914 reputation
517
bio website
location
age
visits member for 3 years, 8 months
seen yesterday

Feb
18
awarded  Commentator
Feb
18
comment How useful is the genetic algorithm for financial market forecasting?
you select your strategy using past data, then you apply it to future data! That way you are absolutely not sure that an optimised GA will perform! It will work only if there is persistence
Feb
18
answered How useful is the genetic algorithm for financial market forecasting?
Feb
16
answered Digital Signal Processing in Trading
Feb
16
answered Is there a standard method for quantifying mean-reversion for use in directional trading?
Feb
15
comment Correlation between prices or returns?
I would add that one may want to look at either correlation of returns or cointegration of prices rather than correlation on prices
Feb
7
awarded  Beta
Feb
7
comment Mean reverting strategies
Thanks for you input. I have been looking with much interest at your website and found your papers very good. I can see that you sometimes look at mean reversion in spread using futures. How do you roll from one contract to another and don't you think that you may break the cointegration each time you roll?
Feb
3
awarded  Student
Feb
2
comment What are the best master programmes for someone interested in a career in quantitative finance?
Ecole Polytechnique?
Feb
2
awarded  Critic
Feb
2
comment How useful is Markov chain Monte Carlo for quantitative finance?
Thanks DavidShor, great question
Feb
2
comment How useful is Markov chain Monte Carlo for quantitative finance?
Same question as @DavidShor. Also even if true, financial applications of quantitative methods are not only pricing.
Feb
2
comment Mean reverting strategies
This link is quite interesting I think arxiv.org/abs/0808.1710
Feb
2
comment Mean reverting strategies
Orstein-Uhlenbeck is a AR(1) if I am not wrong? I would like a more general method. I have tried to fit a time varying ARMA(p,q), which is I think a more general model than Orstein-Uhlenbeck.
Feb
2
answered How does pair trading work?
Feb
2
comment Mean reverting strategies
I agree it is a good debate. But honestly I think we can share technics without the strategies themselves. It is often a matter of how you use the technics...
Feb
1
awarded  Teacher
Feb
1
awarded  Supporter
Feb
1
comment What type of analysis is appropriate for assessing the performance time-series forecasts?
better answer than mine, richardh. I would add that this model will help you deal with linear relationship (pearson autocorrelation) of your residual. But in theory you can remove any type of relationship between them. The only thing you need is to find a model for your residuals and add it to your original forecasting model to remove this deterministic part of the residuals. basically a good model is obtained when the residuals are really random.