Embedding R in Windows applications, and executing R remotely
Embedding R in Windows applications, and executing R remotely
Embedding R in Windows applications, and executing R remotely
Create successful ePaper yourself
Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.
Signal noise decomposition of f<strong>in</strong>ancial data: An<br />
<strong>in</strong>frequent trad<strong>in</strong>g analysis<br />
Helgi Tomasson<br />
The observed transaction prices on a stock market at discrete time po<strong>in</strong>ts are<br />
assumed to be a sample from a cont<strong>in</strong>uous time-value process. The theory of an<br />
efficient market is used as motivation for a r<strong>and</strong>om-walk type model. The fact<br />
that bid-ask spread <strong>and</strong> other microstructure phenomena exist is accounted for<br />
by add<strong>in</strong>g a noise term to the model. Models for elementary detrend<strong>in</strong>g based<br />
on stochastic processes <strong>in</strong> cont<strong>in</strong>uous time are set up. For empirical analysis,<br />
they are formulated <strong>in</strong> state-space form, <strong>and</strong> calculations are performed with<br />
the Kalman-filter recursions. The result is an analytical way of decompos<strong>in</strong>g<br />
the observed transaction price change <strong>in</strong>to a value <strong>in</strong>novation <strong>and</strong> a market<br />
noise component. The respective <strong>in</strong>novation st<strong>and</strong>ard deviations <strong>and</strong> market<br />
noise st<strong>and</strong>ard deviations are easily <strong>in</strong>terpreted. Some alternative stochastic<br />
structures are considered <strong>and</strong> applied to data from the Icel<strong>and</strong> Stock Exchange.<br />
Algorithm is implemented <strong>in</strong> the statistical language R comb<strong>in</strong>ed with Fortran<br />
subrout<strong>in</strong>es.<br />
Keywords: Diffusion processes, state-space models, f<strong>in</strong>ancial data, <strong>in</strong>frequent<br />
trad<strong>in</strong>g