![]() |
ИСТИНА |
Войти в систему Регистрация |
ИПМех РАН |
||
We describe a typical structure of high frequency (low latency) market data. In order to analyze such data one needs to see them through a “frog’s eye view”, i.e. one single market event at a time. Using mathematical theory that goes back to B.B. Mandelbrot, P.K. Clark, V.Yu. Korolev and others we develop an approach based on the “principle of MA”. As an illustration we use the approach to measure real time market risk with modified Value at Risk that is sensitive not only to changes in standard deviation of potential loss, but also to changes in frequency of trading events