Artykuł przedstawia możliwość adaptacji jednej z metod wtórnego próbkowania - powtarzanej K-składowej walidacji krzyżowej, w odniesieniu do oceny efektywności inwestycyjnej strategii portfelowych. Metoda ta pozwala na wnioskowanie statystyczne w sytuacjach, gdy nie jest znana postać analityczna rozkładu badanej cechy lub postać ta nie pozwala na skuteczne zastosowanie dostępnych metod parametrycznych. Badanie empiryczne, które przeprowadzono w celu potwierdzenia stawianej hipotezy nie wskazuje na fakt, iż należałoby jej zaprzeczyć. Zostało ono wykonane na dużym zbiorze danych historycznych pochodzących z amerykańskiego rynku akcji. Dodatkowo w artykule zaprezentowano również sposoby wyznaczania błędów prognozy mierników efektywności inwestycyjnej podczas stosowania metod wtórnego próbkowania.
EN
This paper presents one of possible adaptations of resampling method - repeated K-fold cross-validation - used for portfolio optimization strategies investment efficiency estimation. This method is very helpful with statistical inference determination in situation when analytical form of distribution is not known or analytical form makes using available parametric methods inefficient. Empirical study was conducted to confirm that aforementioned assumption is true and its results show that there is no support for assertion of its falseness. It was performed on large historical data set from American stock market. Moreover, this article presents how to determine prediction errors for investment efficiency measures applied in resampling methods.
Main goal of this paper is to analyse the influence of transaction costs on efficiency evaluation and parameters selection regarding technical analysis investment strategies. The first part of the article gives a brief description on rules of constructing automated investment strategies using technical analysis tools. Also, two efficiency measures that reflect several different features of distribution where presented: Omega and UPR (upside potential ratio). Moreover, a statistical method for testing whether a series of returns obtained in historical verification process comes from the same distribution as the series obtained during optimization process was introduced. Kolmogorov- -Smirnov statistical test is used for that purpose. Next, an investment strategy proposed for analysis was presented. A technical indicator ROC (rate of change) was used to generate signal for entering and exiting market positions. Afterwards, an empirical research was conducted on capital instruments of different markets: stock, currency and commodity. The results of empirical research do not result in clear conclusions. Most of the analysed cases show that considering transaction costs at optimization level results in different strategy parameters selection than not doing so. However, efficiency of such strategies in many cases is not satisfactory despite statistical verification regarding the same distribution of given returns and returns of the most effective strategies in optimization process.
This paper presents the Omega meter - quite new tool to measure investment efficiency and risk. One of its major advantages is the independence from the assumptions made regarding the distribution of investment returns being analysed. It also allows for comparison of investment efficiency and riskiness of different assets. The article focuses on the essential characteristics of the meter and the presents methods for its point estimation. An empirical comparison with other risk adjusted measures of investment efficiency was made. The comparison was performed on investment funds of different categories. Omega meter in some cases allows to obtain more valuable information than the other regarded measures of investment efficiency. However, empirical research shows that efficiency measurement results are not much different when using Omega and Sortino ratio.
JavaScript jest wyłączony w Twojej przeglądarce internetowej. Włącz go, a następnie odśwież stronę, aby móc w pełni z niej korzystać.