Does the choice of estimator matter when forecasting returns?

While the literature concerned with the predictability of stock returns is huge, surprisingly little is known when it comes to role of the choice of estimator of the predictive regression. Ideally, the choice of estimator should be rooted in the salient features of the data. In case of predictive regressions of returns there are at least three such features; (i) returns are heteroskedastic, (ii) predictors are persistent, and (iii) regression errors are correlated with predictor innovations. In this paper we examine if the accounting of these features in the estimation process has any bearing on our ability to forecast future returns. The results suggest that it does.

Expressed emotion as predictor of relapse in patients with comorbid psychoses and substance use disorder

Objective: Expressed emotion (EE) and substance use disorder predict relapse in psychosis, but there is little research on EE in comorbid samples. The current study addressed this issue. Method: Sixty inpatients with a DSM-IV psychosis and substance use disorder were recruited and underwent diagnostic and substance use assessment. Key relatives were administered the Camberwell Family Interview. Results: Patients were assessed on the initial symptoms and recent substance use, and 58 completed the assessment over the following 9 months. High EE was observed in 62% of households. Expressed emotion was the strongest predictor of relapse during follow up and its predictive effect remained in participants with early psychosis. A multivariate prediction of a shorter time to relapse entered EE, substance use during follow up Q1 and (surprisingly) an absence of childhood attention deficit hyperactivity disorder. Conclusions: Since high EE is a common and important risk factor for people with comorbid psychosis and substance misuse, approaches to address it should be considered by treating clinicians.

What measure of temperature is the best predictor of mortality?

Hot and cold temperatures significantly increase mortality rates around the world, but which measure of temperature is the best predictor of mortality is not known. We used mortality data from 107 US cities for the years 1987–2000 and examined the association between temperature and mortality using Poisson regression and modelled a non-linear temperature effect and a non-linear lag structure. We examined mean, minimum and maximum temperature with and without humidity, and apparent temperature and the Humidex. The best measure was defined as that with the minimum cross-validated residual. We found large differences in the best temperature measure between age groups, seasons and cities, and there was no one temperature measure that was superior to the others. The strong correlation between different measures of temperature means that, on average, they have the same predictive ability. The best temperature measure for new studies can be chosen based on practical concerns, such as choosing the measure with the least amount of missing data.

On the endogeneity of the pay-performance relationship in professional soccer

In this reply we show that the Nüesch (2009) comment paper to our initial contribution (Torgler and Schmidt 2007) has several shortcomings. He suggests that professional soccer wages seem to buy talent rather than motivation. We therefore provide a larger set of talent proxies and estimations to check whether this assertion is correct. Our results indicate that his conclusion is problematic. We still observe a strong motivational effect, and in some cases the effect is even larger than the talent effect. A further key problem in Nüesch’s contribution is the fact that he neglects to consider the relevance of the relative salary situation.

The significance of endogeneity problems in crash models: An examination of left-turn lanes in intersection crash models

Crash prediction models are used for a variety of purposes including forecasting the expected future performance of various transportation system segments with similar traits. The influence of intersection features on safety have been examined extensively because intersections experience a relatively large proportion of motor vehicle conflicts and crashes compared to other segments in the transportation system. The effects of left-turn lanes at intersections in particular have seen mixed results in the literature. Some researchers have found that left-turn lanes are beneficial to safety while others have reported detrimental effects on safety. This inconsistency is not surprising given that the installation of left-turn lanes is often endogenous, that is, influenced by crash counts and/or traffic volumes. Endogeneity creates problems in econometric and statistical models and is likely to account for the inconsistencies reported in the literature. This paper reports on a limited-information maximum likelihood (LIML) estimation approach to compensate for endogeneity between left-turn lane presence and angle crashes. The effects of endogeneity are mitigated using the approach, revealing the unbiased effect of left-turn lanes on crash frequency for a dataset of Georgia intersections. The research shows that without accounting for endogeneity, left-turn lanes ‘appear’ to contribute to crashes; however, when endogeneity is accounted for in the model, left-turn lanes reduce angle crash frequencies as expected by engineering judgment. Other endogenous variables may lurk in crash models as well, suggesting that the method may be used to correct simultaneity problems with other variables and in other transportation modeling contexts.

