direct statistical trading a "clearly" defined approach by NTW31

Discussion in 'Strategy Building' started by nukethewhales31, Dec 31, 2008.

  1. Interestingly enough, this is very close to what my setup is. I'm using range bars (I create them on my own using a custom program from a tick datafeed). By definition, a new bar in a range chart is started when the last tick breaks out of the (fixed) range of the previous bar. So now I'm at the open (first tick) of the new bar, with a history of range bars (and the ticks that comprise them) behind me, and I want to find the statistical sweet spot in terms of target/stop (assuming my entry is at the bar open tick (limit entry) or the tick that follows it (market entry). I can program anything, but I need the formulas :)

    Thanks again!

    D.
     
    #431     May 27, 2009
  2. decent week hoping for more movement this usd/MXN trade hasnt unloaded yet
     
    #432     May 29, 2009
  3. statistical sweet spot huh... i dont really understand what you mean when your talking about this rangebar stuff maybe you could throw me a picture. as for formulas im not much of a programmer so i dunno.. i do my stuff by hand slowly and painstakingly.
     
    #433     May 29, 2009
  4. usd/mxn trade im sure you can see why by now.
     
    #434     May 29, 2009
  5. I can send you an example by email. For one thing, I am not sure your method can be applied to what I'm trying to do, so I'll explain more in the email.

    D.
     
    #435     May 29, 2009
  6. lpn

    lpn

    Ok, here are some questions :)

    Initially you used MA(20) to determine the direction in which you can take a trade, but after that you started using tick data. What do you use to determine the direction for tick data?

    It looks like you are using 1h and 3h bars. Smaller intervals seem to have the advantage that you can close position faster, but the potential profit (several pips in your examples) is greatly reduced by the spread. Longer intervals would have the potential for larger possible profits (and losses of course), so the effect of the spread would be less significant. Have you tried different time frames and if so, any idea what may be the optimal setting?

    More questions to come...
     
    #436     May 29, 2009
  7. there is a way to measure everything.. you can always normalize as well.. i havent got your email hope a can help.
     
    #437     Jun 1, 2009


  8. its the same the tick data just makes the entrance and exits more precise.
    I feel strongly that as you narrow done to smaller frames you loose efficiency.. you may be making good trade but eeking out small profits per trade. the ME is smaller... but as you raise timeframes stops get tighter and are in more meaningful places.. profits are at better larger areas... spread is nonsense .. traders feel less stress in the trade... in the higher frames knowing the long flow is important and when things stack like the interest rate differential is in the direction of your hopeful trade you get paid to trade and in some cases can nullify a loss not moving a stop because of the interest rate differential and the duration it took to get there... anyways probabilities are higher in highe time frames and efficiency in the trade is increased... as you go lower these things start to fade away. i like to trade anywhere from the 3hour - the weekly depending... for short term trading 3hour /4 hour is best efficiency i think.
     
    #438     Jun 1, 2009
  9. I've heard that the interest rate difference or "carry direction" is a fundamental driver of pair price direction. I've had my hands full just studying technicals so I never looked hard at currency fundamentals.

    Do you find that the carry direction is a significant contributor to your trade bias? For instance, your example USD/MXN short made sense because being short this pair pays significant interest.
     
    #439     Jun 1, 2009
  10. its very significant but income stream doesnt drive a pair as much as eco data... trading in direction of rate is like buying a house on a loan so you can rent it out for a profit or cover the cost of that loan.. its a more investment way to trade a pair.. for instance usd/mx since wat 02.. some of those years the return based on a 2 times leverage is 18% annual yield.... pays everyday and if wat you make is added to the trade over time .... you are basically compounding the yield daily just like using revenue stream from one investment house to pay for another... but you look at results USD/PESO is a pair that normally stalls.. (just like otherpairs) the Sum of the moves are random. and because of that sometimes its within an area where capital is appreciated and times when its not. anyways my answer is no income streams do not drive price as much as other economic data.

    my usd/mxn trade i used a turn to carry style for ... where i was trading in the direction of the rate because thats where i thought it was going to go.. and while i was waiting ... i was recieving an income stream.. its a very sexy way to trade.
     
    #440     Jun 1, 2009