I introduced myself and what I’m trying to do a little bit in my first post here:
As I work my way through the OpenGamma source code and, aside from being super impressed, I’m starting to appreciate how I can use it for my business. I was shy about asking too many questions here, so I sent a note via your web enquiry form. I’m happy to pay someone for some dedicated time if necessary (I know you guys are busy!), but Soila recommended I start here first so here it goes.
Insurance companies, particularly in Asia, are not too concerned about real-time risk analytics or even pre-trade risk analytics (yet!), at least for ALM purposes. There is still quite some lag, so I don’t see myself using the trading functionality or live data feeds for risk analytics initially, but I certainly like having that functionality in the background waiting to be exploited (which, on second thought, would come in handy for my client’s OTC trades - my mind has been maybe too focused on ALM).
So what I’m currently trying to do is extract enough of the OpenGamma functionality to, independently of any server and UI:
- Define some assets
- I think these should extend ManageableSecurity
- Create a SecurityMaster
- I’m not sure which kind to use. I’m looking at InMemorySecurityMaster and DataSecurityMasterResource.
- If I understand, with the latter, I can define securities in CSV files in my WEBContent/resources folder.
- Build up some positions in these assets using ManageablePosition
- Create a PositionMaster.
- Similar to above, I’m not sure what kind of master to use.
- Put the positions into a portfolio using ManageablePortfolio
- Create a PortfolioMaster
If I got this far, which I’m not even close yet, I would already consider that a huge victory. What have I left out?
Assuming I am able to construct a portfolio master with a portfolio I’m happy with, what else would I then need in order to run it through some cycles for Monte Carlo, etc, using your analytics tools?
Could someone help outline the minimal components I would need in order to, for example, build a bond portfolio with several thousand issuers and feed it a few tens of thousands of yield curves and look at the P&L distribution using OpenGamma?
On a personal note, I am quickly becoming a big fan of OpenGamma and hope I can contribute back in some ways. I see my use of OpenGamma as selling point and will proudly display (with permission beforehand) “Powered by OpenGamma”, or something, on any apps I build that use it.
Two areas I may be able to help include:
- Quantitative Research
- Asset/liability modelling
- Risk and performance analytics
- Market modelling/economic scenario generation
- User experience and data visualisation
The areas I have made most progress so far (although still very early) are in the user interface and data visualisation. I’m developing a re-usable data visualisation library geared for risk analytics based on d3.js, which you might find interesting.
Cheers and thanks for any help!
Eric Forgy, Director
Coherent Capital Advisors, Limited