Possibilities of construction economic-mathematical models of financial market dynamics using local level components

This paper is devoted to presentation of application local level and local level with drift models in order to solve a problem of developing common methods and techniques for creating economic-mathematical models for forecating dynamics of financial market time series. In paper, basing on initially introduced forecasting model that takes into account phenomenon of price deviation from its equilibrium value, as one of the basic technical analysis tool, there were proposed two additional economic-mathematical models, containing local level and local level with drift components. Complicated by the corresponding components, additional models were evaluated by classical Kalman filtering, using data of three exchange rate markets, US dollar against ruble (USDRUB), euro (EURUSD) and Swiss franc (USDCHF) for the year of 2017, and demonstrated higher forecasting capabilities, judging by percentage of correct forecast directions, in comparison with the initial model, which in turn confirms potential possibility of their real economic use, in particular, on binary options market, presented in paper as example.