2
$\begingroup$

I intend to thoroughly prepare for an internship that I will start in a couple of months, and therefore wanted to clarify what topics I need to study and some recommended references for them.

The description for the internship is as follows: the intern will be a part of research for quant trading strategies, will be given a live project and will learn how to develop and backtest trading strategies for equities, forex, commodities, etc.

As for my existing background, I once did a project in mathematical finance on option pricing in Markov modulated markets (which I think may not be related a whole lot to the internship) and I also learned a bit of C and Python programming languages earlier. I have absolutely no prior experience in the area of quant trading.

I'd be grateful for any advice regarding what to study, and for suggestions on what book(s) or papers (e.g. which among the quant papers on sites like arxiv.org) to read, or if there are any websites/software that simulate the kind of work I will need to undertake.

Thanks in advance!

|improve this question
$\endgroup$
  • 1
    $\begingroup$ From a finance point of view is too difficult to answer, as it depends on the kind of trading your firm does (why not ask them?). But in general, improving your skills in programming and working with data is a good idea. Python is more and more used in this kind of work. $\endgroup$ – Alex C Jan 18 '16 at 15:14
3
$\begingroup$

I did a similar internship at a quant equity shop and based on my experience, I think there are a few common aspects to such work, which you can try and work on, to have a more productive internship experience:

1) handle on programming language - check with the firm what programming language they'd want you to work in. And in case, you have never worked in it, spend some time getting used to it. Some firms can be flexible on this but some want you to work in specific language only (e.g in my case I learnt SAS, since the firm mostly used SAS)

2) broader understanding of 'quant trading' and what you are after - incase you have had a coursework on this, revise that. Or pick a book to get a general idea about the discipline. e.g. Inside the black box by Rishi Narang, could be a useful read

3) experience with data - its good if you have had some prior experience with the dataset you'd be using over the internship, but in your case, since you don't know what you'd be working on - equity, forex, commodities or something else, don't worry about it

4) experience with backtesting - you could practice by running a backtest of a simple well documented strategy (such as value/ momentum/ Fama- French factors etc). This will give you a better hold on some of the things that you should look out for, such as avoiding 'look ahead' bias, selecting relevant securities etc.

Also, you should check with your guide at the company, if there are any relevant research papers/ background reading material, that they would like you to read in the meantime.

All the best.

|improve this answer
$\endgroup$
  • $\begingroup$ Thanks a lot! There's just one point I want to clarify: how exactly should I go about getting experience with backtesting? I'm guessing this would involve 2 parts: reading the theory behind backtesting a particular strategy (HOW to backtest), and then obtaining data (which may just be some dummy dataset for practice purposes) on the basis of which that strategy can be tested (WHAT to test). If possible could you suggest some reference where I can find theory on the "how to backtest" part and some source where I can get a dummy dataset? $\endgroup$ – u23 Jan 18 '16 at 22:44
  • 1
    $\begingroup$ Read Meb Faber's paper "A Quantitative Approach To Tactical Asset Allocation" and try to backtest that using data from Yahoo Finance and either an Excel spreadsheet or a python program. $\endgroup$ – Alex C Jan 18 '16 at 23:57
  • $\begingroup$ It'll depend on what data you have access to - if you can get CRSP data (basically price data for all US stocks), you can backtest 'momentum' strategy and compare results with those published on Kenneth French's website (it also has an explanation on how to do it, else u can refer the research paper). For 'value' or Fama-French factors (esp SMB or HML), you also need access to Compustat data. You wont get exact results - but if you do it carefully enough and use similar dataset, you'll get very close. $\endgroup$ – Uditg_ucla Jan 19 '16 at 20:37
  • $\begingroup$ While it definitely depends on what kinds of things your firm does, if you have no prior experience I'd definitely make sure to at least become conversant at a broad level in the different order types and execution algorithms, which should help you gain familiarity with basics of market microstructure (which is a really cool animal in and of itself). Two books I'd recommend are Johnson's "Algorithmic Trading & DMA" and Tomasini & Jackle's "Trading Systems," in addition to any other relevant "soft" reading that interests you. $\endgroup$ – Jacob Amos Feb 16 '16 at 23:58

Not the answer you're looking for? Browse other questions tagged or ask your own question.