Trading vs. Quant Research

  • Thread starter Thread starter Zeuge
  • Start date Start date
Oh wow that reading list is so comprehensive. I really liked Shreve in general. I'm about halfway through and I feel so good when I read it because I actually understand line by line and that's how the concepts stay in my head.

Andy Nguyen: I've heard of Boost and Quantlib, actually I already have Boost installed and I know Quantlib runs with Boost. But what exactly can one do in efforts to get practical finance coding experience (ie. coding a strategy for pairs trading) without a platform? (Like, for example, what exactly could I make the code do without real-time info? I *could* get real-time info if I had a platform, but I would need one which works with Boost and Quantlib then, so that it would take my C++ code as input. Do you see what I'm trying to ask here?)
 
I have huge work exp in programming(Java) and know C++. As both are very similar(in main conecpts of OOPs and polymorphism) and are much different(in memory management-garbage collection etc), I think I should be fine with C++ within a few weeks.

My main query is that whether there are any programming-finance challenges/competitions etc where you can showcase your skills and can get involved with "real" application of programming in finance. That is something very useful for any MFE guy.

Before I graduate from current MFE program, I want to develop "great" skills in practical applications.
 
Oh wow that reading list is so comprehensive. I really liked Shreve in general. I'm about halfway through and I feel so good when I read it because I actually understand line by line and that's how the concepts stay in my head.

Andy Nguyen: I've heard of Boost and Quantlib, actually I already have Boost installed and I know Quantlib runs with Boost. But what exactly can one do in efforts to get practical finance coding experience (ie. coding a strategy for pairs trading) without a platform? (Like, for example, what exactly could I make the code do without real-time info? I *could* get real-time info if I had a platform, but I would need one which works with Boost and Quantlib then, so that it would take my C++ code as input. Do you see what I'm trying to ask here?)

You don't need real-time info , get a database of all the stocks/bonds/etc you want.
Lets say over 2011, and run your programs using that data, your algorithm doesn't know if its running on real-time or on historical data.
If your algorithm is able to predict the next historical datapoint than there is a good chance it will do so in real-time too.
 
Back
Top