Value-at-Risk Prediction: A Comparison of Alternative Strategies

Given the growing need for managing financial risk, risk predictionplays an increasing role in banking and finance. In this studywe compare the out-of-sample performance of existing methodsand some new models for predicting value-at-risk (VaR) in aunivariate context. Using more than 30 years of the daily returndata on the NASDAQ Composite Index, we find that most approachesperform inadequately, although several models are acceptableunder current regulatory assessment rules for model adequacy.A hybrid method, combining a heavy-tailed generalized autoregressiveconditionally heteroskedastic (GARCH) filter with an extremevalue theory-based approach, performs best overall, closelyfollowed by a variant on a filtered historical simulation, anda new model based on heteroskedastic mixture distributions.Conditional autoregressive VaR (CAViaR) models perform inadequately,though an extension to a particular CAViaR model is shown tooutperform the others.     

FISCAL DEFICITS AND MACROECONOMIC PERFORMANCE IN DEVELOPING COUNTRIES

Although fiscal adjustment was urged on developing countriesduring the 1980s to lead them out of economic malaise, considerableuncertainty remains about the relations between fiscal policyand macroeconomic performance. To illustrate how financial markets,private spending, and the external sector react to fiscal policies,the behavior of holdings of money and public debt, private consumptionand investment, the trade balance, and the real exchange rateis modeled for a sample of ten developing countries. The studiesfind strong evidence that over the medium term, money financingof the deficit leads to higher inflation, while debt financingleads to higher real interest rates or increased repressionof financial markets, with the fiscal gains coming at increasinglyunfavorable terms. Consumers respond differently to conventionaltaxes, unconventional taxes (through inflation or interest andcredit controls), and debt financing, in ways that make fiscaladjustment the most effective means of increasing national saving.Private investment—but not private consumption—issensitive to the real interest rate, which rises under domesticborrowing to finance the deficit. Contrary to the popular presumption,in some countries private investment increases when public investmentdecreases. There is strong evidence that fiscal deficits spillover into external deficits, leading to appreciation of thereal exchange rate. Fiscal deficits and growth are self-reinforcing:good fiscal management preserves access to foreign lending andavoids the crowding out of private investment, while growthstabilizes the budget and improves the fiscal position. Thevirtuous circle of growth and good fiscal management is oneof the strongest arguments for a policy of low and stable fiscaldeficits.      

Analyst forecast characteristics and the cost of debt

We examine the relation between analyst forecast characteristics and the cost of debt financing. Consistent with the view that the information contained in analysts’ forecasts is economically significant across asset classes, we find that analyst activity reduces bond yield spreads. We also find that the economic impact of analysts is most pronounced when uncertainty about firm value is highest (that is, when firms have high idiosyncratic risk). Our findings are robust to controls for private information in equity prices and level of corporate disclosures. Overall, the results indicate that the information contained in analyst forecasts is valued outside the equity market and provide an additional channel in which better information is associated with a lower cost of capital.     

Inductive research in last-mile delivery routing: Introducing the Re-Gifting heuristic

The last-mile problem presents a daunting challenge for many logistics service providers, especially some 7000 small, localized operations for whom the cost of complex software solutions is often prohibitive. As a result, last-mile dispatchers rely on simple heuristics to ensure adequate customer service at an acceptable cost. This research effort extends prior qualitative work by developing and testing a simple vehicle routing heuristic, based on behaviors observed in practice, that prioritizes customer service over cost against other simple vehicle routing heuristics across a variety of environments using simulation. The results support the inclusion of a customer service focus in vehicle routing and the addition of such heuristics to existing algorithm portfolios, specifically in urban areas with well-developed highway systems.     

A Model of Bargaining over Hazardous Waste Cleanup

We put forward a formal model of a bargaining problem in which two parties suspected of contaminating the environment are responsible for clean-up costs. If the parties do not negotiate an agreement on a cost allocation, one will be imposed by the government. This process is commonly used in environmental cleanups performed under the Comprehensive Environmental Response, Compensation and Liability Act (CERCLA or Superfund). Passed by the US Congress in 1980 and administered by the Environmental Protection Agency (EPA), CERCLA provides the U.S. government with authority to manage releases (or threatened releases) of hazardous substances that may endanger public health or the environment. We conclude that potentially responsible parties will be induced to settle only in the face of specific allocations of clean-up and explicit threats. For example, at the Middlefield–Ellis–Whisman Superfund site in Mountain View, California, the responsibilities of the different parties for soil and groundwater contamination are understood by all, yet our negotiation model predicts that without the threat that additional costs will be imposed, agreement on the allocation of clean up costs will never be reached.     

Exploring the relationship of perceived automotive salesperson attributes, customer satisfaction and intentions to automotive service department patronage: The moderating role of customer gender

This study examines the relationship of perceived automotive salesperson attributes and customer satisfaction behavior in the purchase experience leading to the patronage of a service department in an automotive dealership. Logistic regression analyses indicate that gender moderates the relationship of the perceived salesperson attributes and customer satisfaction as well as intentions for service department patronage. The model distinguishes users from non-users of the service department and does this better for females than for males. Finally, the paper concludes with a discussion of the managerial implications and directions for future research.     

Credibility lost: attempting to reclaim an expert identity in an HR professional context

Professional insecurity is a long‐standing concern within HR, with claims to expertise seen as critical to credibility. Considering HR as an epistemic community and drawing on the identity work literature, we examine an identity threat to, and subsequent response by, a training and development (T&D) team. Based on ethnographic exposure to their practice, we explore how team members experience the threat and follow their attempts to re‐establish their position in the local epistemic community, the HR department. We examine both individual and collective identity work, considering how both the identity threat and subsequent responses are embedded within T&D and HR practice more broadly. Through this analysis, we offer academic insight on the nature of HR practice and the construction of claims to expertise.     

Strong measures of group coherence and the probability that a pairwise majority rule winner exists

A Pairwise Majority Rule Winner (PMRW) exists for a voting situation if some candidate can defeat each of the remaining candidates by Pairwise Majority Rule. The PMRW would be very appropriate for selection as the winner of an election, but it is well known that such a candidate does not always exist. This paper concludes a series of studies regarding the probability that a PMRW should be expected to exist in three-candidate elections, by introducing the notion of a strong measures of mutually coherent group preferences. In order for voting situations to be reasonably expected to fail to have a PMRW in a three-candidate election, voters’ preferences must be generated in an environment that is far removed from the situation in which there is a strong-overall-unifying candidate. So far removed, that it is extremely unlikely that a PMRW will not exist in voting situations with large electorates for a small number of candidates.