On the Inefficiency of Bang-Bang and Stop-Loss Portfolio Strategies

We show in this article that bang-bang portfolio strategies where the investor is alternatively 100% in equity and 100% in cash are dynamically inefficient. Our proof of this result is based on a simple second-order stochastic dominance (SSD) argument. It implies that this is true for any decision criterion that satisfies SSD, not necessarily expected utility. We also examine the stop-loss strategy in which the investor is 100 percent in equity as long as the value of the portfolio exceeds a lower limit where the investor switches to 100 percent in cash. Again, we show that this strategy is inefficient under second-order risk aversion. However, a slight modification of it–in which all wealth exceeding a minimum reserve is invested in equity–is shown to be an efficient dynamic portfolio strategy. This strategy is optimal for investors with a nondifferentiable utility function.     

Residual-based specification of the random-effects distribution for cluster data

We propose a method for specifying the distribution of random effects included in a model for cluster data. The class of models we consider includes mixed models and frailty models whose random effects and explanatory variables are constant within clusters. The method is based on cluster residuals obtained by assuming that the random effects are equal between clusters. We exhibit an asymptotic relationship between the cluster residuals and variations of the random effects as the number of observations increases and the variance of the random effects decreases. The asymptotic relationship is used to specify the random-effects distribution. The method is applied to a frailty model and a model used to describe the spread of plant diseases.     

Friendship and Delinquency: Selection and Influence Processes in Early Adolescence

Positive association of relevant characteristics is a widespread pattern among adolescent friends. A positive association may be caused by the selection of similar others as friends and by the deselection of dissimilar ones, but also by influence processes where friends adjust their behavior to each other. Social control theory argues that adolescents select each other as friends based on delinquency. Differential association theory, on the other hand, argues that adolescent friends influence each other's delinquency levels. We employ new statistical methods for assessing the empirical evidence for either process while controlling for the other process. These methods are based on ‘actor‐oriented’ stochastic simulation models. We analyze longitudinal data on friendship networks and delinquent behavior collected in four waves of 544 students in 21 first‐grade classrooms of Dutch secondary schools. Results indicate that adolescents select others as friends who have a similar level of delinquency compared with their own level. Estimates of the social influence parameters are not significant. The results are consistent with social control theory but provide no support for differential association theory.     

A distance measure for choice functions

This paper discusses and characterizes a distance function on the set of quasi choice functions. The derived distance function is in the spirit of the widely used Kemeny metric on binary relations but extends Kemeny’s use of the symmetric difference distance to set functions and hence to a more general model of choice.