SERVING THE QUANTITATIVE FINANCE COMMUNITY

Search found 264 matches

  • 1
  • 2
  • 3
  • 4
  • 5
  • 18
by ScilabGuru
February 9th, 2004, 5:24 pm
Forum: Numerical Methods Forum
Topic: Testing Monte Carlo Results
Replies: 4
Views: 190161

Testing Monte Carlo Results

<t>I am sorry, there is no any problem of unit roots when you generate lognormal model. Unit root problem arises in more advanced econometric time seris when return depends on the previos returns. It comes from disrete linear system theoyr whe is staded that the stable (not going to infinity) linear...
by ScilabGuru
February 9th, 2004, 4:41 pm
Forum: Technical Forum
Topic: Overfitting Data
Replies: 12
Views: 190324

Overfitting Data

<t>Well, of course, if your parameter is 0.000000....1, or simply speaking the data is not sensitive to your parameter than nothing can help you.But if you are going to work with the same type of the data in future - then this just means the in the future your data will not be sensitive pretty much ...
by ScilabGuru
February 9th, 2004, 3:13 pm
Forum: Technical Forum
Topic: Overfitting Data
Replies: 12
Views: 190324

Overfitting Data

<t>QuoteOriginally posted by: scholarQuoteOriginally posted by: rezasometimes you have 4 parameters and 5000 data points and you still cannot get a decent estimate !Or even 1 parameter and 25 years of daily data: just try to estimate a mean reversion rate from IR time series...Well, this just means ...
by ScilabGuru
February 9th, 2004, 2:38 pm
Forum: Technical Forum
Topic: Transaction Costs in Portfolio Optimization Problems
Replies: 6
Views: 189870

Transaction Costs in Portfolio Optimization Problems

<t>The question what you mostly ineterested are in. Bidask spread and other commision are modeled differently. If you trade relatively frequently - bidask spread becomes a dominant, otherwise other factors are more essential Cvitanic has a set of paper devoted to "friction market". He also set quite...
by ScilabGuru
February 4th, 2004, 7:29 pm
Forum: Student Forum
Topic: Regression linest() function returns error.
Replies: 8
Views: 191489

Regression linest() function returns error.

<t>QuoteOriginally posted by: dronaTrevor,yes, I do have an instances or probably more than one, where qualitative factor(s) all zeros.I donot get any problems when I run in matlab though !regardsHow many equations you have? BTW it is a bit pathalogical to work with categorical type data (1/0) in st...
by ScilabGuru
February 3rd, 2004, 5:50 pm
Forum: Careers Forum
Topic: Who needs H1 the first time can relax until the next fiscal year
Replies: 22
Views: 191050

Who needs H1 the first time can relax until the next fiscal year

To what I understand for both. Cry, students... It is not applied only for one who transfers H1 from 1 company to another...
by ScilabGuru
February 2nd, 2004, 9:46 pm
Forum: Technical Forum
Topic: overfitted technical trading rule
Replies: 6
Views: 189391

overfitted technical trading rule

<t>QuoteOriginally posted by: mtleitaoCan anyone suggest ways of detecting if a technical trading rule suffers from overfitting?ThanksIf it too complicating than almost surely it is overfitted...No simple rules, only testing and testing..., but don't forget that testing itslef introduce additional i...
by ScilabGuru
February 2nd, 2004, 9:32 pm
Forum: Technical Forum
Topic: Overfitting Data
Replies: 12
Views: 190324

Overfitting Data

<t>There are a lot of techniques how to balance the ability of a model to fit the data in sample and complexity, namely inability to fit it well out of sample. The Occam rasor principle claims that between two models explaining similary the data, the simple one is preferable. Quantitatively, there a...
by ScilabGuru
February 2nd, 2004, 9:17 pm
Forum: Careers Forum
Topic: Who needs H1 the first time can relax until the next fiscal year
Replies: 22
Views: 191050

Who needs H1 the first time can relax until the next fiscal year

65000 H1-quota on this financial year almost surely is ended in February... Any ideas, how is to solve the problem ?
by ScilabGuru
January 29th, 2004, 11:29 pm
Forum: Student Forum
Topic: Stationarity- ARIMA
Replies: 5
Views: 189865

Stationarity- ARIMA

log returns have couple of proprties making them popular:1. They are close to relative returns since log(x_2/x_1)=log(1+x_2-x_1/x_1))~ x_2-x_1/x_12. They are additive like absolute returns, namely log(x_k/x_1)=log(x_k/x_n) +log(x_n/x_1)
by ScilabGuru
January 29th, 2004, 6:50 pm
Forum: Student Forum
Topic: Monte carlo simulation variance
Replies: 8
Views: 189519

Monte carlo simulation variance

<t>It was mentioned already but let me explain differently. You have N instance of a random variable (MC simulation) from the same distribution. It variance indeed convereges to constant to the true value of the variance of this distribution. But now you take sample mean:sum Y_i /n . Theoretical var...
by ScilabGuru
January 29th, 2004, 3:12 am
Forum: Technical Forum
Topic: How to combine different conditional return distributions?
Replies: 9
Views: 189434

How to combine different conditional return distributions?

<t>One more way to assume that you return model is the composition of few models. In this case you don't have to make sum of distribution but you have to specify n probabilties for switching of models. Namely, your return will be with probability P_1 from distribution R_1with probabilit P_2 from dis...
by ScilabGuru
January 23rd, 2004, 8:47 pm
Forum: Off Topic
Topic: Bush's new immigration plan?
Replies: 56
Views: 191618

Bush's new immigration plan?

Quotelike most women Hmm
by ScilabGuru
January 23rd, 2004, 5:51 pm
Forum: Off Topic
Topic: Bush's new immigration plan?
Replies: 56
Views: 191618

Bush's new immigration plan?

<t>QuoteOriginally posted by: Energetic<brNo I'm not because earlier you volunteered to answer my question:QuoteI infer that you regard the discrimination by skills and abilities as legitimate. (I do too, no beef here.) May I ask you why?You never gave me a satisfactory answer but since you entered ...
by ScilabGuru
January 23rd, 2004, 5:27 pm
Forum: Off Topic
Topic: Bush's new immigration plan?
Replies: 56
Views: 191618

Bush's new immigration plan?

I am only the first half of (a) And you?
  • 1
  • 2
  • 3
  • 4
  • 5
  • 18
GZIP: On