Graduating and Lost - Need Advice

[dinogoesrawr]

Looking for some advice on next steps.

Graduating from LSE with a degree in Stats/Finance this summer. Have obtained a 2:1 already, could get a first depending on how the exams go this summer. Got a 1st class in calculus/linear algebra modules (I find it 'naturally' easy) and some coursework modules that involved learning 'programming' (R, AIMMS, @Risk Excel add on which involved Monte Carlo simulations). 2:1 usually in finance (corp finance not quantitative finance) and economics/actuarial subjects.

Interned in IBD/M&A and thought it was ok. Interned with Facebook/Google/Apple on the business side (not the engineering side). Was encouraged to find something more challenging by senior staff in both cases. I also want something more technical (only the modelling was fun during the internships for me) but don't feel I have the skill set yet (or the brain haha).

These signs are pointing to a career in markets - however - I do not follow the financial news (e.g. oil prices/macroeconomics). I only read about M&A deals that appeal to me e.g. Facebook buying WhatsApp.

Problems:
1/ LSE is definitely not Cambridge/Imperial for math so perhaps I'm not actually that good at math I know people at Imperial/Cambridge who are olympic gold medalists in math etc - I'm not that at all. This is why I chose a career in IBD/tech - I thought the 'mix' of technical/mushy would help.
2/ We only learn R programming (for regression) in stats - no MATLAB etc - so I do not feel set up well for the working world as most technical jobs looks for much more technical skill.
3/ Its too late to apply for a Masters for the September 2014 entry and I'm graduating without an internship/job lined up.

What do I do with my life as I feel completely lost?

 

[bigbadwolf]

1) Look for a job.
2) Start the long and arduous process of learning to code (C/C++/Python/Matlab), along with some numerical analysis.

On a side note, math is overrated in finance, and Imperial is overrated (though Cambridge really is in a class of its own).

 

[dinogoesrawr]

1) Look for a job.
2) Start the long and arduous process of learning to code (C/C++/Python/Matlab), along with some numerical analysis.

On a side note, math is overrated in finance, and Imperial is overrated (though Cambridge really is in a class of its own).

Thanks for the quick reply.
1) What type of job should I look for? I'm trying tech firms and banking, based off my previous experience, but nothing is panning out.
2) Are there courses for coding or do I self-teach?

 

[bigbadwolf]

Thanks for the quick reply.
1) What type of job should I look for? I'm trying tech firms and banking, based off my previous experience, but nothing is panning out.
2) Are there courses for coding or do I self-teach?

I haven't lived in England for close to 20 years now so I'm not competent to advise you. I suspect that with a degree in stats/finance you will have a lot of competition for the jobs you might think of applying for.

With regard to coding, there are courses but I think every expert coder is largely an autodidact regardless of whether he has or has not taken formal courses. If I'm right, then the main thing is to know which books to get hold of and make a little study plan. For starters, that is. Formal courses are an intro to syntax with some lightweight exercises and projects thrown in. If you think of, say, C++ as a language, that's analogous to teaching some phrases and basic vocabulary to tourists. It's not command of the language, i.e., not the ability to use it with dexterity, to be able to translate ideas and algorithms into working code with ease. That comes with time, practice, repetition, and a gradual extension of one's powers. There are no shortcuts.

 

[Liam]

Looking for some advice on next steps.

Graduating from LSE with a degree in Stats/Finance this summer. Have obtained a 2:1 already, could get a first depending on how the exams go this summer. Got a 1st class in calculus/linear algebra modules (I find it 'naturally' easy) and some coursework modules that involved learning 'programming' (R, AIMMS, @Risk Excel add on which involved Monte Carlo simulations). 2:1 usually in finance (corp finance not quantitative finance) and economics/actuarial subjects.

Interned in IBD/M&A and thought it was ok. Interned with Facebook/Google/Apple on the business side (not the engineering side). Was encouraged to find something more challenging by senior staff in both cases. I also want something more technical (only the modelling was fun during the internships for me) but don't feel I have the skill set yet (or the brain haha).

