How many factors are important in U.K. stock returns?

I use the sequential approach of Harvey and Liu ([2018]. Lucky factors (Working Paper). Duke University) to build linear factor models in U.K. stock returns among a set of 13 candidate factors using individual stocks and three groups of test portfolios between July 1983 and December 2017. My study finds that the Market factor is the dominant factor in reducing mispricing in individual stocks and test portfolios regardless of the pricing error metric used. The Market factor has a bigger impact when using a value weighting pricing error metric. Whether a second factor is used or not depends upon which metric is used for mispricing and the time period examined. My study finds support for a two-factor model for the whole sample period of the Market factor and the Conservative Minus Aggressive (CMA) factor of Fama and French ([2015]. “A five-factor asset pricing model.” Journal of Financial Economics 116: 1–22) when giving greater weight to the mispricing of larger companies.     

Optimal granularity for portfolio choice

Many optimization-based portfolio rules fail to beat the simple 1/N rule out-of-sample because of parameter uncertainty. In this paper we suggest a grouping strategy in which we first form groups of equally weighted stocks and then optimize over the resulting groups only. This strategy aims at balancing the trade-off between the benefits from optimization and the losses from estimation risk. We rely on Monte-Carlo simulations to illustrate the performance of the strategy, and we derive the optimal group size for a simplified setup. Furthermore, we show that estimation risk also has an impact via the criterion by which the assets are sorted into groups (like the expected excess returns or betas), but does not negate the grouping approach. We relate our work to linear asset pricing models, and we conduct out of sample back-tests in order to confirm the validity of our grouping strategy empirically.     

How should we estimate value-relevance models? Insights from European data

We study the consequences of unobserved heterogeneity when employing different econometric methods in the estimation of two major value-relevance models: the Price Regression Model (PRM) and the Return Regression Model (RRM). Leveraging a large panel data set of European listed companies, we first demonstrate that robust Hausman tests and Breusch-Pagan Lagrange Multiplier tests are of fundamental importance to choose correctly among a fixed-effects model, a random-effects model, or a pooled OLS model. Second, we provide evidence that replacing firm fixed-effects with country and industry fixed-effects can lead to large differences in the magnitude of the key coefficients, with serious consequences for the interpretation of the effect of changes in earnings and book values per share on firm value. Finally, we offer recommendations to applied researchers aiming to improve the robustness of their econometric strategy.     

Forecasting US consumer price index: does nonlinearity matter?

The objective of this article is to predict, both in sample and out of sample, the consumer price index (CPI) of the US economy based on monthly data covering the period of 1980:1–2013:12, using a variety of linear (random walk (RW), autoregressive (AR) and seasonal autoregressive integrated moving average (SARIMA)) and nonlinear (artificial neural network (ANN) and genetic programming (GP)) univariate models. Our results show that, while the SARIMA model is superior relative to other linear and nonlinear models, as it tends to produce smaller forecast errors; statistically, these forecasting gains are not significant relative to higher-order AR and nonlinear models, though simple benchmarks like the RW and AR(1) models are statistically outperformed. Overall, we show that in terms of forecasting the US CPI, accounting for nonlinearity does not necessarily provide us with any statistical gains.     

上海金融发展与经济增长关系的实证研究

本文从实证上研究上海金融发展与经济增长之间的长期和短期关系。协整检验、格兰杰因果关系检验和误差修正模型估计结果表明,上海金融发展与经济增长存在着长期稳定的均衡关系,上海金融发展的规模、结构、效率与经济增长之间存在单向因果关系,即经济增长是金融发展规模的格兰杰原因、金融发展结构是经济增长的格兰杰原因、经济增长是金融发展效率的格兰杰原因,滞后2期的金融发展的效率是经济增长的格兰杰原因。从短期来看,上海金融发展对经济增长的影响也是比较显著的。基于以上研究结论,本文提出完善上海金融结构与提高金融效率的政策建议。     

新疆农村金融发展与经济增长关系的实证分析

本文依据农村金融发展和农村经济增长的实际数据,依次使用平稳性检验、协整检验和格兰杰因果检验等计最经济学方法,对新疆农村金融发展和经济增长进行了实证检验.结果表明:新疆农村金融发展的指标和农村经济增长之间均存在着显著的长期正向相关关系,新疆农村经济是农村金融的Granger原因,反之却不成立.     

多元化集团中的母合优势分析

以2002～2006年实施多元化战略的集团公司为研究对象,利用分层线性模型的实证检验结果表明:母公司显著地影响了上市子公司的经营绩效,反映了母公司对多元化战略实施的重要作用.但通过进一步分析发现,对于实施非相关多元化战略的集团公司而言,母公司对子公司的干预更具有价值性,结果多元化集团中的母合优势分析支持了"母合优势"理论;而对于实施相关多元化战略的集团公司而言,母公司的干预会损害子公司的经营绩效.     

农村社会保障支出对农村居民消费的影响的实证研究

从社会保障对居民消费影响机制的微观基础出发,运用协整分析并建立向量误差修正模型对我国农村社会保障支出的消费效应进行了实证研究.结果表明,我国农村社会保障支出没有对农村居民消费支出产生促进作用.因此,应加快农村社会保障制度的改革,促进农村居民消费,拉动内需.     

人民币现金境外需求规模的间接测算研究:1999-2008

本文基于1978年至2008年的年度数据和缺口估算法,通过构建一个货币需求函数的计量模型,对1999年至2008年的人民币境外持有规模进行了估算,估算结果表明,在1999年至2008年间,境外人民币的需求占人民币现金总量的比率均值为7.62%,且流通规模呈逐年递增之势。最后,本文进行了总结并提出相关政策建议。     

区域耕地数量变化预测方法的对比研究

研究目的:对区域耕地数量变化预测方法进行对比分析,科学合理确定区域耕地数量。研究方法:通过Pearson相关分析法筛选影响耕地变化的关键因子,利用改进的BP神经网络算法、灰色模型和多元线性回归模型法对费县耕地数量变化进行预测。研究结果:改进的BP神经网络算法相对传统的灰色模型和多元线性回归模型等预测方法具备较高的预测精度。研究结论:改进的BP神经网络算法是进行耕地数量变化预测的较好方法,具有实际指导意义,其预测结果可以为当地相关部门合理地确定耕地保有量、推进耕地保护提供科学依据。