A fast and scalable ensemble of global models with long memory and data partitioning for the M5 forecasting competition

This work presents key insights on the model development strategies used in our cross-learning-based retail demand forecast framework. The proposed framework outperforms state-of-the-art univariate models in the time series forecasting literature. It has achieved 17th position in the accuracy track of the M5 forecasting competition, which is among the top 1% of solutions.     

Price matching and product differentiation strategies considering showrooming

Showrooming, a phenomenon in which customers use brick-and-mortar stores to assess products and then purchase them from online retailers (o-retailers) for lower prices, is considered a great threat to traditional retailers (t-retailers). To combat showrooming, many t-retailers have executed price matching which enables customers to pay o-retailers' lower prices for the identical product. To avoid direct competition with t-retailers who execute price matching, many o-retailers have begun to sell differentiated products from t-retailers, which weakens the information advantage to customers from practicing showrooming. Motivated by these observations, we construct a duopoly game, where a t-retailer and an o-retailer sell products in a same category, to study the profitabilities of product differentiation and price matching in the context of showrooming. The results show that in the scenario without price matching, the o-retailer is likely to benefit from product differentiation only when the o-retailer's differentiated product is more popular with customers than the t-retailer's product. However, in the price matching scenario, the o-retailer also has the opportunity to benefit from product differentiation when the o-retailer's differentiated product is less popular with customers than the t-retailer's product, and product differentiation can be a win-win strategy for the two retailers under certain conditions. Considering the o-retailer's product differentiation decision, the t-retailer is only likely to execute price matching if the non-digital attributes of the product category sold by two retailers are not very obvious.     

粤港澳大湾区制造业集聚与区域创新

以2009—2020年粤港澳大湾区11个城市为研究对象，探讨制造业集聚与区域创新的关系。研究发现，制造业集聚显著提高区域创新水平，可通过提高企业竞争这一路径实现。进一步分析发现，金融业集聚对区域创新发展具有显著的正向关系，金融产业和制造业产融协同集聚显著提高区域创新水平。研究结果表明，制造业集聚加剧产业内部企业创新竞争，带动区域创新发展，且金融业在创新中起到积极的推动作用。指出应鼓励区域企业创新竞争，优化营商环境，为产业融合发展提供保障。     

Doing more and doing better are two different entities: Different patterns of family control and environmental performance

Family firms bear two types of agency costs, including type I and type II agency problems, in corporate environmental practices: (1) Outside executives at family firms hesitate to engage in environmental strategies, which can lead to drops in profits; (2) Controlling families employ opportunistically environmental management to achieve their interests. We argue that a primary cause for the agency problems lies on ineffective internal corporate governance at family firms, which can cause loss of managerial (or power) balance between outside executives and family executives. Our findings show that family firms with ownership and strategic control (FSC), which family executives and outside executives monitor and constrain each other, can achieve the highest environmental performance. Moreover, external controls, including product market competition and provincial environmental regulations, substitute effective internal control of FSC. The environmental performance premium of FSC is more prevalent when the production market competition is lower. Family firms with ownership, operational, and strategic control (FOSC) can achieve higher environmental performance within a province with more stringent environmental regulations.     

Does local government competition aggravate haze pollution? A new perspective of factor market distortion

Haze pollution has become a new threat to China's sustainable development, but it may be that local government behaviour can play an important role in the prevention and control of pollutants. A dynamic spatial autoregressive (SAR) model is used to study the relationship between local government competition and haze pollution. To further explore the indirect impact of factor market distortion on haze pollution and control potential endogeneity problems, a newly developed intermediary effect model that incorporates the characteristics of the generalized method of moments (GMM) is utilized to explore how factor market distortion indirectly affects haze pollution. The research results show that regional haze pollution in China is characterized by significant spatial correlation, and local government competition has a positive impact on haze pollution; that is, local government competition exacerbates haze pollution. In general, local government competition not only directly leads to an increase in haze pollution but also further intensifies it by distorting the local factor market, and the intermediary role of factor market distortion is approximately 7.04%. The results of the regional inspection found that competition among local governments in the eastern region did not lead to haze pollution, and distortion of the factor market did not exist as an intermediary effect. However, both direct and intermediary effects are significant in the central and western regions. Therefore, an official performance appraisal system that includes ecological constraints should be established to guide the benign transformation of local government competition, and an environmental management mechanism must be developed for joint prevention and control to reduce haze pollution. In addition, the free flow of factors and marketization are equally important.     