Commentaries and Responses to "The Driver Behaviour Questionnaire as a predictor of accidents: A meta-analysis" [Commentaries lead by Anders af Wahlberg; Responses lead by J.C.F. de Winter]

The following discussion is in response to a 2010 article published in the Journal of Safety Research by J.C.F. de Winter and D. Dodou entitled “The Driver Behaviour Questionnaire as a predictor of accidents: A meta-analysis” (Volume 41, Number 6, pp. 463-470, available on sciencedirect.com). The editors are pleased to provide a forum for this exchange and welcome further comments.

Clinical utility of mental state screening as a predictor of intellectual outcomes 6 months after diagnosis of type 1 diabetes

Background Screening tests of basic cognitive status or ‘mental state’ have been shown to predict mortality and functional outcomes in adults. This study examined the relationship between mental state and outcomes in children with type 1 diabetes. Objective We aimed to determine whether mental state at diagnosis predicts longer term cognitive function of children with a new diagnosis of type 1 diabetes. Methods Mental state of 87 patients presenting with newly diagnosed type 1 diabetes was assessed using the School-Years Screening Test for the Evaluation of Mental Status. Cognitive abilities were assessed 1 wk and 6 months postdiagnosis using standardized tests of attention, memory, and intelligence. Results Thirty-seven children (42.5%) had reduced mental state at diagnosis. Children with impaired mental state had poorer attention and memory in the week following diagnosis, and, after controlling for possible confounding factors, significantly lower IQ at 6 months compared to those with unimpaired mental state (p < 0.05). Conclusions Cognition is impaired acutely in a significant number of children presenting with newly diagnosed type 1 diabetes. Mental state screening is an effective method of identifying children at risk of ongoing cognitive difficulties in the days and months following diagnosis. Clinicians may consider mental state screening for all newly diagnosed diabetic children to identify those at risk of cognitive sequelae.

Deprivation is an independent predictor of 1-year mortality in outpatients with chronic obstructive pulmonary disease

Deprivation is linked to increased incidence in a number of chronic diseases but its relationship to chronic obstructive pulmonary disease (COPD) is uncertain despite suggestions that the socioeconomic gradient seen in COPD is as great, if not greater, than any other disease (Prescott and Vestbo).1 There is also a need to take into account the confounding effects of malnutrition which have been shown to be independently linked to increased mortality (Collins et al).2 The current study investigated the influence of social deprivation on 1-year survival rates in COPD outpatients, independently of malnutrition. 424 outpatients with COPD were routinely screened for malnutrition risk using the ‘Malnutrition Universal Screening Tool’; ‘MUST’ (Elia),3 between July and May 2009; 222 males and 202 females; mean age 73 (SD 9.9) years; body mass index 25.8 (SD 6.3) kg/m2. Each individual's deprivation was calculated using the index of multiple deprivation (IMD) which was established according to the geographical location of each patient's address (postcode). IMD includes a number of indicators covering economic, housing and social issues (eg, health, education and employment) into a single deprivation score (Nobel et al).4 The lower the IMD score, the lower an individual's deprivation. The IMD was assigned to each outpatient at the time of screening and related to1-year mortality from the date screened. Outpatients who died within 1-year of screening were significantly more likely to reside within a deprived postcode (IMD 19.7±SD 13.1 vs 15.4±SD 10.7; p=0.023, OR 1.03, 95% CI 1.00 to 1.06) than those that did not die. Deprivation remained a significant independent risk factor for 1-year mortality even when adjusted for malnutrition as well as age, gender and disease severity (binary logistic regression; p=0.008, OR 1.04, 95% CI 1.04 to 1.07). Deprivation was not associated with disease-severity (p=0.906) or body mass index, kg/m2 (p=0.921) using ANOVA. This is the first study to show that deprivation, assessed using IMD, is associated with increased 1-year mortality in outpatients with COPD independently of malnutrition, age and disease severity. Deprivation should be considered in the targeted management of these patients.

Response to comments on "Productivity is a poor predictor of plant species richness"

Pan et al. claim that our results actually support a strong linear positive relationship between productivity and richness, whereas Fridley et al. contend that the data support a strong humped relationship. These responses illustrate how preoccupation with bivariate patterns distracts from a deeper understanding of the multivariate mechanisms that control these important ecosystem properties.