[R-sig-finance] stochastic dominance problem
spencer.graves at pdf.com
Fri Feb 10 02:50:07 CET 2006
I just got one hit from RSiteSearch("stochastic dominance"), but that
discussed some other kind of dominance.
If a function to check for that exists in R, it exists under some
other name. However, checking the definition (e.g.
"http://cepa.newschool.edu/het/essays/uncert/increase.htm"), I think
"stochastic dominance" is so specific it is likely to be of more
theoretical than practical interest. The conditions should be easy
enough to check for theoretical distributions. However, in real
portfolio management, I doubt if stochastic dominance is of much
practical value, because interest would primarily focus on a list of
alternatives among which there exists no clear dominance for all
possible utility functions. Instead, people are most concerned with
finding strategies that seem to be better for a given level of risk.
Are you familiar with Rmetrics (www.rmetrics.org)? The "fPortfolio"
and other packages associated with Rmetrics might interest you.
Rmetrics is breathtaking in scope but difficult to get into. If you are
new to R but are intereseted in analyzing financial time series, I
suggest you start with ch. 14 of Venables and Ripley (2002) Modern
Applied Statistics with S, 4th ed. (Sprinter). Next I recommend you
work the vignettes in the zoo package. From there, you could try many
things, including the "dse" bundle and Rmetrics. For a good reference
on statistics in finance, I recommend Ruey Tsay (2005) Analysis of
Financial Time Series, 2nd. ed. (Wiley). It has only a little R, but
the contents are, I think, quite important.
hope this helps.
hendry raharjo wrote:
> i have a problem quite similar to portfolio selection problem.
> i am trying to do a stochastic dominance test to rank these options using R,
> the criteria is 'the-higher-the-better' since the random variable is 'profit':
> Alt A ~ Normal (mean=10, stDev=2)
> Alt B ~ Normal (mean=8,stDev=4)
> Alt C ~ Uniform (a=8,b=12)
> the question is to order these alternatives according to the stochastic
dominance rules, say Second-order Stochastic Dominance and Third-order
Dominance. Has anyone ever had this kind of problem and solved it using R?
> i am new in R as well as in finance field, i really appreciate your
suggestion/ help with this problem.
> Thank you,
> (i tried to ask to r-help list people, but it seems that i have an
email sending problem since it fails)
> [[alternative HTML version deleted]]
> R-sig-finance at stat.math.ethz.ch mailing list
More information about the R-sig-finance