Assessing the discriminatory power of loss given default models

For banks using the Advanced Internal Ratings-Based Approach in accordance with Basel III requirements, the amount of required regulatory capital relies on the banks'' estimates of the probability of default, the loss given default and the conversion factor for their credit risk portfolio. Therefore, for both model development and validation, assessing the models'' predictive and discriminatory abilities is of key importance in order to ensure an adequate quantification of risk. This paper compares different measures of discriminatory power suitable for multi-class target variables such as in loss given default (LGD) models, which are currently used among banks and supervisory authorities. This analysis highlights the disadvantages of using measures that solely rely on pairwise comparisons when applied in a multi-class setting. Thus, for multi-class classification problems, we suggest using a generalisation of the well-known area under the receiver operating characteristic (ROC) curve known as the volume under the ROC surface (VUS). Furthermore, we present the R-package VUROCS, which allows for a time-efficient computation of the VUS as well as associated (co)variance estimates and illustrate its usage based on real-world loss data and validation principles.     

Shrinkage Estimators for Prediction Out-of-Sample: Conditional Performance

We find that, in a linear model, the James–Stein estimator, which dominates the maximum-likelihood estimator in terms of its in-sample prediction error, can perform poorly compared to the maximum-likelihood estimator in out-of-sample prediction. We give a detailed analysis of this phenomenon and discuss its implications. When evaluating the predictive performance of estimators, we treat the regressor matrix in the training data as fixed, i.e., we condition on the design variables. Our findings contrast those obtained by Baranchik (1973 Baranchik , A. J. ( 1973 ). Inadmissibility of maximum likelihood estimators in some multiple regression problems with three or more independent variables . Ann. Statist. 1 ( 2 ): 312 – 321 .[Crossref], [Web of Science ®] , [Google Scholar]) and, more recently, by Dicker (2012 Dicker , L. ( 2012 ). Dense signals, linear estimators, and out-of-sample prediction for high-dimensional linear models. arXiv:1102.2952 [math.ST].  [Google Scholar]) in an unconditional performance evaluation.     

Life table technique for the working ages

Some results of efforts to design an integrated formula system for age-specific death rates, survivals and expectation of life are presented. The report deals only with the active ages 15–80. The system is based on the assumption that the age-specific central mortality rate (m _x ) can be expressed satisfactorily by means of a polynomial m = a ₀ + a ₁ x + a ₂ x ²+ ….     

Crime,deterrence and unemployment in Greece: A panel data approach

This study empirically examines the relationship among crime, deterrence and unemployment in Greece. A regional dataset over the period 1991–1998 was collected and analysed. Our econometric methodology follows the Generalized Method of Moments (GMM) estimator applied to dynamic models of panel data. The results show that property crimes are significantly deterred by higher clear-up rates and that unemployment increases crime. For violent crimes, however, the effect of the clear-up rate and unemployment are found to be generally insignificant. Finally, our results may provide support to policy makers in forecasting criminal activity in the current economic downturn under a wave of harsh austerity measures, budget cuts and increased unemployment.     

Exact combinatorial algorithms and experiments for?finding maximum k-plexes

We propose new practical algorithms to find maximum-cardinality k-plexes in graphs. A k-plex denotes a vertex subset in a graph inducing a subgraph where every vertex has edges to all but at most k vertices in the k-plex. Cliques are 1-plexes. In analogy to the special case of finding maximum-cardinality cliques, finding maximum-cardinality k-plexes is NP-hard. Complementing previous work, we develop exact combinatorial algorithms, which are strongly based on methods from parameterized algorithmics. The experiments with our freely available implementation indicate the competitiveness of our approach, for many real-world graphs outperforming the previously used methods.     

The mediating effects of leader integrity with transparency in communication and work engagement/performance

We examined in a 3-month longitudinal study how leader behavioral integrity relates to individual follower work engagement, and how that relationship, in turn, connects to performance. We hypothesized that ratings of leader behavioral integrity would mediate the relationship between leader transparent communication and follower work engagement, which would also have a positive relationship with performance. We tested our hypotheses using data collected from military cadets (n = 451), who each rated their respective leader. Our findings show that followers who rated their leaders as exhibiting more transparent communication at Time 1, also rated themselves as more engaged in their work role at Time 2 (3 weeks later), and that their perceptions of leader behavioral integrity mediated that relationship. Follower engagement also positively related to third-party ratings of follower performance at Time 3 lagged 3 months. We discuss the implications of these findings for research on leader integrity, authenticity, follower engagement, and performance.     

Objective Bayesian analysis of spatial data with uncertain nugget and range parameters

The authors develop default priors for the Gaussian random field model that includes a nugget parameter accounting for the effects of microscale variations and measurement errors. They present the independence Jeffreys prior, the Jeffreys‐rule prior and a reference prior and study posterior propriety of these and related priors. They show that the uniform prior for the correlation parameters yields an improper posterior. In case of known regression and variance parameters, they derive the Jeffreys prior for the correlation parameters. They prove posterior propriety and obtain that the predictive distributions at ungauged locations have finite variance. Moreover, they show that the proposed priors have good frequentist properties, except for those based on the marginal Jeffreys‐rule prior for the correlation parameters, and illustrate their approach by analyzing a dataset of zinc concentrations along the river Meuse. The Canadian Journal of Statistics 40: 304–327; 2012 © 2012 Statistical Society of Canada