Model of FX rate volatility, based on fractal features of financial time series (B.Poutko, A.Didenko, M.Dubovikov.)

Paper develops volatility forecasting model of RUR/USD exchange rate. To forecast volatility we decompose it to components, characterizing fractal structure of financial time series. Using regression analysis we confirm quasi-cyclical time structure for one of the fractal parameter. We then discuss capacity of the method to predict volatility, including forecasting market transition to unsteady state.

(Applied Econometrics, vol. 36(4), pages 79-87R code and dataset is available upon request: alexander.didenko (at) or from my profile).

Continue reading

When do financial analysts make mistakes?

While helping one of my students, Oleg Karapaev, in his struggle with a paper on determinants of divergence between “fair” and observed prices in stocks (written as a part of his bachelor research project), I’ve made some observations perhaps worth sharing with blogosphere. I used some R code and ggplot2ggthemes by Jeffrey Arnolds to represent findings visually (important: to replicate code in full you should have an access to Bloomberg and Rbbg package to download data).

Historical vs. Predicted price for Microsoft shares

Continue reading

Using constant volume scale for modeling fractal characteristics of financial time series (A.Didenko, M.Dubovkiov, B.Poutko)

We propose to use intrinsic time scale based on volume when measuring fractal dimension of financial time series. (Dubovikov, 2004) introduces a new method of measuring fractal dimension which is superior to other methods, including well-known Hurst index in terms of  speed of asymptotic. As a downside, estimates obtained with new method, are noisy and hard to predict, which in turn complicates its use in practice. We demonstrate that sampling time-series across volume scale, instead traditional physical time scale, could significantly improve predictability of fractal dimension.

(Published in Russian in “Science in Modern Information Society IV” , Fall 2014, ISBN 978-1-50232-179-4)

Algotrading in Bloomberg, R, and beyond (Financial University, Spring-Fall 2014). 

Class description: introductory course on developing trading algorithms and algotrading industry as a whole. Class is held in financial lab with 9 Bloomberg terminals; students are learning to fast-prototype algorithms using R language and real data from markets. Focus is on practice, but good understanding of underlying theory is a must. Knowledge of Bloomberg platform and R language is an advantage.

Prerequisites: basics of financial markets, technical and fundamental analysis, financial mathematics, modern portfolio theory, statistics and probability theory.

Continue reading