• C++ Programming for Financial Engineering
    Highly recommended by thousands of MFE students. Covers essential C++ topics with applications to financial engineering. Learn more Join!
    Python for Finance with Intro to Data Science
    Gain practical understanding of Python to read, understand, and write professional Python code for your first day on the job. Learn more Join!
    An Intuition-Based Options Primer for FE
    Ideal for entry level positions interviews and graduate studies, specializing in options trading arbitrage and options valuation models. Learn more Join!

Turning trading hobby into Career

Joined
9/19/13
Messages
4
Points
11
I am a young stay at home Mum that has developed an interest in High Frequency trading. I am constantly reading blogs, articles, books, you name it trying to educate myself. I have recently decided to enter into formal education and try and develop a career in something that really interests me. So far I find that I am really drawn to algorithmic trading.

At this time in our lives my family and I are located in a regional Australian town for my husbands work. So I am restricted to distance learning programs. I would prefer a UK based distance learning degree. London school of economics and Politics will allow me to do Mathematics and Economics in 2-3 years. On completion we will most likely be living in London where I would like to do a Masters in Computer Science and most likely an additional Masters Financial systems Engineering, Hopefully from UCL. This is not my most desirable education pathway. I am hoping that I can avoid doing a PhD unless it really interests me.

I would like to get into work as a High frequency trader or something alike. At the very least I would like to be fully equipped to work from home and become the Bread winner. Is all this education worth it if I am content in trading from home?
 
I am going to need to ask the question, have you ever programmed anything? Programming skill is not something you pick up in a class, it is more akin to art, you should be very passionate about it. HFT is a very fast paced field it will be hard for most people to believe a stay at home mom can keep up. Could you present a counter-argument please?
 
I would not expect a stay at home Mum to keep up either. But I think a tertiary educated young woman may have a chance. Would you have the same idea about an accounting graduate. I have my Bachelors in Accounting already. I find that boring as all bat shit. Probably because it is not fast paced. I have also been day trading with CMC Markets for 2 years now with good returns. I am ready to move on to the next challenge and learn more. No, I have not programmed anything on my own. My brother has started showing me basics and I have enrolled in a free course in object orientated programming to see how go.

I have recently been introduced to Ninja Trader by a friend of a friend with a common interest. He was able to learn a lot on his own and learn programming through some night courses. I plan to get comfortable with this software in the short term but I would love to learn more. This is why I was thinking that computer science education would be beneficial to me.

I understand that HFT is fast paced but with the right education and software I should think that any person would stand a chance. My question was would it be worth it for me to get formal education requirements or to just use an already established HFT trading platform.
 
Well I asked the question to see specifically where you were in Computer science knowledge/passion. I currently have friends who spend around 9 hours a day just doing math/programming. The thing about programming at that level is NO ONE can teach you how to do it.

If you wanted the education as badly as is necessary to succeed in that space then you would have realized that all the education you need is all over the internet free of charge(have you seen coursera.org?/stack exchange?/Stachurski's website?/youtube?/numerous pdf's online?). It's not as if you need some gold banner hanging up in your living room. If you want knowledge it is there; it is not exclusive to attending a top university.
 
Well you answered my question then. I was unsure as to whether it would be enough but now I know I am on the right track in self education. Well when I get to 'that' level in programming we will see if I can handle it. At the very least I'm sure I will get by using a software like Ninja Trade.
 
Once you reach the level of technical fluency (many people never do), you will be able to read HFT literatures. The rest is up to you to implement the idea and find your own.
You do not need credentials to run your own shop. But you need a lot of knowledge and experience (by losing money ;) )
Unless you discover your edge, retail traders fight up hill battle against institutional players who invest millions in infrastructure alone.
Don't expect to turn on the machine and the money will roll in.
 
Last edited:
Dear finance mun,
1) Retail software can't program HFT algo simply because they can't load nano seconds tick feeds.
2) There is no universities out there teaching you the secret behind HFT. Even if you can find any out there, those were merely theories that explained the phenomena.
3) The inventor who created HFT never disclosed the actual theory just like KFC never disclosed their secret recipes.
4) If you can program an algo that do scalping along with other institutional players, prove it with live trading. Scalping does not requires HPC infrastructural. Back testing algo is not reliable to avoid curve fitting.
5) No formal education is requires for building algo trading strategies.
6) EXPOSURE in real trading house is a must. People do not hire what you have learned from school, but what you can do with your experience and knowledge.
 
