Backtesting Stochastic Mortality Models

Abstract

This study sets out a backtesting framework applicable to the multiperiod-ahead forecasts from stochastic mortality models and uses it to evaluate the forecasting performance of six different stochastic mortality models applied to English & Welsh male mortality data. The models considered are the following: Lee-Carter’s 1992 one-factor model; a version of Renshaw-Haberman’s 2006 extension of the Lee-Carter model to allow for a cohort effect; the age-period-cohort model, which is a simplified version of Renshaw-Haberman; Cairns, Blake, and Dowd’s 2006 two-factor model; and two generalized versions of the last named with an added cohort effect. For the data set used herein, the results from applying this methodology suggest that the models perform adequately by most backtests and that prediction intervals that incorporate parameter uncertainty are wider than those that do not. We also find little difference between the performances of five of the models, but the remaining model shows considerable forecast instability.     

US elections and monthly stock market returns

Using monthly market returns over a period of 104 years, we investigate possible relationships between stock market performance and various occurrences in American elections. Unlike most prior studies, we find little relationship between the two. In the relatively few cases where we do find statistically significant relationships, the degree of explanatory power is quite small. Specifically, market returns do not appear to vary based on partisan control of the government, a result that is robust to the inclusion or exclusion of macroeconomic control variables. Further, the often-discussed “second-half” effect, which predicts higher returns during the second half of a given presidential term, turns out to be both weaker and less straightforward than is commonly believed. Overall, neither election results nor the election cycle appears to offer much help in predicting stock market returns.      

Dilemmas in the management of temporary work agency staff

Temporary jobs account for an increasing proportion of new engagements in the UK labour market, with temporary work agencies or 'labour market intermediaries' occupying a central role in the regulation of entry into some organisations. Such evolving arrangements have been found to have their contradictions, even for the host organisation. This article explores the internal and external pressures to use a temporary work agency as a means of recruiting labour at host organisations. It considers some of the HRM issues that stem from the use of such workers, including the tendency to devolve HRM to the managers of such agencies operating within the host organisation. Central to this article is a consideration of the potential sustainability of organisations' use of temporary agency workers, engaging with this concern from the perspective of organisational cost‐effectiveness.     

The role of food service in tourist satisfaction

The paper is based on empirical investigation carried out amongst a 341-sample population visiting the Black Sea resorts of Romania in August 1997. The purpose of the study was to investigate the role and importance of food service in tourist satisfaction and to note any differences in satisfaction levels between regional groups. Analysis of the findings revealed that significant differences existed between three tourist groups’ satisfaction perceptions of value for money, quality of food, number of dishes, standard of food service, variety of dishes, presentation of food and speed of service in general; and of bread, coffee, meat and soup in particular. Overall results further indicated that quality of food, value for money, variety of dishes, attractiveness of surroundings and presentation of food were the attributes that most affected the overall food service experience in Romania. The paper concludes that food service is an important contributor to tourist satisfaction and that there are significant differences in satisfaction levels with food service between eastern and western European, and Romanian tourists.     

The Demand for Broad Money in the United Kingdom, 1878–1993

Using annual data from Friedman and Schwartz (1982), Hendry and Ericsson (1991a) developed an empirical model of the demand for broad money in the United Kingdom over 1878–1975. We update that model over 1976–1993, accounting for changed data definitions and clarifying the concept of constancy. With appropriate measures of opportunity cost and credit deregulation, the model's parameters are empirically constant over the extended sample, which was economically turbulent. Policy implications follow for parameter nonconstancy and predictive failure, causation between money and prices, monetary targeting, deregulation and financial innovation, and the effect of policy on economic agents' behavior.     

Common and Exclusive Agencies and Partial Delegation

Within the context of common agency, the article examines separationand partial delegation. The first entails breaking the multiple-taskagency into smaller single-task agencies, whereas the secondallows principals to substitute some of their efforts for theagency's. With respect to the first alternative, it is shownthat common agency can be sustained against separation by thepresence of strong complements. For the second alternative,principals partially delegate their effort and retain the rightto influence agency output at a later stage of the game. Althoughthis tends to weaken agency incentives, principals may preferthis version of common agency than one under full delegationand may prefer it to an exclusive one.     

Discrimination by Gender and Disability Status: Do Worker Perceptions Match Statistical Measures?

We explore whether perceptions of discrimination are related to ordinary statistical measures. The majority of disabled respondents report feeling some discrimination due to their disability, the majority of women feel some discrimination because of their gender, and a surprising number of men also report some discrimination. We do not find a strong link between perceptions of discrimination and measured discrimination perhaps because those who perceive discrimination feel that it occurs along other dimensions than pay. However, we do find a connection between whether a person feels his or her income is inadequate and measured discrimination for all groups studied.     

HMO coverage of cosmetic procedures: Response to market competition

Theory suggests that insurance policies should not cover purely cosmetic procedures. This paper attempts to explain empirical variation in coverage of a purely cosmetic procedure. Health maintenance organizations (HMOs) are modeled in a Cournot competition with two decisions—the number of policies and the limits placed on medical care utilization. Comparative statics for an individual market suggest, first, that restrictiveness is positively associated with the number of HMOs. Second, restrictiveness is positively associated with the predisposition to demand managed care products. Third, each effect is strengthened by increasing the other parameter. Restriction of tattoo ablation, a purely cosmetic procedure, is examined empirically. The first and second predictions are supported at low levels of predisposition and numbers of HMOs, while the third prediction is not supported. The data were obtained under a grant from the Office of Technology Assessment (HMO Laser Procedure Coverage) and a University of Michigan Rackham Pre-Doctoral Fellowship (InterStudy and County Business Patterns).     

Does the law of one price hold in China? Testing price convergence using disaggregated data

This study examines to what extent prices diverge across China and how long it takes prices to converge following idiosyncratic shocks. We consider monthly data using disaggregated goods prices from 36 cities in China. Following [Imbs, J., Mumtaz, H., Ravn, M., and Rey, H., 2005. PPP strikes back: aggregation and the real exchange rate. Quarterly Journal of Economics 120, 1–44.], we use two estimation methods: a fixed effect method when considering goods individually and a mean group estimation specification for a panel including all goods simultaneously. The mean group method also accounts for dynamic heterogeneity across goods. Impulse response functions are obtained to calculate half-lives. With both methods, we find half-lives of only a few months or less, supporting the conjecture that convergence rates within a country are faster than rates estimated in an international context. However, the half-life reported here is still shorter than that for other studies using disaggregated intracountry data. Perhaps a lower degree of specialization and market differentiation in developing countries creates a greater potential for price convergence.