Free Code Experiment

Discussion in 'Trading' started by Avid_Consumer, Jun 1, 2005.

  1. I would like these patterns to dissolve. Maybe others can get some use from them. trade em with your friends, sell em to someone, have fun
     
  2. Apply this to 1 min ES bars. Curves on all the indices to varying degrees.


    input:
    ETime(0),
    XTime(8);

    if closed(0) < opend(0) then
    if time = calctime(1600, (-1 * ETime)) then
    buy this bar;

    if time >= calctime(931, XTime) and time < calctime(1600, (-1 * ETime)) then
    sell this bar;
     
  3. Apply this to ES daily bars. Also curves on a few different indices to varying degrees.

    buy next bar l - 1 point limit;
    sellshort next bar h + 1 point limit;
     
    • fade.png
      File size:
      29.1 KB
      Views:
      374
  4. What does this mean???

    :confused:
     
  5. it's really simple, it's just going long 'ETime' minutes before 4pm EST, and selling the position 'XTime' minutes after 9:30am EST. You can drop the inputs and it's a little more intuitive:

    if time = 1600 and closed(0) < opend(0) then
    buy this bar;

    if time >= 931 and time < 1600 then
    sell this bar;

    works just as well, arguably better
     
  6. mmillar

    mmillar

    This is a well known 'nonsense' system. Most of the trades won't be hit (because they are limit orders) and, by definition, the ones that are hit will be the overall worst trades.
     
  7. since it's making less than 300 trades in 2.5 years, taking the tick slip on each side won't kill you. It's pretty insensitive to the slip. If you apply the concept to like a 15 min ES bar where the curve looks much better, then yes, the slip becomes insurmountable. The ideal frequency for ES fades is probably somewhere inbetween
     
  8. Not true. I have been running a similar system with quite a success.
     
  9. sgr

    sgr

    Interesting, Avid. I'm looking at your 'overnight' system.

    My understanding that @ES would be a non-adjusted continuous contract. So holding overnight over the contract changeover would show a nice gap, that could not actually be traded. One of the pitfalls of backtesting with continuous contracts.

    Steve
     
  10. true, i see what you mean. discount any profits made on rollover days. i haven't tallied it, but i doubt they're acting as any sort of crutch in this instance ..although it's a good point because i have seen situations where they account for an inordinate amount of profit. good point wrt an important vulnerability in all overnight work based on continuous contracts
     
    #10     Jun 1, 2005