Questions for ET QUANTS

Discussion in 'Educational Resources' started by Jeffrey A., Jul 3, 2006.

Thread Status:
Not open for further replies.
  1. This is a big open question for running statistics and data mining on various stock market data.

    Looking for a big nudge in possible recommendations of books or web links that would set up a novice in starting out for learning this process. Maybe available software too.

    For example,

    Looking over the WSJ, I have always been curious how to set up a very simple test procedure for their reporting of Money Flows.
    This breaks out two groups of 15 stocks+indices into Issue Gainers and Issue Losers with this formula: Money Flow figures are the daily volume of composite uptick trades minus the dollar value of downtick trades.

    Let's say I take(6/29 data) the biggest gainer stock Conoco +82.3 vs. largest loser stock AIG -58.0.

    I'm assuming you guys would run a whole bunch of statistical runs to test after X days later the largest gainer is X% percent higher/lower or vice versa. Maybe test for correlation/alpha and play the stock off an index. Sum the total Money Flow for the 15 stocks in each Gainer/Loser column as a ratio for a possible predictive aspect for market turns? Etc., etc.

    This is the type of stuff I would like to learn how to do. But I would like to get started in an example related to this type of stock market data and relevant worthwhile statistics of ZScore, etc.

    I need to start someplace. Suggestions?

  2. First I would think REAL hard about whether or not all the uptick transactions occurred at the ask and all the downtick at the bid. Validity, validity.
  3. You need to be aware that there are no quants here.

    There are pretender quants, and people who are quants in their dreams, also there are many fantasy quants and quasi quants, but NO, None, Zero real quants here at ET.

    You may however find some real quants (or quants in training) over at Paul Wilmott's site. If I were you I would look for Aaron Brown at;

    Sorry about Hypopotomus, periodically he stops his meds and breaks out of the facility he lives in. I am sure that soon they will find and sedate him, then we won't have to put up with his rambling disjointed pronouncements.

    Good luck,
  4. You seem to be very well informed...
  5. Somehow he knows very well everyone's background on ET! :D
  6. I am sorry.

    You must be a quant

    Please go ahead and answer the man's question

    Speaking just for myself, I am never to proud to learn something new.

    I'll take notes.

  7. How do you know this? Please give us a definite description of the illusive QUANT. As you may recall, some of the more adventurous pretender ET folks have gotten themselves stuck into this QUANT stuff already.

    As to Hypostomus Plectosomus, after many years you still have not grasped much of his wisdom. You simply must have invited his artful commentary with your pretension of knowing what is a QUANT.

    nononsense :cool:
  8. A "quant" is a person qualified to work in finance using quantitative methods.

    A "quant" working on Wall Street always possess a graduate degree (usually a Phd) in quantitative methods, math, computer science, physics or similar discipline. They are always graduates of one of the top 10 schools.

    That leaves everyone here "out" and by extension, unless someone shows up to claim that they are qualified, there are NO quants here.

    I realize that my description is oversimplified, but that is how it has to be for the majority of folks that frequent this site.

    Wonderful talking with you again.


  10. ozzy


    Don't you people sleep?

    Here's my defenition of a quant.

    A skýnny human beýng usually wýth no athletýc abýlýty of eýther pakýstaný, chýnese or mýddle eastern herýtage whose parents have braýnwashed hým to belýeve that the only way to succeed ýn lýfe ýs to get a PhD and then keep slavýng away crunchýng numbers.

    I was surrounded by these panzýes ýn school. Dýme a dozen.
    #10     Jul 4, 2006
Thread Status:
Not open for further replies.