Mr Rufus is on target. I'd boil it down to: If you have money you're willing to use as working capital, you can hire good people and get down to business. If you don't, you can bring in good people as equity partners, which also works. Either way, if you have an edge, everthing will work out fine.
You are using an out of date browser. It may not display this or other websites correctly.
You should upgrade or use an alternative browser.
You should upgrade or use an alternative browser.
career paths of automated trading system developers
- Thread starter hfoptimizer
- Start date
BearTrading
ET Sponsor
Quote from hfoptimizer:
My story so far...
I develop automated trading strategies for a company that will remain anonymous. Out of respect I will avoid saying or suggesting anything negative except of course to acknowledge there are problems that have caused me to threaten to quit twice and brings me to this post.
I've developed automated trading systems for about 10 years, initially on my own, then for an employer later. I've now reached a high enough level as an system developer where profitability is no longer the only major issue. I now feel that honesty, trust and doing things right matter as least as much. I now know that developing profitable trading systems is actually a scientific process. Of course there are many do's and dont's, one must be pragmatic and motivated with a decent software and math background. Given all that, automated trading systems can be done with surprising ease and efficiency. I've been told I'm the most efficient and productive system developer at my company. My trading strategies from several years ago built the track record that now sustains the company.
I run a computing cluster simulating billions of trades per second testing 100s of different strategies per second and always push for more efficiency. I also push to build a stronger team and hire some top-notch software engineers because I know the value of such people and because the codebase is becoming too difficult for me personally to manage. Can't utilize CPU resources properly without good people writing good algorithms.
I could go on. But most will get the point. Without going into negative details it is becoming quite clear I should quit and start my own operation as I had years ago. I find myself thinking about business strategy and building better morale, trust and a solid team at least as often as I think about designing a better trading system analysis framework.
Anyone else in a similar situation?
We have a an algorithmic trading arm at Bear that we think could be something that you are looking for.
It is comprised of a cluster of system architects, developers, coders, testers, quants and traders.
Please feel free to contact us (PMs are always fine as well) to see if there is a potential fit.
Thanks.
I am sorry if this is not the exact place to post this topic..
I am an IT developer, and curious to know what is it like to be a trading system developer, what is front office?
Can someone please give examples of what is involved in a trading system developer's work life during a day?
(I was confused when someone told me that developers have to interact with traders in a typical firm. Is the software customized for each trader? )
Any pointers or any information about the examples of role of a trading systems developer will be highly appreciated.
I am an IT developer, and curious to know what is it like to be a trading system developer, what is front office?
Can someone please give examples of what is involved in a trading system developer's work life during a day?
(I was confused when someone told me that developers have to interact with traders in a typical firm. Is the software customized for each trader? )
Any pointers or any information about the examples of role of a trading systems developer will be highly appreciated.
hfoptimizer,
I can't help you with your problem but I would love to hear a bit more about how you're using clusters to make trading decisions. What type of software are you using to do this?
Runningbear
I can't help you with your problem but I would love to hear a bit more about how you're using clusters to make trading decisions. What type of software are you using to do this?
Runningbear
Wow, old thread.
I think Rufus nailed it. After a decade of working with both startups (as founder as well as seed investor) and large corporates.... Truth is, entrepreneurship is both easy and hard. For those who "get it", launching a business is the easiest thing in the world: find your competitive advantage, implement it, aim high.
For those who haven't been in that type of environment, I understand why it's daunting. A lot of very smart people have never setup an office, deployed a colocated system, talked to customers, configured a cvs archive from scratch, negotiated rent, designed a business card, written a budget, etc etc etc.
At the end of the day, for those with the financial and emotional buffer needed to give this a try.... I encourage you to go for it. For those of us leveraging technology, you really don't need a "team". This is still the golden age for garage startups, and you will figure it out as you go. Use third world freelancers, if you really need help with grunt work (or that business card!).
