Statistical Decision Problems and Bayesian Nonparametric Methods

This paper considers parametric statistical decision problems conducted within a Bayesian nonparametric context. Our work was motivated by the realisation that typical parametric model selection procedures are essentially incoherent. We argue that one solution to this problem is to use a flexible enough model in the first place, a model that will not be checked no matter what data arrive. Ideally, one would use a nonparametric model to describe all the uncertainty about the density function generating the data. However, parametric models are the preferred choice for many statisticians, despite the incoherence involved in model checking, incoherence that is quite often ignored for pragmatic reasons. In this paper we show how coherent parametric inference can be carried out via decision theory and Bayesian nonparametrics. None of the ingredients discussed here are new, but our main point only becomes evident when one sees all priors—even parametric ones—as measures on sets of densities as opposed to measures on finite-dimensional parameter spaces.     

Sovereign ceilings “lite”? The impact of sovereign ratings on corporate ratings

Although credit rating agencies have gradually moved away from a policy of never rating a corporation above the sovereign (the ‘sovereign ceiling’), it appears that sovereign credit ratings remain a significant determinant of corporate credit ratings. We examine this link using data for advanced and emerging economies over the period of 1995–2009. Our main result is that a sovereign ceiling continues to affect the rating of corporations. The results also suggest that the influence of a sovereign ceiling on corporate ratings remains particularly significant in countries where capital account restrictions are still in place and with high political risk.     

Exploring sources and outcomes of trust and commitment to nonprofit organizations: The case of Amnesty International Portugal

The interactions between customers and organizations have been regarded as a core aspect in cocreating and developing new products and practices. The current study explores these interactions in the context of citizens and nonprofit organizations. The study comprises two parts: (a) a qualitative approach using interviews and the analysis of social networks to better understand the organization's social media marketing practices and corresponding outcomes and (b) a quantitative approach to analyze antecedents and outcomes of trust and being committed to participate in a nonprofit organization such as Amnesty International. From the qualitative approach, the findings highlight that Amnesty International uses social media (Facebook) to share information, and citizens tend to consider sharing posts, but they do not express new ideas or interact with the organization. These findings contribute to structure the model analyzed in the second stage of the research. Regarding the qualitative approach, the results show that organizational credibility is the most significant driver of trust followed by communicative ability. Affectively commitment to participation and relationship perception act as outcomes of overall commitment.     

Collaboration at work: An integrative multilevel conceptualization

The term collaboration has been used throughout a variety of research disciplines to describe multiple types of interaction; yet, a unified, comprehensive definition of the construct remains elusive. This lack of clarity regarding the distinctions and commonalities between collaboration and other interaction concepts has resulted in conceptual confusion that affects practice and research in human resource management. Practitioners see collaboration as more of a buzzword than as an effective human resource strategy. Previous theory development efforts have not yet taken a comprehensive multidisciplinary approach. This has resulted in failure to integrate key themes across disciplines into an overall view of collaboration, which is a commonplace practice in business and military sectors alike. This paper describes a multidisciplinary conceptualization of collaboration and discusses the implications of this integrative theory to human resource management and strategy development as well as future research efforts.     

Risk factors in Australian bond returns

This study examines the risk factors in Australian bond returns. The study quantifies bond liquidity and estimates a liquidity risk factor in the Australian setting. We develop a three‐factor asset pricing framework that uses term, default and liquidity risk factors to explain the variation of Australian bond returns. Our findings corroborate the US evidence on the pervasiveness of these risk factors faced by bond investors. The three‐factor model developed in this study has practical applications when calculating the cost of debt, evaluating the performance of an active bond fund manager and hedging underlying risk in a bond portfolio.     

A Proposed Model for Consumer-Based Brand Equity Analysis on Social Media Using Data Mining and Social Network Analysis

Online social networks have expanded their “virtual borders,” making the Internet more like an environment of social interaction than a business tool. However, even before the emergence and expansion of social media, marketing professionals were interested in identifying consumers' perceptions about brands. Thus, operational models have been proposed to facilitate such a task. Those models, however, can be expensive and inconvenient, since the models use questionnaires for data collection. To help overcome this problem, this article proposes a model for brand equity analysis from the consumer perspective expressed in social networks using opinion mining techniques and social network analysis. The application of the proposed model on data collected from Twitter made it possible to analyze five brand equity dimensions: brand awareness, brand loyalty, perceived sentiment, perceived quality, and brand associations. The results reached by the application of the model show that brand equity can be analyzed from data retrieved from virtual social networks, disclosing how consumers perceive brands in such an environment, without using questionnaires, enabling different brands in different contexts. Those data can be analyzed under both objective and replicable criteria for each of the brand equity elements that make up the model.     

Quantile serial dependence in crude oil markets: evidence from improved quantilogram analysis with quantile wild bootstrapping

We examine the quantile serial dependence in crude oil prices based on the Linton and Whang’s quantile-based portmanteau test which we improved by means of quantile wild bootstrapping (QWB). Through Monte Carlo simulation, we find that the quantile wild bootstrap-based portmanteau test performs better than the bound testing procedure suggested by Linton and Whang. We apply the improved test to examine the efficiency of two crude oil markets – WTI and Brent. We also examine if the dependence is stable via rolling sample tests. Our results show that both WTI and Brent are serially dependent in all, except the median quantiles. These findings suggest that it may be misleading to examine the efficiency of crude oil markets in terms of mean (or median) returns only. These crude oil markets are relatively more serially dependent in non-median ranges.     

Stock market interdependence between Australia and its trading partners: does trade intensity matter?

We investigate the extent and manner of stock market interdependence between Australia and its trading partners and examine whether this is affected by trade intensity. Based on trade intensity, we classify Australia’s trading partners into major, medium and minor partners. We hypothesize that markets with greater (lower) trade intensity will be more (less) interdependent with Australia. We perform correlation (unconditional and conditional) analyses between Australia and its trading partners. Our results indicate that most of the markets that are highly correlated with Australia are its major trading partners. We conduct panel regression analysis to investigate whether trade intensity has any impact on the stock market correlations between Australia and its trading partners. The results show that trade intensity significantly and positively affect the correlations of Australia with its major trading partners. Thus, the results confirm our hypothesis that trade intensity drives stock market interdependence between Australia and its trading partners.     

How pervasive is the effect of culture on stock market linkages? Evidence across regions and economic cycles

We conduct a comprehensive study on the effect of culture on stock market linkages. With data on 25 national stock markets, a quantile regression model is used to estimate the determinants of market linkages using culture variable/s such as language, religion and Hofstede’s cultural dimensions while controlling for distance, economic and legal variables. Further, we test whether these effects hold across regions and if changes are detected during periods of market crisis. We also test if market liquidity, an indicator of market efficiency, diminishes the impact of culture on market linkages. The main conclusion is that culture preferences shape investor choices, which affects integration between stock markets. The equity markets with similar cultural traits tend to increase market linkages; however, we observe differences across regions. Furthermore, liquidity and economic uncertainty fail to have an impact on the significance of culture variable/s as determinants of market linkages.