R&D Investments, Debt Capital, and Ownership Concentration: A Three-Way Interaction and Lag Effects on Firm Performance in China's Pharmaceutical Industry

The existing literature has yet to provide consistent evidence on the relationship between R&D investments and firm performance. The current study attempted to fill this gap in the literature by examining the effect of lag structure and the moderating role of financial governance, in terms of debt capital and ownership concentration, on the returns of R&D. Analyzing a sample of China's pharmaceutical firms from 2009 to 2018, we found that the effect of R&D upon growth begins in the second year after R&D spending and increases thereafter. There exists a vigorous debate about the choice between debt and ownership structure. To fill this gap, we proposed a three-way interactive effect. The results suggest that firms that invest heavily in R&D may achieve their highest performance when the use of debt capital and the extent of ownership concentration are both low. This study contributes to the R&D investments and financial governance literature by reconciling previous mixed evidence about the returns of R&D and the debt–equity choices on R&D investment decisions.

Language and discounting future value: language as a moderator of internationalization–innovation performance relationships

PurposeResearch on relationships between firms' degree of internationalization (DOI) and innovation performance has been mixed, and moderators of these relationships need to be explored. We focus on patents granted as an indicator of innovation performance and explore the moderating role of firms' home-country languages on the DOI–innovation performance relationship. We argue that in countries with languages that always require speakers to grammatically mark the future, firms will focus less on the future value of patents, which should moderate DOI–innovation performance relationships.Design/methodology/approachWe use an unbalanced panel consisting of 567 firm-year observations of 64 different large automotive suppliers from the year 2007 through 2019. This database was analyzed using negative binomial models with a 3-years lag structure and firm controls.FindingsResults show a U-shaped DOI–innovation performance relationship, but only for firms from countries that do not always require speakers to grammatically mark the future. Additionally, a firm's status as being from a country where dominate languages always require speakers to grammatically mark the future has a direct negative relationship with innovation performance.Research limitations/implicationsLimitations are that the sample included a large number of firms from one country (i.e. Japan) and focused on a single industry. Additionally, we used a narrow operational definition of innovation performance (i.e. patents) and relied upon a single methodology. In terms of implications, we identify one moderator that helps explain mixed results of past DOI–innovation performance research, and we identify a direct relationship between language and innovation performance. Hence, future research in this area should control for the dominant language type of firms' home countries.Originality/valueTo our knowledge, this is the first study to examine how language moderates DOI–innovation performance relationships and also relates to innovation performance directly.

Lag structure of BOLD signals within white matter in resting-state fMRI

Previous studies have demonstrated that BOLD signals in gray matter in resting-state functional MRI (RSfMRI) have variable time lags. We investigated the corresponding variations of signal latencies in white matter within 1393 subjects (both sexes included) from the Brain Genomics Superstruct Project. We divided the dataset into ten equal groups to study both the patterns and reproducibility of latency estimates within white matter. We constructed time delay matrices by computing cross-correlation functions between voxel pairs. Projections of voxel latencies were highly correlated (average Pearson correlation coefficient = 0.89) across the subgroups, confirming the reproducibility and structure of signal lags in white matter. We also applied a clustering analysis to identify functional networks within white matter. Analysis of latencies within and between networks revealed a similar pattern of inter- and intra-network communication to that reported for gray matter. Moreover, a unidirectional path, from inferior to superior regions, of BOLD signal propagation was revealed by higher resolution clustering. The variations of lag structure within white matter are associated with different sensory states (eyes open vs eyes closed, and eyes open with fixation vs. eyes closed). These findings provide additional insight into the character and roles of white matter BOLD signals in brain functions.Significance StatementFunctional MRI (fMRI) has had major impacts on clinical and basic neuroscience, and it has been used extensively to study the functional role and spatiotemporal organization of gray matter in different states. However, functional MRI signals from white matter have usually been ignored or even identified as artifacts. We used fMRI data from 1393 subjects to demonstrate (1) fMRI BOLD signals in white matter are robustly detectable in a resting state and exhibit a reproducible, spatiotemporal organization, similar to gray matter; (2) functional networks within white matter can be obtained by applying clustering analysis on the white matter connectivity matrix; (3) the pattern of signal latencies within and between networks resembles the results for gray matter. Further studies on the Beijing EOEC dataset II also revealed that the variations of latencies within white matter alter with different sensory (visual) states. Our findings demonstrate the that resting-state BOLD signals within white matter should be incorporated into comprehensive models of brain function.

Econometrics and econometric modeling

The textbook covers a wide range of issues related to econometric modeling. Regression models are the core of econometric modeling, so the issues of their evaluation, testing of assumptions, adjustment and verification are given a significant place. Various aspects of multiple regression models are included: multicollinearity, dummy variables, and lag structure of variables. Methods of linearization and estimation of nonlinear models are considered. An apparatus for evaluating systems of simultaneous and apparently unrelated equations is presented. Attention is paid to time series models. Detailed solutions of the examples in Excel and the R software environment are included. Meets the requirements of the federal state educational standards of higher education of the latest generation. For undergraduate and graduate students studying in the field of "Economics", the curriculum of which includes the disciplines "Econometrics"," Econometric Modeling","Econometric research".

Price and income elasticities for alcohol in Sri Lanka

Aims: To study the effect of changes in income as well as prices of beer and arrack, on alcohol sales in Sri Lanka during the period 1981–2017. Design: The analyses were conducted by means of ARIMA time series analysis for arrack and beer separately. Measures: Yearly data on the sales of beer and arrack in the entire country were used. National GDP figures were used as a proxy measure of income, and yearly price data for arrack and beer were from Colombo, the largest metropolitan area in the country. Results: No short-term effects of changes in price or income were found on sales. However, changes in the price of beer were significant, with a lag structure, which implies a delayed effect of price changes on the sales of beer. A significant cross-price effect of changes in the price of beer on arrack sales was found as well. Conclusions: Sales of alcohol in Sri Lanka are not affected by price changes to the same extent as in high income countries. Most likely, the explanations will be found in the different drinking cultures between Sri Lanka and these countries.