This paper focuses on the performance of Lithuanian life-cycle second-pillar pension funds. Every such fund first specifies its benchmark and then attempts to follow the benchmark in some way. This is a form of regulation, meaning that every such fund is somehow regulated and controlled by the central bank authorities. The goal of this paper is twofold: (i) to analyse the returns of the pension funds with respect to their benchmarks and (ii) to determine whether less strict regulation leads to a better outperformance of the fund with respect to the benchmark. In order to achieve this, we introduced a new performance measure called the dominance-tracking index, which combines the ideas of almost stochastic dominance relations and tracking errors. While the tracking error and its modifications measure the strength of the regulation, almost stochastic dominance provides information about preferences between the funds and their benchmarks. Therefore, the new index was constructed in such a way as to take into account both approaches. The empirical section of the study then presents the results separately for the considered pension managers and participants’ age groups as usual in the life-cycle pension funds analysis. Finally, by taking into account various periods, we studied the effects of the COVID-19 crisis.

This paper presents a stress testing technique based on a hidden Markov regime switching model and scenario generations. Firstly, we assume that investor’s risk preferences and beliefs are primarily observed in main world indices and then transmitted to other markets such as IInd pillar life-cycle pension funds. This enables both regime (market state) identification and regime-switching detection. Second, we apply stress to the transition matrix by increasing probabilities of moving to a worse market state, i.e., we increase the transition probabilities on the upper-right side of the main diagonal and decrease the diagonal probabilities correspondingly. Third, future scenarios (evolution) of returns of pension funds are generated using stressed transition matrices and basic risk and performance measures of the final wealth are presented. Finally, three different strategies for pension managers are considered and compared to each other. The results show how the best strategy depends on the stress level.

This study comprehensively evaluates the ability to borrow machine learning modeling techniques for SMEs, utilizing a sample of the Baltic States with many variables. The study aims to assess the applicability of access to credit modeling techniques for SMEs. This is the first study in which a large–scale assessment has been carried out in the Baltic States sample, covering five years of credit applications from SMEs to a depository institution. The results showed that Gradient Boosting produces the most accurate results. Gradient Boosting demonstrated better results than the benchmark Logistic Regression as well as other advanced machine learning models, including Random Forests and Multilayer Perceptron models. The method showed the highest accuracy of the overall receiver operating characteristic (ROC) curve and the associated area under the curve (AUC) (ROC–AUC) and Average Precision values, as well as other discriminatory threshold values, compared to alternative methods.