This paper illustrates the effects of item-nonresponse in surveys on the results of multivariate statistical analysis when estimation of productivity is the task. To multiply impute the missing data a data augmentation algorithm based on a normal/Wishart model is applied. Data of the German IAB...

The sentiment of news predicts the short-term stock market performance of individual companies. We find that this association is solely due to the idiosyncratic informational content of an article. We transparently quantify the association between news sentiment and stock market performance of...

Artificial intelligence (AI) can improve the resiliency of both individual financial institutions and the financial sector as a whole. Yet, its users must balance benefits and risks of these technologies. This paper illustrates five focus points which financial institutions should consider when...

We construct an index for economic policy uncertainty in Denmark using articles in Denmark's largest financial newspaper. We adapt a Latent Dirichlet Allocation (LDA) model to sort articles into topics. We combine article-specific topic weights with the occurrence of words describing uncertainty...

We construct novel proxies of physical and transition climate risks by conducting textual analysis of climate-change news over the period 2000-2018. This analysis uncovers four textual variables related to the topics of U.S. climate policy, international summits, natural disasters, and global...

The Romanian vegetable market is in continuous development, with consumers preferring local vegetables to the detriment of imported ones. Even though there is a very good potential and a fairly varied supply of fresh vegetables on the domestic market, some of the producers have given up making...

Regression, matching, control function and instrumental variables methods for recovering the impact of education on individual earnings are reviewed for single treatment and sequential multiple treatments with and without heterogeneous returns. The sensitivity of the estimates once applied to a...

The classical Salant-condition for the comparison of the mean interrupted spell length of the stock of unemployed and the mean completed spell length of the corresponding flow can be substantially weakened: for a NWUE [New Worse than Used in Expectation] distribution the former is greater than...

Using administrative panel data of health insurants, we estimate the effects of low birth weight on health service utilization among children and young adults between birth and 21 years old. To account for time-invariant heterogeneity of mothers, we use sibling fixed- effects estimation. We find...

Applying a probabilistic causal approach, we define a class of time series causal models (TSCM) based on stationary Bayesian networks. A TSCM can be seen as a structural VAR identified by the causal relations among the variables. We classify TSCMs into observationally equivalent classes by...