Community Forum : General Discussion Forum
Welcome Guest   
 
 Subject : SD approaches in quantitative finance.. 03/11/2019 09:35:42 AM 
David Robottom
Posts: 5
Location
I feel that a software package like Stella/iThink would make a very useful backtesting environment for quant finance strategies.
Before I re-invent the wheel, is anyone aware of any previous work in this area please?
Best regards
David Robottom
 Subject : Re:SD approaches in quantitative finance.. 03/12/2019 09:28:24 AM 
Jean-Jacques Lauble
Posts: 16
Location
What kind of finance? And why stella/I think sould be better than any other package? i think that today all SD packages have become good with the time, and the difference does not come from the package but from the people using them.
 Subject : Re:SD approaches in quantitative finance.. 03/12/2019 10:48:19 PM 
David Robottom
Posts: 5
Location
Thanks Jean-Jacques,
Like you, I am more interested in how SD can improve my thinking in this area. I use iThink simply because it is the package I have available. What I would really like to understand is whether others have used SD in the quantitative financial area, for example for backtesting individual investment strategies, or for comparative analysis of different alternatives, and whether there is anything already published in this area.
Best regards
David
 Subject : Re:SD approaches in quantitative finance.. 03/13/2019 02:37:49 AM 
Jean-Jacques Lauble
Posts: 16
Location
Hi David

To carry on this discussion, I must ask you a question. You are interested in finance as a researcher or will you apply your thinking to real problems with the hope to improve their solutions?
 Subject : Re:SD approaches in quantitative finance.. 03/14/2019 06:48:45 AM 
David Robottom
Posts: 5
Location
Very definitely real problems. Namely, optimisation of real-world trading algorithms.
Best regards
David
 Subject : Re:SD approaches in quantitative finance.. 03/14/2019 01:46:47 PM 
Jean-Jacques Lauble
Posts: 16
Location
But are you a consultant, working for a trading company, or working for your own trading?
 Subject : Re:SD approaches in quantitative finance.. 03/15/2019 06:26:32 AM 
David Robottom
Posts: 5
Location
Hi again,
This is just for my own trading and to see if this area can benefit from an SD approach. If it works and there is no information "out there" of work already done, I plan to circulate a note about the results.
Best regards
David
 Subject : Re:SD approaches in quantitative finance.. 03/15/2019 08:39:42 AM 
Jean-Jacques Lauble
Posts: 16
Location
Hi David

Did you already study that problem with other methods? There are plenty of studies about it, using all kinds of methods. What is your level of instruction in mathematics, programming, general simulation, quantitative sciences?

Best

JJ
 Subject : Re:SD approaches in quantitative finance.. 03/16/2019 10:17:37 PM 
Muhammad Mutasim
Posts: 1
Location
Hi everydody. Im a new user of sd.im working on my research paper using ithink 9.0. csn any body tell me how can i use delay variable in my model??also how can i get the latest complete version of ithink? Thanks
 Subject : Re:SD approaches in quantitative finance.. 03/18/2019 10:59:16 AM 
Mr. Leonard Malczynski
Posts: 34
Location
There are many free materials on using delays in system dynamics. Look for the MIT Road Maps and the Guided Study program.

In addition many textbooks have sections on delays. I believe that John Sterman's Business Dynamics and the series of modeling texts from Springer on system dynamics modeling, e.g. Dynamic Modeling by Hannon and Ruth have DVDs with models, including those using delays with Stella/iThink. If you want Studio or Vensim examples just ask.

Specific to iThink, the best place to go is to the isee systems website.
 Subject : Re:SD approaches in quantitative finance.. 03/18/2019 02:43:34 PM 
Tom Fiddaman
Posts: 1
Location
@Muhammad - You should post this unrelated question in a separate thread.

@David - I'd suggest starting with a search of conference papers, e.g. google:
site:systemdynamics.org inurl:conferences finance trading

Ventana has done some work on aspects of trading strategy using both aggregate and agent models, though I can't talk about the details. It works.

However, I think your purpose is particularly challenging. Normally, one can construct a model that generates synthetic data from first principles, and then use that to test estimation methods and strategies. But that doesn't work so well in short term finance, because it's hard to construct a convincing model, for a variety of reasons. OTOH using data is problematic as well, because it's hard to distinguish the structure from the noise that won't recur. So, it does seem like there would be a niche for a simulation that enforces some reality checks, complementing the data that might capture aspects of correlation structure etc. that aren't in the model.
 Subject : Re:SD approaches in quantitative finance.. 04/16/2019 06:31:49 AM 
David Robottom
Posts: 5
Location
Thank you for this comprehensive response, Tom. I have just returned from travelling back to the UK, so please accept my apologies for a late acknowledgement.
There is a lot of material in your reply, which I will now go away and digest...
Best regards
David
 
# of Topics per Page