Forums  > Careers  > Master's degree but no full-time work experience. Where to start?  
     
Page 1 of 2Goto to page: [1], 2 Next
Display using:  

Riane


Total Posts: 3
Joined: Nov 2015
 
Posted: 2015-11-09 15:14
Hello everyone,

I'm a person who, right after getting my degree in economics, decided to go get a master's degree in statistics. I think this gives me a great education, but leaves me in a weird spot.

I'm quite good at modeling and coding, I can write SQL and R in my sleep. Next semester I'm taking a class on VBA and I have some background in Python too. In addition to this, I have had five internships totaling 1 year 2 months of internship experience.

Hiring managers do respect this, but apparently it is not a substitute for full-time work experience. I can't get a job that requires 1-2 years of experience.

So where do I start? I've been applying to risk/quant/data science jobs mostly so far.

Any ideas?

HitmanH


Total Posts: 463
Joined: Apr 2005
 
Posted: 2015-11-09 16:28
Where are you based?

Also - quite good at coding? In what language? As SQL/R aren't really coding languages - and looks like only starting with VBA (don't) and Python...

MCgeneratedname


Total Posts: 143
Joined: Dec 2013
 
Posted: 2015-11-09 17:48
Welcome.

A word of warning as you seem to be interested in a statistical/quantitative field: You might have noticed skimming through the careers section that many areas in finance are closed for people without PhDs. Stats/Quant is one of them.


Riane


Total Posts: 3
Joined: Nov 2015
 
Posted: 2015-11-09 18:10
ok MCgeneratedname, that's what I needed to hear.

say I wanted to set myself up to get a phd in a couple of years (stats, econ, finance, computational finance, etc.) so that eventually, I can transition into a quantitative role.

Where would be the best place to start then? I really do like to code so i'd want something with a lot of that.

MCgeneratedname


Total Posts: 143
Joined: Dec 2013
 
Posted: 2015-11-09 18:42
Best place to start a PhD? The best university you can get into. Preferably with the best known guy in your field as supervisor. You might want to avoid the softer fields like economics... People have little respect for them.

Best place to start working? Prop shops, the large CTAs and hedge funds.

But all of this is rather obvious. The real question is should you do a PhD just to get a quant role? The answer to this is NO! Because you need a good PhD in pure math / stats. But to get a good one you need to be intrinsically motivated and not motivated by career goals. There are some threads on PhD necessity on here...

The other problem is that even if you get a very good PhD you are still competing with an army of similar candidates.

Coding is nothing special to be honest. It's your daily bread and butter. So it is rather something you pick up along the way than a special skill that distinguishes you.

EspressoLover


Total Posts: 337
Joined: Jan 2015
 
Posted: 2015-11-11 00:34
Depends if you're at a target school. Masters or even bachelors in stats with a lot of CS classes from Wharton, Harvard and Stanford frequently are recruited for top-tier hedge fund and bulge-bracket quant positions.

Good questions outrank easy answers. -Paul Samuelson

Riane


Total Posts: 3
Joined: Nov 2015
 
Posted: 2015-11-11 00:42
I'm not from a target school lol. I go to a top 20 state school and will have to network my ass off to get to whoever will hire me.

EspressoLover


Total Posts: 337
Joined: Jan 2015
 
Posted: 2015-11-11 01:04
You might have better luck in the Chicago prop scene. They tend to be less Ivy-discrimatory. You just have to be a lot more selective and do some background research before accepting an offer. There's a lot of joke shops out there, that you don't want to have on your resume. (There's also a lot of fantastic ones).

The biggest comparative advantage you have from a state school is the much larger size of your alumni network. Your best bet is to reach out to alumni contacts at the places you might be interested in. For example Penn State always seems to have a surprisingly large Wall Street presence, if not just because Penn Staters love other Penn Staters, and there are so many of them everywhere.

Also I'll add that if you've had 5 internships, you should reach out to people from those. Even if you don't want to work for any of those companies, you're likely to have contacts that have moved on to other places, or at least know people. Getting a single recommendation even from a low-level employee, gives you a massive advantage relative to an unsolicited resume drop.