I see everyone concern. I'm still determined to give this a good shot though. Perhaps I will aim a little lower in the mean time. I will take tertiary level courses in computer programming though. I may never become an expert but I intend to learn as much as I can. I'm not looking for a secret I understand that once you build up the skills its trial and error and that some people naturally have the gift and others don't. I would eventually like to be able to build my own algorithmic trading strategies. It might take years but it is a goal I can work towards.

What are your thoughts on programs like Rizm in the meantime?
 
it is a retail platform just like any other out there.

If you build a retail algo trading strategies, you make no difference from other since there a lot of 3rd parties ready made algo.

Nobody hire retail algo trading strategies.
 
It's good that you are profitable day trading, big plus. However, when you go and try to implement a strategy, you realize that you do a lot of things intuitively or in your head, and these things can be exceptionally hard to quantify and/or program. Even if you can write down/quantify the strategy more or less, it can take a long time to actually build up the infrastructure for trading. And not all of it is just grinding away work, you often have to make hard design decisions.

So basically, what I would say is -- the fact that you are making money in discretionary trading is a big plus, but the leap from discretionary to algorithmic is a big one, especially if you have no experience writing software.
 
It takes anything from 5-20 years to become an experienced programmer. It much more difficult than maths (you don't learn maths, you know it or don't). Progamming is a skill you have to learn, like getting belts in judo or karate.
 
It takes anything from 5-20 years to become an experienced programmer. It much more difficult than maths (you don't learn maths, you know it or don't). Progamming is a skill you have to learn, like getting belts in judo or karate.
So true

BTW I bought your book :D

Well I disagree with many things , one of the main core problem of C++ is the lack of visualization of data.
The point is so relevant, that today most new language are coming with DAta Visulization Package.

Sorting, Interpolating, Matric computation, etc... that's all things you won't need to do anymore... but using C++ you still lack of vision of what you doing even when you are a praticionner.
 
...one of the main core problem of C++ is the lack of visualization of data.
The point is so relevant, that today most new language are coming with DAta Visulization Package.

Do you mean Visualization Library? C++ is a very low level language. Those libraries do exist but are not included in the standard of the language (and they will never be):

- OpenGL http://www.opengl.org/
- Qt http://qt-project.org/
- wxWidgets http://www.wxwidgets.org/
- Windows has its own thing (I don't know what's called nowadays)
- etc...
 
YOu don't have to be a master programmer to design trading algorithms. Depending on complexity you could do with high school math. Implementing the algorithms also goes from easy to rocket science. So you need to develop an idea first of what you want to do and at what level.

(edit: While it may not be hard to develop any trading algorithm, the difficult part is creating a profitable one.)
 
Last edited:
Do you mean Visualization Library? C++ is a very low level language. Those libraries do exist but are not included in the standard of the language (and they will never be):

- OpenGL http://www.opengl.org/
- Qt http://qt-project.org/
- wxWidgets http://www.wxwidgets.org/
- Windows has its own thing (I don't know what's called nowadays)
- etc...
No ...
My point is , Visual C++ or any Debugged doesn't allow you to manipulate data..

You could say that the JIT is emulating the needs but with very poor capabilities... it take no effort to actually ship the data as external data allow manipulation into a shell.
Doing that in interpreter is not even a problem at all... just because C++ is compiled shouldn't mean , you need to be into a running process and manually drop the data into a shell yourself to explore it.

Debugger in C++ should always be a shell in a first place and leave programmer free to explore results and intermediate and then ... continue the process.
 
No ...
My point is , Visual C++ or any Debugged doesn't allow you to manipulate data..

[...]

Debugger in C++ should always be a shell in a first place and leave programmer free to explore results and intermediate and then ... continue the process.

That's why there's R. You can explore your data and develop the algorithm in R. Then you can deploy in either Java or Python (there's no reason to use C++ unless you have specific performance or interface requirements).
 
Back
Top