Quant-trader career advice

Quote from CalVolibrator:

Are you feeling you are actually lacking the skill sets to make the transition or is it that you are just not given the opportunity? When you say electronic trading desk you mean the equity DMA team? Because if that is the case and you show you are hungry to transition plus bring solid programming skills to the table then I do not see why you could not make the transition. I would not bother with course work and university degrees at this point. Strengthen your programming, analytical skills and bother traders at desks with questions and offer them in exchange to do projects for them. I would say that is your number one best bet. I would never reject someone who approaches me at my desk and offers to work on a project in exchange to learn something. In fact, if the project result was impressive it would be my first candidate should I have room to expand.

Please do not listen to people who only have retail experience, nor listen just to my advice, but I am fairly certain that you will hear the same sort advice from most other professional traders.

P.S.: When I say do not listen to them I do not mean do not listen to anyone without professional experience. But please stay away from those who suggest to use Multicharts and all other kinds of crap or suggest you to quit your job. Ridiculous advice.
More...

Probably not the best choice of boards to post my question, thanks for being one of the few posters with a genuine reply.

I'm in the FX space and despite the way I may have come across, am in a fairly senior position. My view is that the market will continue to become more and more technology dependent. Coming from a trading background, feel like I will be at a disadvantage in several years time unless I make an effort to increase my technical abilities. So my question regarding post grad study wasn't so much to transition into another role because I'm unhappy with my current one, it was more to provide me with a structured way to learn new abilities which may come in handy in the future (coding, data mining, backtesting, etc). Have looked into doing an M.sc in Computing (Machine Learning) but haven't been able to find a course I could do part time while working.
 
So a fair description of what you current do is working at an eFx desk or a desk that is fairly mechanical fx but may turn into a pure eFX desk in the future?

eFX is not and never will be heavily quantitative even if they fully automate processes. If you want to transition to a pure quant desk in the future away from fx then yes, maybe the best route would be to undergo a Financial Engineering program. However, if you want to stay in fx space and just want to make sure you are not made obsolete then I still believe that strengthening your programming and modeling skills on the side would most likely suit you best. You could start with time series analysis in R. There are countless online opportunities available regarding that, why do you need to look for a course type of learning experience, people in financial space hardly care about the difference.

Quote from xroads84:

Probably not the best choice of boards to post my question, thanks for being one of the few posters with a genuine reply.

I'm in the FX space and despite the way I may have come across, am in a fairly senior position. My view is that the market will continue to become more and more technology dependent. Coming from a trading background, feel like I will be at a disadvantage in several years time unless I make an effort to increase my technical abilities. So my question regarding post grad study wasn't so much to transition into another role because I'm unhappy with my current one, it was more to provide me with a structured way to learn new abilities which may come in handy in the future (coding, data mining, backtesting, etc). Have looked into doing an M.sc in Computing (Machine Learning) but haven't been able to find a course I could do part time while working.
More...
 
You'd be wasting your time with ML if your singular goal is to get into quant trading. On the hand if it's something you're actually interested in by all means.
 
CalVolibrator - eFX is correct. Why do you think it will never be heavily quantitative in the future? Any good online resources you would recommend for learning to use R?

slickpick - ML was just an option I was considering. Like I mentioned before, the only reason I would consider doing a course is because of the structure it provides to learning, as well forces me into some sort of commitment to accomplishing my goals. Feel if I try to do this in my own time, will be very easy to make excuses and drift. Not easy to come home after a long day at work and motivate myself to continue working.
 
If you think your math skills are weak, go with M.fin

As far as programming goes, most quants use python, matlab, r and maybe light c++ with exceptions. If you have a programming background I feel you can learn most of the programming on your own.

Knowing how to backtest strategies effectively is one of the most important skills. New ideas are great until the market moves against you.

I agree with what most people have said: don't quit your day job, do this in parallel, and getting the job is all about networking.