Good questions outrank easy answers. -Paul Samuelson

Lebowski


Total Posts: 70
Joined: Jun 2015
 
Posted: 2015-11-11 01:11
> Depends if you're at a target school. Masters or even bachelors in stats with a lot of CS classes from Wharton, Harvard and Stanford frequently are recruited for top-tier hedge fund and bulge-bracket bank positions.

I'll second that. Despite my hiring woes discussed a couple threads below, now that more companies have started getting back to applicants I've found some pretty good opportunities with only a bachelors in Math/Stats + a CS concentration from a target school. A PhD or equivalent experience is a requirement for certain roles, but by no means for all. Anecdotally, there seems to be more hard PhD requirements in the UK.

Perhaps as an undergrad I should bite my tongue and just stick to confirming the above post, but I'll offer my experience...feel free to disregard it.

R and SQL won't hurt your resume, but a compiled language and some OOP experience would probably go a long way toward convincing recruiters that you can code--a valuable skill for entry level roles. In my experience, most trading jobs will give you a "coding challenge" of some kind during the interview process that will be pretty difficult to take down without some elementary data structures and algorithms knowledge. Python is a perfectly acceptable object oriented language, but so many people know a little Python but can't address basic interview questions that come up time and time again like "what's polymorphism" or "abstract class" vs "interface" that I'm hesitant to say it will be sufficient to attract recruiters without the requisite OOP background. The good news is that you can do a lot to prepare for the coding aspect of interviews fairly quickly, so be sure to brush up on data structures as it's fairly low hanging fruit. Some shops like Ketchum and Allston seem to only care about R for entry level trading roles, so maybe those might be good places to look.

Good luck. I struggled a lot on my first few trading interviews, but just keep improving and you'll eventually find something good if you're smart and passionate enough.

a路径积分


Total Posts: 80
Joined: Dec 2014
 
Posted: 2015-11-11 05:13
> Depends if you're at a target school. Masters or even bachelors in stats with a lot of CS classes from Wharton, Harvard and Stanford frequently are recruited for top-tier hedge fund and bulge-bracket quant positions.

By the way, UPenn and Harvard do not have strong CS programs and you don't usually take non-Wharton classes at Penn because of constraints about elective coursework and credit.

EspressoLover


Total Posts: 337
Joined: Jan 2015
 
Posted: 2015-11-11 09:02
> By the way, UPenn and Harvard do not have strong CS programs

I don't disagree, but that's more relevant for a PhD than a Bachelors. General prestige school prestige matters much more than specific department strength at the undergraduate recruiting level. A CS major from Harvard is at least at order of magnitude more likely to get a front-office quant-ish job at Goldman or Citadel than a CS major from UIUC.

> [Y]ou don't usually take non-Wharton classes at Penn because of constraints about elective coursework and credit.

On this point, I think you're misinformed. Many Penn undergrads dual-major across the Wharton and Engineering schools (including Stats/CS) and graduate in four years.

Good questions outrank easy answers. -Paul Samuelson

a路径积分


Total Posts: 80
Joined: Dec 2014
 
Posted: 2015-11-11 10:42
> On this point, I think you're misinformed. Many Penn undergrads dual-major across the Wharton and Engineering schools (including Stats/CS) and graduate in four years.

Fair enough, I was under the impression you were referring to Wharton MBAs.

> A CS major from Harvard is at least at order of magnitude more likely to get a front-office quant-ish job at Goldman or Citadel than a CS major from UIUC.

However, this statement is only true if you're talking about a firm such as GOOG, DB, IMC, Jane St or HRT - not GS or Citadel.

Conditioned on being a CS college grad who is working at Citadel, the individual is around twice as likely to be from UIUC than from Harvard.

Granted that (1) the CS undergrad program is about 2-3x as large as Harvard's depending on the year in the last 3 years and (2) there's substantial self-selection - I don't know about the UIUC crowd but the Harvard crowd is much warmer towards Google and SV startups than Citadel or Goldman. But with these numbers an educated guess is that conditioned on being a Harvard CS college grad who is applying to Citadel, you are just nearly as likely to accepted as a UIUC CS grad who is applying to Citadel. Certainly not an order of magnitude in difference.

