forecasting methodology?

[tobias_lietz@westlb.de]

hi, all

hope that someone can help me:

what i want to do is to provide robust forecasts for generating scenarios
to measure market risk beyond periods of two years (e.g. 10 years).
what i have are time series on a monthly basis. unfortunately the history
of some time series is shorter than 10 years.

has someone experience how to do this best? maybe garch/arch/arima........?
or has someone papers/files/......which could help me?

best regards and many thanks

tobias


mailto:tobias_lietz@westlb.de