I think Rufus nailed it. After a decade of working with both startups (as founder as well as seed investor) and large corporates.... Truth is, entrepreneurship is both easy and hard. For those who "get it", launching a business is the easiest thing in the world: find your competitive advantage, implement it, aim high.
For those who haven't been in that type of environment, I understand why it's daunting. A lot of very smart people have never setup an office, deployed a colocated system, talked to customers, configured a cvs archive from scratch, negotiated rent, designed a business card, written a budget, etc etc etc.
At the end of the day, for those with the financial and emotional buffer needed to give this a try.... I encourage you to go for it. For those of us leveraging technology, you really don't need a "team". This is still the golden age for garage startups, and you will figure it out as you go. Use third world freelancers, if you really need help with grunt work (or that business card!).
Quote from heech:
Truth is, entrepreneurship is both easy and hard. For those who "get it", launching a business is the easiest thing in the world: find your competitive advantage, implement it, aim high.
For those who haven't been in that type of environment, I understand why it's daunting. A lot of very smart people have never setup an office, deployed a colocated system, talked to customers, configured a cvs archive from scratch, negotiated rent, designed a business card, written a budget, etc etc etc.
Finding a real competitive advantage is the hardest part. I've seen many smart people fool themselves, then fool others into thinking there is a competitive advantage.
I've seen some that start off with absolutely no competitive advantage, but are good at setting up an office, talking to customers, and sales materials. It quickly takes on the illusion of a competitive advantage. Sometimes they then discover a competitive advantage even when they started with none.
hfoptimizer -
I would compartmentalize everything -
Write a business plan based on contingencies:
1) Getting capital - find a seed firm or find a startup partner
2) Developing / executing strategies
3) find help via coding website - odesk etc.. for moderate to low priced coders..
4) find professional to manage the technical aspect of trading stocks or futures... accounting / backoffice. You could likely outsource this if you had a reputable Broker and big enough size.
Obviously if you test that many strategies simultaneously you are using something like KDB+, otherwise it doesn't scale and you end up with a lagging system susceptible to data delay or some backtesting artifacts..
You also know that scaling your ideas is really the only viable mechanism for you to evaluate your trading strategy and the resulting potential size of the business.
When you have those topics well defined and understand, potential investors and employees will "get it" and follow suit.
I would compartmentalize everything -
Write a business plan based on contingencies:
1) Getting capital - find a seed firm or find a startup partner
2) Developing / executing strategies
3) find help via coding website - odesk etc.. for moderate to low priced coders..
4) find professional to manage the technical aspect of trading stocks or futures... accounting / backoffice. You could likely outsource this if you had a reputable Broker and big enough size.
Obviously if you test that many strategies simultaneously you are using something like KDB+, otherwise it doesn't scale and you end up with a lagging system susceptible to data delay or some backtesting artifacts..
You also know that scaling your ideas is really the only viable mechanism for you to evaluate your trading strategy and the resulting potential size of the business.
When you have those topics well defined and understand, potential investors and employees will "get it" and follow suit.
Sorry I should have updated this thread earlier.
psytrade:
I have a partner. We have some viable strategies. Seeding capital has already been offered.
I am already an experienced coder and strategy developer. No need to hire anyone.
Definitely no need for KDB+ either. Market data and computational intermediates are easily cached on disk or in memory on each cluster node.
Yet, there is a problem that holds me back. Strategy development is exponentially more effective with more sophisticated data. Knowing what I know, I will wait to pull the trigger until I have access to the data I want. Otherwise I will wait and do something else.
psytrade:
I have a partner. We have some viable strategies. Seeding capital has already been offered.
I am already an experienced coder and strategy developer. No need to hire anyone.
Definitely no need for KDB+ either. Market data and computational intermediates are easily cached on disk or in memory on each cluster node.
Yet, there is a problem that holds me back. Strategy development is exponentially more effective with more sophisticated data. Knowing what I know, I will wait to pull the trigger until I have access to the data I want. Otherwise I will wait and do something else.