I estimate there's nearly 5-8x as many Harvard grads at Goldman than UIUC grads, however. This is largely driven by the fact that, conditioned on being at a front office job at Goldman, you're more likely to have studied finance than CS, or economics than CS. And economics is the most popular concentration at Harvard by far, whereas UIUC grads mostly choose to major in some form of engineering. Also, CS is still a rather small program at Harvard despite its recent growth (probably due to the pizza allure of CS50), these numbers are driven more by the AM bunch than CS.

> I don't disagree, but that's more relevant for a PhD than a Bachelors.

Inevitably, some high school student will find this thread on Google when he or she's deciding whether to go to Harvard or MIT and I have to clarify my points.

Now on the contrary, while Harvard doesn't have a strong CS program, I actually recommend studying CS at Harvard whether as an undergraduate or grad student as strongly as any of the top 10 'commonly ranked' programs if you want to make a career of CS.

Harvard's alumni network in academic CS is probably second only to MIT's, and the faculty to student ratio is several times better than, say, MIT's. Odyssey is years ahead of MIT's HPC resources and running test cases on a dedicated SB quad socket machine with 1.5 TB memory is something you likely won't get to do at MIT. (This still doesn't mean it's a strong program or a large department, though.) Now, if you're optimizing for sex life or living community though, you should probably consider MIT. And of course, to credit the valid part of EspressoLover's post, you should base college choice largely on the people and college on a whole, not your department.

goldorak


Total Posts: 1048
Joined: Nov 2004
 
Posted: 2015-11-11 11:27
Reminder to self: do not send your daughter studying in the US. There, they really think you actually learn things at school. Strange country.


If you are not living on the edge you are taking up too much space.

MCgeneratedname


Total Posts: 143
Joined: Dec 2013
 
Posted: 2015-11-11 12:45
oh... American. Then forget everything I said Wink
The situation here is different due to an over-supply of PhDs.

Reminder to self: Ask goldorak if he is willing to home school my children.

Lebowski


Total Posts: 70
Joined: Jun 2015
 
Posted: 2015-11-11 13:23
Not to thread jack, but since there are a couple folks who are directly involved in hiring on here and the conversation has shifted, I thought I'd ask. Do recruiters really care about the strength of departs so much for undergrads or does the strength of school carry more weight in your mind? I've always imagined that the strength of the department applied more to research and graduate studies.

I don't doubt that you're 2x as likely to be from UIUC than Harvard at Citadel or 8x as likely to be from Harvard at GS, but I also feel like once you attach a certain level of name brand to yourself--say like top 10 or top 25 US News undergrad rankings--what you do in an interview and where you actually want to work is probably a much stronger indicator of what job you'll get than the actual name of the school. That doesn't invalidate any of the above statements, it just merely adds that companies care about how you do in interviews and students have free will...not necessarily groundbreaking information.

goldorak


Total Posts: 1048
Joined: Nov 2004
 
Posted: 2015-11-11 15:23
@MC. I have a baby sitting business. I lazily called it the "the baby turtle traders". You are free to apply. Babies who already got a degree in economics or/and floaties for swimming are not considered as I tend to consider them as definitely lost to the dark side.

If you are not living on the edge you are taking up too much space.

sv507


Total Posts: 165
Joined: Aug 2010
 
Posted: 2015-11-11 20:40
@riane, there is a kaggle winton capital challenge.. maybe that would be something you could try your hand at and if you do well on the public leaderboard use it on your cv

mtsm


Total Posts: 220
Joined: Dec 2010
 
Posted: 2015-11-11 21:11
thanks for pointing this out. very interesting and how ridiculous their blurb on the kaggle page is...

mtsm


Total Posts: 220
Joined: Dec 2010
 
Posted: 2015-11-11 21:28
@goldorak: very funny you say that about the US, because it's so much more true about Europe in general

a路径积分


Total Posts: 80
Joined: Dec 2014
 
Posted: 2015-11-12 05:27
> Do recruiters really care about the strength of departs so much for undergrads or does the strength of school carry more weight in your mind? I've always imagined that the strength of the department applied more to research and graduate studies.