The only thing I would add is that if you want to wind up on the buy side, I suggest transitioning laterally at an IB first to get experience and a proven P+L history. I find the buy side to be less forgiving of mistakes.
 
No one has told you what quantitative trading is....it is not just backtesting trading ideas, that is only a small piece of the puzzle.

Quantitative trading is the statistical analysis of a trading instrument and/or a basket of trading instruments. After coming to a possible conclusion on how a market or instrument behaves you then implement a trading algorithm to exploit this statistically found behavior. The programming of the algorithm doesn't have to be very complicated.

Focus on statistical analysis of time series data. Most educational institutions focus on derivatives which is not day or swing trading related. There are very few resources available on quantitative analysis. Most quantitative trading resources focus on implementing and backtesting the algorithms and thats only half of it. Commong implementations of machine learning or A.I. have not been very useful/profitable.

A retail trader would use tools such as R or Excel for analysis and trading platforms such as NinjaTrader or Multi-Charts for trading algorithm implementation in C#.
 
Quote from CalVolibrator:

So a fair description of what you current do is working at an eFx desk or a desk that is fairly mechanical fx but may turn into a pure eFX desk in the future?

eFX is not and never will be heavily quantitative even if they fully automate processes. If you want to transition to a pure quant desk in the future away from fx then yes, maybe the best route would be to undergo a Financial Engineering program. However, if you want to stay in fx space and just want to make sure you are not made obsolete then I still believe that strengthening your programming and modeling skills on the side would most likely suit you best. You could start with time series analysis in R. There are countless online opportunities available regarding that, why do you need to look for a course type of learning experience, people in financial space hardly care about the difference.
More...

Look, what good does jumping to r time series analyses do for him?

His quantitative techniques aren't developed or that he's spoken about any strats to use. This isn't about being a tool, it's about that I've found coding programs is useless so buy OTS technology especially mc because the data sets are so much more useful that the skillset is inside anybody who has played programmatic complex computer video games so each logic set he wants to use will need development and it is the automation here that's fairly new to the question of how to be a quant because they are the commanders making weapons and not the investors trusting advisors to know how to use them.

This arbitrary promotion of continuing education is unnecessary if you only want to make weapons and not all the ones that should be used before the nukes are broke out to end all war. That's just my own attitude between self-determination of analysis and how that makes my ideas different than the useless autobot programming mid-point dipdrop chicken scratch of half a pip for someone's peso-dollar conversion.

I guess it pays the bills, though, if you're in the business but not anywhere near what he's talking about as a quant.
 
I would argue with you if you had ever worked as quant in a professional environment which you obviously have not. eFX is a very straight forward business, it has nothing to do with deriving pricing algorithms such as one would as part of an exotic options trading desk. eFx is all about knowing your customers and to either widen or tighten spreads as a function of the sophistication level and pnl generated by said clients. Its sad but the truth. Stop suggesting him to use retail tools that have no place whatsoever in a professional sell-side setup.

Quote from bwolinsky:

Look, what good does jumping to r time series analyses do for him?

His quantitative techniques aren't developed or that he's spoken about any strats to use. This isn't about being a tool, it's about that I've found coding programs is useless so buy OTS technology especially mc because the data sets are so much more useful that the skillset is inside anybody who has played programmatic complex computer video games so each logic set he wants to use will need development and it is the automation here that's fairly new to the question of how to be a quant because they are the commanders making weapons and not the investors trusting advisors to know how to use them.

This arbitrary promotion of continuing education is unnecessary if you only want to make weapons and not all the ones that should be used before the nukes are broke out to end all war. That's just my own attitude between self-determination of analysis and how that makes my ideas different than the useless autobot programming mid-point dipdrop chicken scratch of half a pip for someone's peso-dollar conversion.

I guess it pays the bills, though, if you're in the business but not anywhere near what he's talking about as a quant.
More...
 
You must log in or register to reply here.
Back
Top