领导干部自然资源资产离任审计与企业绿色并购——来自中国重污染型上市公司的经验证据

基于2011~2017年我国A股重污染型上市公司经验数据，实证检验了领导干部自然资源资产离任审计对企业绿色并购行为的影响，研究发现：领导干部自然资源资产离任审计可显著促进辖区内企业实施绿色并购；而当地区市场竞争程度越高时，越会抑制企业在领导干部自然资源资产离任审计试点实施背景下的绿色并购扩张行为；进一步地，企业绿色并购会显著提高企业价值，且这一影响对并购后当年的企业价值影响最为显著；基于异质性分析方面，领导干部自然资源资产离任审计会约束不同产权性质企业的环境治理行为，促进其采取绿色并购，实施绿色发展；同时领导干部自然资源资产离任审计会更有利于东部地区企业实施绿色并购，而对中西部地区的企业影响不明显。     

提高出口退税能够“稳就业”和“稳外贸”吗?

近年来,全球贸易保护主义抬头加剧了各国之间的贸易摩擦。本文以中美贸易为例,探讨中国出口退税政策在稳就业、稳外贸以及应对贸易摩擦中的作用机制。首先基于1994-2020年季度数据发现:中国就业的波动较平稳;净出口则呈现高波动特征;中国就业和净出口呈现弱顺周期。这些周期特征与其他国家存在显著差异。其次构建了一个含有不完全金融市场和价格不完全传递的对称两国开放经济DSGE模型,数值模拟发现:在本国出口退税冲击和它国技术冲击下,模型能够较好地拟合中国就业和净出口的周期特征,其中财富效应和由贸易条件变化引起的支出转移效应是主要的内在传导机制。借助模型对中美贸易摩擦的反事实实验发现:中国单方面提高1%出口退税时,中国就业增长0.05%,净出口增长0.28%,呈现较强持续性,提高出口退税能够稳就业和稳外贸;当中国提高1%出口退税和美国提高1%进口关税时,中国就业增长0.03%,净出口增长0.16%,呈现较强持续性,勒纳中性不成立,此时出口退税在稳就业和稳外贸中的作用尽管有所削弱,但依然有效。     

Tax evasion by tax deferment: Sham litigation with an informal credit market

We show theoretically how tax evasion is facilitated by informal credit market through tax deferment. Our model is empirically based. Using sham litigation, tax evaders earn a higher rate of return than the stipulated penalty rate for tax evasion while the government loses tax revenue. We propose an upfront part–payment of the disputed amount of tax as a solution to the form of tax evasion we describe.     

M5 competition uncertainty: Overdispersion,distributional forecasting,GAMLSS, and beyond

The M5 competition uncertainty track aims for probabilistic forecasting of sales of thousands of Walmart retail goods. We show that the M5 competition data face strong overdispersion and sporadic demand, especially zero demand. We discuss modeling issues concerning adequate probabilistic forecasting of such count data processes. Unfortunately, the majority of popular prediction methods used in the M5 competition (e.g. lightgbm and xgboost GBMs) fail to address the data characteristics, due to the considered objective functions. Distributional forecasting provides a suitable modeling approach to overcome those problems. The GAMLSS framework allows for flexible probabilistic forecasting using low-dimensional distributions. We illustrate how the GAMLSS approach can be applied to M5 competition data by modeling the location and scale parameters of various distributions, e.g. the negative binomial distribution. Finally, we discuss software packages for distributional modeling and their drawbacks, like the R package gamlss with its package extensions, and (deep) distributional forecasting libraries such as TensorFlow Probability.