I cannot speak for all companies. It varies from shop to shop. But here's the way I see it.

Regarding campus recruiting, our recruiters have to maximize their ROI so they plot a roadmap that inevitably passes only through universities that have produced good interview hit rates in the past; this strongly biases the % hired to the 'top-ranked' (whatever this means) schools, since only these schools get to hear the name, but this doesn't mean that coming from a 'top-ranked' school brings any edge at all given that you have applied.

We don't distinguish between schools during the resume review.

One primary reason is that for a competitive position, we are looking for reasons not to reject rather than reasons to accept. We pick on the strongest signals and don't have the time to see which school you're from, and it's easy to meet sufficient condition to reject even without knowing which school you're from.

Now, people say that recruiting is a crapshoot and that my way of approach of filtering resumes is BS... This claim confounds two parts.

Empirically I have a 80+% accuracy at predicting which resumes don't clear round 1 interview even if I had passed all of them through. This part's easy.

Now the actual part that we find hard: What I, and presumably most recruiters, have no idea how to predict is how well a person will perform at the job given that he has passed through all the rounds and we have certain qualitative gauge on his interview performance. This doesn't mean that resume review is poor executed.

a路径积分


Total Posts: 80
Joined: Dec 2014
 
Posted: 2015-11-12 05:27
>Repeat post<

goldorak


Total Posts: 1048
Joined: Nov 2004
 
Posted: 2015-11-12 08:16
@mtsm: UK is not in Europe, and France is just and will always just be, well, France. Wink

If you are not living on the edge you are taking up too much space.

Lebowski


Total Posts: 70
Joined: Jun 2015
 
Posted: 2015-11-12 08:28
@a___ thanks for your insight and explaining a bit about your approach.

Follow up question: If the hard part is indeed assessing how well even your best applicants do in their given jobs, what is your position on internships? I'm done with this process at this point barring anything disastrous happening with background checks, etc, so this is just a curiousity. I've noticed that a lot of the small HFT shops ostensibly similar to yours are very reluctant to give internships, but what better indication of full time performance could exist than bringing prospective hires in for a few months and test driving them? I can't imagine why so few small HFT shops do this. It's not as if an intern is going to do anything competitive with your IP, if even see it. I imagine some of the smaller shops just aren't equipped to handle babysitting an intern, but even a lot of the elite mid sized shops like Teza, Headlands, and Jump don't offer internships. Another limitation is obviously that prop trading training takes a long time, but I think 10-12 weeks will still be sufficient for a company to assess a prospective hire's aptitudes and potential to learn the remaining material if given a full time offer. Would you care to comment on why so few HFT shops offer internships or discuss your position on the subject if that last step of associating good cumulative interview performance and full time job performance is indeed the tough part?

Lebowski


Total Posts: 70
Joined: Jun 2015
 
Posted: 2015-11-12 08:28
(Deleted repost)

a路径积分


Total Posts: 80
Joined: Dec 2014
 
Posted: 2015-11-12 12:31
Your hypothesis is mostly correct:

> I imagine some of the smaller shops just aren't equipped to handle babysitting an intern[.]

The only reason to take on an intern is an investment in a potential hire and HR. If you don't babysit an intern enough, then it defeats the purpose since the intern will go back telling his/her classmates that the experience was uninteresting. But if you babysit an intern, it becomes time-consuming and takes time off your short-term concerns. Smaller firms like Teza and Headlands have a lot of short-term concerns.

Where your hypothesis fails is in explaining Jump. Jump is in a different class. They're much larger so they have fewer short-term concerns. Jump used to offer internships in its Chicago office. They still do in London. In their case, it's just a matter of cost benefit analysis as they are operating in the pareto frontier of business objectives.
Previous Thread :: Next Thread 
Page 1 of 2Goto to page: [1], 2 Next