Help understanding

[jgills]

Hello everyone. I am a new member to Quantnet, but I've been browsing the forums for a few weeks now. I am a senior at a University in the United States majoring in finance and I recently interned at Merrill Lynch in New York. I have a vast understanding of the concepts in finance and capital markets, but the math behind everything does not seem to be part of the curriculum. Its more of how to use excel to do things for you or to use a financial calculator to calculate things. Please do not discount my intellect if you find the questions I ask elementary. I ask because I want to be informed, not ridiculed.

To start, I've become very interested in MFE programs after various books I've read. The main influences being from Benoit Mandelbrot and Nassim Taleb. As I mentioned, I don't want to be seen as naive, but I have a few questions that I need clarified. If you think there is a book I can read that will answer my questions instead of having you answer them all, please let me know.

I understand that MFE programs are a combination of the three fields: Math, Programming and Finance. This is why I'm coming to you guys for answers. I have the finance down great, but I've got to learn the harder subjects now.

I've realized that learning to program is vital to becoming a quant. Is this mainly to build models? If not, what is the main purpose of becoming so proficient at programing? What does it exactly mean to build a model through programming? I obviously have limited knowledge of programming, which is why I'm asking these questions.

My knowledge in math is limited to what has been required at my school, which is Calculus I and II. What else is recommended for me to try and teach myself? Should I try to learn programming first? or more math?

Is there a benefit to learning more math before programming or vice versa?

Thanks in advance for any responses. I truly do appreciate it.
Josh

 

[Lyosha]

Linear Algebra is a must (and a good start), but it's only scraping the surface. Differential equations, analytical probability and stochastic calculus round out the things you really need to know to not be completely lost... and are... much more difficult in my opinion (throw in calculus based statistics in there somewhere as well...)...

For a decent idea of what it means to "program a model", I would suggest Joshi's C++ patterns in finance book. Amazon.com: C++ Design Patterns and Derivatives Pricing (Mathematics, Finance and Risk) (9780521721622): M. S. Joshi: Books: Reviews, Prices more

Note that this book is not enough if you want to actually learn/understand object oriented design patterns as it essentially only deals with singleton and factory. A different, more in depth book should be used for that purpose.

Also note that it assumes basic understanding of C++, so you might want to get some practice with the basics first. People here tend to like Savitch's book, and it's pretty good but wordy. Just about anything from your local barnes and noble should do the trick of getting you up to speed enough to look into Joshi's book.

The amount of programming you need to know depends on what it is that you want to do. Different roles require different amounts of programming prowess, but a general rule is that more is not less.

You can probably learn math and programming simultaneously. I don't think there is much to gain by learning them in some order.

 

[jgills]

You have much to learn by way of math - Linear Algebra is a must (and a good start), but it's only scraping the surface. Differential equations, analytical probability and stochastic calculus round out the things you really need to know to not be completely lost... and are... much more difficult in my opinion (throw in calculus based statistics in there somewhere as well...)...

For a decent idea of what it means to "program a model", I would suggest Joshi's C++ patterns in finance book. Amazon.com: C++ Design Patterns and Derivatives Pricing (Mathematics, Finance and Risk) (9780521721622): M. S. Joshi: Books: Reviews, Prices more

Note that this book is not enough if you want to actually learn/understand object oriented design patterns as it essentially only deals with singleton and factory. A different, more in depth book should be used for that purpose.

Also note that it assumes basic understanding of C++, so you might want to get some practice with the basics first. People here tend to like Savitch's book, and it's pretty good but wordy. Just about anything from your local barnes and noble should do the trick of getting you up to speed enough to look into Joshi's book.

The amount of programming you need to know depends on what it is that you want to do. Different roles require different amounts of programming prowess, but a general rule is that more is not less.

You can probably learn math and programming simultaneously. I don't think there is much to gain by learning them in some order.

Thanks for such a fast response Alexei. I haven't taken the GRE yet, but I know with proper review I will do very well. Should I focus on getting a high quant score on the GRE before I begin my endeavors into the more complex math and programming?

I guess what I'm really looking for is conformation to the plans I come up with in my head from someone who has already been through it.

Thanks