Serious question about automated quant trading for the person of average intelligence

Discussion in 'Trading' started by Kovacs, Dec 14, 2009.

  1. Kovacs

    Kovacs

    So I visited www.wilmott.com, expecting more rigorous, empirical discussions on trading. To my disappointment, I got exactly that. Most of it went over my head. I understood what they were talking about at a high-level, but the details baffled me.

    I'm not innumerate, but my math skills are merely average. My trading now just consists of executing around support and resistance, and this other world seems completely alien.

    What is the best way for a guy like me to get into this stuff so I can start building models for automated trading?
     
  2. What do you mean by "average math skills"? That has a radically different meaning in Wilmott World than in, say, ET world.
     
  3. Kovacs

    Kovacs

    Fair enough.

    By average, I mean that, if I work at it, I can understand the math behind all the popular tools of quant trading. Covariance, expected values, stochastic processes, autoregression, etc.

    Most of it isn't that complicated. The symbology just looks intimidating, but once you start breaking it down it seems straightforward.

    I guess what I mean is--I have no math talent.
     
  4. It looks like they have a Student forum on there. Why not throw up a post on that and see what replies you get?
    Do you have specific strategies you are looking to implement or just trying to get a 30000' view?
    I'm an engineer, not a quant, but I'd approach it the same way I would an engineering problem: Pick a task you wish to accomplish and then identify the knowledge you are lacking to complete that task. Just skimming that website for a few minutes makes it appear that you could spend months with your nose buried in a math textbook (not necessarily a bad thing, but I'm assuming you want to focus on trading). I find that forums are most helpful if you require answers to specific questions, and I'm sure that one is no different.
     
  5. Consider bypassing the parts that are inductively based.

    If you find anything based on science (deductive processes), then drill down a little bit.
     
  6. Wilmott is a different ballpark. This is where the real meat is if you want to learn.

    ET is like Community college arts program in Georgia.

    Wilmott is more like The MIT Advanced Physics program.

    If you do not enjoy Math for the sake of it and are capable of working differential calculus like a person adds his grocery bill I would suggest not driving yourself crazy.

    You have to like Math if you want to master it.
     
  7. byteme

    byteme

    Mark Joshi, a well-known author and academic in this field has a bunch of advice for wannabe quants on his website. More specifically, he has a recommended reading list here:

    http://www.markjoshi.com/RecommendedBooks.html

    Essentially, here's what you'll need to do:

    1) Learn some Real Analysis (e.g. Bryant, Rudin etc.)

    2) The some Probability (e.g. Resnick, Chung etc.)

    3) Perhaps brush up on some Linear Algegra.

    4) Then you're ready for some intro Fin Math books (e.g. Joshi, Neftci etc.) These will introduce you to Stochastic Calculus etc.

    5) Oh, you'll need to learn some C++ or perhaps Python to code your models in.

    You can self-study all of the above or take a Masters (Mathematical Finance, Financial Mathematics, Financial Engineering, Quantitative Finance - depending on your slant) or one of the specialist certificates e.g. Certificate in Quantitative Finance www.cfq.com

    None of the above will prepare you for automating anything though - just the math and coding the models.

    Good luck.
     
  8. don't waste your time on any of this rubbish.

    Wilmott is full of geeks who are only good at pricing arbitrage free positions. Pricing not trading.

    put it this way - perhaps their 'leader' - Wilmott himself - is a failed trader and makes money by printing endless formulas on books and arranging courses. He himself knows the pitfalls of current mathematical thinking in finance.

    'quant' has become the biggest cliche buz word in the world.

    some 'quants' are pure pricing monkeys, other 'quants' are purely programmers, some 'quants' are also risk managers, while some other 'quants' build trading models.

    the last group is no different to a normal trader. unless you have no rule in your trading and completly trade on your emotions, then you are trading a model. the 'quant' word makes it exciting and sells well to clients.
     
  9. There's often a broad confusion between "what do I need in order to start trading" and "what do I need in order to work at a bank creating derivatives". This post might be an example.
     
  10. I'm with byteme, but then again, I am one of those evil Wilmott peeps :)...
     
    #10     Dec 15, 2009