants of current account imbalances. We find evidence of slope heterogeneity likely rendering the fixed-effect estimator inconsistent. Mean group estimation is followed to overcome the latter difficulty. Both estimation techniques are evaluated by means of in- and out-of-sample criteria. The mean group estimator outperforms the fixed-effect approach and, moreover, only three economic variables are found to impact on the current account balance significantly. These are the government budget balance, the domestic output gap and the changes of the terms of trade. We finally propose a parsimonious dynamic model of the current account, including only these variables and illustrate that it has similar predictive accuracy as country-specific benchmark models. JEL no. F32, C23, C53     

遗失物拾得人报酬请求权及标准研究

新颁布的物权法规定了遗失物拾得制度,虽然立法者并没有按照国际惯例在物权法中充分肯定所有遗失物拾得人报酬请求权,然而通过一些调查分析,我们发现这与我们的国情又是相符的。看来,对所有遗失物拾得人报酬请求权的肯定还要循序渐进。     

Unusual operating cash flows and stock returns

The current financial reporting of cash flows from operations does not present individual sources of these cash flows, making it difficult for investors to assess a firm’s future performance. I hand-collect individual cash flows from unusual operations and examine their characteristics for predicting future cash flows. The results show that the unusual individual cash flow items contain a significant incremental predictive ability for future cash flows. Additional return tests show that stock prices fail to fully reflect their predictive value, suggesting that the current reporting practice may mislead investor perceptions of a firm’s cash generating ability and investors could benefit from a more explicit presentation of cash flows from operations.     

Predicting European Takeover Targets

This article extends the Palepu (1986) acquisition likelihood model by incorporating measures of a technical nature, e.g. momentum, trading volume as well as a measure of market sentiment. We use the proposed model to predict takeover targets in a large sample of European and cross‐border merger and acquisition deals and validate its performance on an in‐ and out‐of‐sample basis. The robustness of the proposed model is investigated across several dimensions. In addition we explore the ability of the model to form the basis of successful takeover timing investment strategies. The results of our empirical analysis suggest that the proposed model predicts European takeover targets with relatively high accuracy and is able to determine portfolios that earn significant returns which are not explained by conventional risk factors.     

从现代产权制度着眼 深入推进国有林权改革

在论述国有林区建立现代产权制度重要意义的基础上,对伊春的林权制度改革目标和内容进行了阐述。认为伊春林区实行的以职工家庭对国有林地进行承包经营为主要内容的产权形式,是国有林区的一种优越的经济制度选择;其核心目标是在国有森林资源领域真正建立起归属清晰、责权明确、保护严格、流转顺畅的现代产权制度。     

水价对中国北方城市居民用水需求影响分析

介绍城市居民用水需求变化的背景，对北方城市居民用水需求进行了预测。得出结论：水价对于减少城市居民用水需求的作用非常显著；城市化的高速发展会对城市居民用水造成巨大压力，对城市居民用水价格、用水管理手段和农业节水制度的积极改革，为解决城市居民用水供需缺口提供可能。     

基于GRA&BPNN的广西粮食产量预测研究

[目的]研究粮食产量的影响因素并以其相关性为基础预测粮食产量对实现广西粮食产业的"做强做优"具有重大意义。通常情况下粮食产量与种植技术发展水平、农田耕地面积、土地肥力、气候等诸多因素相关,但是在样本数据不足、数据间关联度不明显的情况下则无法采用回归分析、灰系统等常用预测方法。[方法]文章应用灰色关联分析方法得到水库水量、农田有效灌溉面积、第一产业从业人口、播种面积、除涝面积等5个与广西粮食产量关系最为密切的因子变量,同时取2004~2012年的数据作为学习、训练样本,以2013~2014年的数据为试报样本,并以此建立BP神经网络粮食预测模型。[结果]检验结果表明运用本模型预测粮食产量具有较高的精度和良好的泛化性。[结论]根据模型结果,该文提出提升广西粮食产业发展的可行性建议,即加强水库的管理、引导与粮食产业相关的产业、稳定粮食种植面积、加强洪涝灾害的防御和治理能力、推进农业信息化。     

基于GNNM(1.1)模型的货运量及货运周转量预测

通过各种算法对货运指标进行有效预测,对于把握未来货运发展趋势有着非常重要的作用。文章将灰色预测与神经网络预测方法进行了有机结合,建立了一个基于灰色神经网络的预测(GNNM)模型。通过模型对货运量及货运周转量进行了预测,得到了较满意的结果,表明了模型具有较高的可靠性及实用性。     

基于LVQ神经网络的电信企业客户流失预测模型研究

随着人工智能的快速发展,人工神经网络被广泛地运用到分类预测领域。文章首先明确了客户流失的定义及其分类,然后分析了LVQ神经网络的基本原理,最后从研究样本的确定、预测变量的选取、模型的训练及评估三个方面构建了基于LVQ神经网络的电信企业客户流失预测模型,以期为电信企业客户流失预测模型的设计提供一定的借鉴意义。     

基于自组织数据挖掘的电子商务客户流失预测模型

为解决电子商务客户流失预测中的高维、非线性问题,本文将自组织数据挖掘理论(SODM)引入客户流失预测,提出一种新颖的基于自组织数据挖掘的电子商务客户流失预测模型。该方法将自组织数据挖掘中的客观系统分析算法(OSA)和改进分组数据处理网络(GMDH)集成起来进行电子商务客户流失预测。首先利用OSA算法选择出重要的电子商务客户流失关键属性,然后将训练样本送入改进GMDH网络进行学习和训练,进而对测试样本客户流失状态进行判别。将该方法应用于某网上商店客户流失预测实证分析,预测结果验证了该方法对包含多种因素影响的电子商务客户流失预测具有优势,基于自组织数据挖掘的电子商务客户流失预测模型具有较强的实用性和可操作性。