These signs are pointing to a career in markets - however - I do not follow the financial news (e.g. oil prices/macroeconomics). I only read about M&A deals that appeal to me e.g. Facebook buying WhatsApp.

Problems:
1/ LSE is definitely not Cambridge/Imperial for math so perhaps I'm not actually that good at math I know people at Imperial/Cambridge who are olympic gold medalists in math etc - I'm not that at all. This is why I chose a career in IBD/tech - I thought the 'mix' of technical/mushy would help.
2/ We only learn R programming (for regression) in stats - no MATLAB etc - so I do not feel set up well for the working world as most technical jobs looks for much more technical skill.
3/ Its too late to apply for a Masters for the September 2014 entry and I'm graduating without an internship/job lined up.

What do I do with my life as I feel completely lost?

Wouldn't look at gold medallists - a lot of hot air. Skills is what it's about and some of these unis shy way from coding. In modelling or stats applications within work having done this stuff is useful, but judiciously using the right areas efficiently isn't something that requires gold medals.

I've noticed, of late, a lot of arrogant twats graduating with the right stuff from these unis - friends that hire never employ these people. To be fair, having been a quant I would say that it's a job where the percentage of the job that is covered by MSc or PhD is much, much higher than other jobs I've done, so their arrogance is understandable. But there's a lot of important stuff around modelling that's missed even in MFE projects e.g. building assumptions from scratch and even choosing which model to use. Also getting models approved, getting risk on board is something that requires the right attitude. Show you understand that from your internships and show humility then your lack of medals is insignificant.

Firstly - at some point think study and forget jobhunting. Even with good markets I've seen people go nowhere with jobs and fail their degree because they're so busy interviewing. While rest of us passed and took months to be employed, which felt like shit but was better than winding up with a failed degree. Sad reality of the situation.

As previous poster said study plan is most important. Coding simple games and models would get more attention than any solving equations. I covered stuff like file i/o at home doing this, but never once used this in a course. There's also books on C++ design patterns that cover proper practical cases. I despise the word "course" - maybe head paperclip counter at Random Blah Blah Firm needs some stupid management course, but many coding courses miss things like edge cases and other issues you'll only consider within a job or a proper project. Perhaps try find a public dataset and then work on that. A quant project would work, but this and the study plan will take months - this is the toughest maths I've seen. There are other fields where hardcore stats are needed which you might enjoy more. But ultimately this isn't a shortcut - as the previous poster said, none exist. Maybe fire up a few Hadoop clusters and build something like a fraud analyser, there's loads of other ideas. Also don't be so bothered about not having things "lined up", it's not med school where there's standardised routes, don't be fooled by ideas like "milkround" hiring. Competition is one thing but what's asked of a maths grad these days is such that most maths people I know took months to get jobs with one or two exceptions. Firms that hire quants therefore don't expect you to have something "lined up" but will expect you to be looking for a career not a "job", the passion and work towards it is what matters.

From a practical perspective R is great - might not be as advanced as C++ but within things like film recommendation R algorithms can be way faster than other alternatives like Python or Hive/SQL. Also given how easy it is, I would rather write simple R code than C++ where C++ is not needed. Depends on what jobs you're looking at, also bear in mind, in spite of what people say, sectors like quant or IT will still have issues with cash for training hence some monstrous job specs to cover that. This issue is ignored by crude articles which go along the lines of lots of jobs => quids in.

Also be aware if you can code some line managers won't be too bothered if you've never seen VBA or SQL - they can be taught to an experienced coder within weeks, if not days. Same with Python if you've done R. Not so much with C++ or Java - learn that through project/study and as the previous poster says practice, practice, practice.

 

[Liam]

There are no shortcuts.

I've done various mathematical and finance roles for 9 years and that's the lesson I've learnt the most.

As a recruiter said to me once 6 month courses in coding or modelling "don't cut the mustard".

 

[aero_malik]

I'll piggyback on the coding comments. Pick up decent and comprehensive textbooks in algorithms and data structures in C++, after your introduction to the language. Learn how to work with real-world datasets. I/O is very important. Learn when to ignore C++ and stick to a more straightforward language like R, MATLAB or Excel VBA.