Publikace na Matematicko-fyzikální fakulta |
2012
Abstrakt
In the contribution financial data (mainly profits/losses, P&L) are analyzed using both new built-in statistical functions and own procedures. Selected parametric P&L distributions that seem to be suitable for modelling but are not easily analytically tractable are investigated.
It is shown that in many cases the mixture of three symmetric distributions is sufficiently flexible to fit the data successfully. For the estimation purposes the built-in EstimatedDistribution function is used.
The statistical tests include tests of independence and tests for symmetry, among others.