Existing methods of technical analysis is the empirical essence of speculation. Their effectiveness is only confirmed with the methods of mathematical statistics. In fact, technical analysis methods can improve the probability of the conclusion of profitable trades. Thus, there is a need to create a handler function of price changes to filter out the random component and the deterministic component separation in order to isolate the point of a trend change and optimizing the time point of opening and closing positions. As a solution to this problem a stochastic handler function has been developed that determines the change in the price of an asset with the method of the Markov theory of nonlinear filtering. Evaluating the effectiveness of the synthesized receiver was based on numerical modeling of processing the mixture of the deterministic signal and the white Gaussian noise. It is shown that the designed processor allows for filtering out the random component of the price of liquid shares traded on MICEX, with the aim of developing a profitable trader strategy. The resulting graph of Markov random process evaluation λ (t) at the output of the filter periodically intersects the graph of the original random process. In this case, the intersection of graphs only occurs if the change of direction (trend) estimates λ (t). Thus, in the ideal case, assuming that the transaction can be performed precisely at the time of intersection of the graphs, each transaction is at least break even. The article poses the problem of closing the gap between real conditions of the transactions and ideal ones. To do this, one must solve the problem of the fractal properties of markets and limit the possibilities of using high-frequency trading robots and algorithms.
|Translated title of the contribution||FILTRATION OF RANDOM COMPONENT ASSET PRICE BY METHOD OF MARKOV FILTRATION|
|Number of pages||17|
|Journal||Вестник УрФУ. Серия: Экономика и управление|
|Publication status||Published - 2015|
Level of Research Output
- VAK List