Home > Technical Topics > Strategy Development > Simple System for Beginners Monthly Results

Simple System for Beginners Monthly Results

  1. Date Entry Exit P/L(less commission) Trade

    1 556.50 556.50 5.00- Long
    2 558.50 564.20 565.00 Long
    3 559.70 559.20 45.00 Short
    7 566.80 566.90 5.00 Long
    8 563.30 562.00 125.00 Short
    9 564.20 566.00 175.00 Long
    10 567.70 569.80 205.00 Long
    13 573.10 572.50 65.00- Long
    14 570.20 571.40 125.00- Short
    15 568.10 567.80 35.00- Short
    16 574.20 573.70 55.00- Long
    17 572.70 572.30 35.00 Short
    20 572.30 571.00 135.00- Long
    21 575.20 577.00 175.00 Long
    22 567.30 565.20 205.00 Short
    24 568.90 566.30 265.00- Long
    27 561.10 558.60 245.00 Short
    28 562.20 565.80 355.00 Long
    29 569.70 571.00 130.00 Long
    30 572.70 572.10 65.00- Long

    NET $1,570.00
     
  2.  
  3. This is not a comment on how good/bad the system is. Simply the EL code and the results I obtained. My data is from CME's official T&S files, that I extract with a perl script. This data used is back adjusted. The latest contract obviously isn't adjusted, but still deviates from the results posted by AnomalyResearch.

    Let me know if anyone disagrees with my data / code. Although the data for september / earlier is adjusted, and *would* have different entry / exit prices, the *net results in a given day are guaranteed to be the same*.

    Commission is $4.80 r/t. Slippage is $5 to get in, $5 to get out. In effect, I am betting I would, on average get a hit of one tick on half my transactions. This is very close to what I actually see trading mechanical systems on ER2.

    Enjoy! :)





    EL Code:

    Code:
    {********************************************************************************
    * Tested working with ts2000i. Needs modification for TS 7/8 etc.
    * Use on 1 min bars (Not necessary - I just coded this on 1 min bar
    ********************************************************************************}
    input:
    	{All times in CST}
    	ETime(1300),
    	CtxToTrade(1),
    	MaxEntriesInADay(1), LastEntryTime(1459), EODExitTime(1500);
    
    var:
    	BuyStop(0), SellStop(0);
    		
    {*******************************************************************************}
    if(Date <> Date[1]) then
    begin
    	BuyStop = 0;
    	SellStop = 0;
    end;
    
    {***Entry***********************************************************************}
    if(Time = ETime) then
    begin
    	{Round down to nearest valid tick - this needs to be changed for
    	 contracts that doesn't trade in 0.1 tick interval, of course}
    	BuyStop = floor((OpenD(0) * 1.0033) / 0.1 + 0.5) * 0.1;
    	SellStop = floor((OpenD(0) * 0.9967) / 0.1 + 0.5) * 0.1;
    end;
    
    if(MarketPosition = 0 and Time <= LastEntryTime 
    and EntriesToday(Date) < MaxEntriesInADay and BuyStop <> 0 
    and SellStop <> 0) then
    begin
    	Buy ("LE") CtxToTrade contracts next bar at BuyStop stop;
    	Sell ("SE") CtxToTrade contracts next bar at SellStop stop;
    end;
    
    {***Stop************************************************************************}
    if(MarketPosition = 1) then
    	ExitLong("LX") CtxToTrade contracts total next bar at SellStop stop
    else if(MarketPosition = -1) then
    	ExitShort("SX") CtxToTrade contracts total next bar at BuyStop stop;
    
    {***EOD Exit********************************************************************}
    if(Time = EODExitTime) then
    begin
    	if(marketposition = 1) then
    		ExitLong ("EODLX") next bar at market
    	else if(marketposition = -1) then
    		ExitShort ("EODSX") next bar at market;
    end;
    
    SetExitOnClose;
    
    

    ===============

    =====Monthly=====
    May ($1,256.00)
    June $1,249.20
    July ($556.40)
    August $2,654.40
    September 784.00
    October 405.20 (till today; day not over yet)
    ===============

    =====Trades=====
    See attached spreadsheet
    ===============
     
  4. Nice job. I enjoyed looking at the spreadsheet.
     
  5. Are you accounting for stops?? Looking at the spreadsheet it looks like it is ignoring the rules put forth. For example, the day of the largest drawdown (in your spreadsheet) 5/12/2004 you state that the short signal was entered at 531.50 so the stop would be ruffly in the 535 area (long signal) yet your exit was way past the stop at 544.90 resulting in a huge loss. Using the parameters in the system the largest loss should be capped to around 3.8 pts max (not accounting for slippage) this figure is the total range between the two signals respectively. Please explain those results, maybe I am missing something here, thanks. :)
     
  6. There is no stop after you enter the position. The stops are only to enter. On 05/12, Open was 543.10. So the sell stop was at about 541.30. But we enter the position only after 1300 CST. So the entry was 532 ish. After entry, there is no stop. There is no cap for losses - none. According to the rules anyways..

    Review the system rules, and it would be clear.

    BTW, I didn't put together the spreadsheet by hand - it's what TradeStation spits out. But it should be alright, mostly.

    I am not aware of any bugs in my coding.

    If anyone haven't guessed already, it's a very good base for a workable system. I have modified it a bit, which goes live with a few contracts next month. But I wouldn't trade this as is. What you have here is a set of rules. A set of rules is not a strategy. If you modify this rules to something you can live with + put in money management, and make sure it complements your other systems, that is something you can call a "system" :)
     
  7. Interested to see the results for October. Looks like the last two days were losses.
     
  8. Perhaps you need to review the rules.

    "The remaining unexecuted stop is your stop loss."

    :)
     
  9. Sigh... that's what you get posting at dead of night... I need more sleep. You are right, and my code posted earlier in this thread reflects that. My code & posted results are right, but my post on 10/28 night re:stops is wrong.

    My apologies.
     
  10. Do you have the results for October yet? Thanks.
     
  11. October net : $2558.80

    10/29 ($114.80)
    10/28 ($144.80)
    10/27 $355.20
    10/26 $305.20
    10/25 ($114.80)
    10/22 $315.20
    10/21 $425.20
    10/20 $125.20
    10/19 $225.20
    10/18 $125.20
    10/15 ($264.80)
    10/14 $175.20
    10/13 ($34.80)
    10/12 $35.20
    10/11 $0.00
    10/08 $265.20
    10/07 $285.20
    10/06 $295.20
    10/05 ($104.80)
    10/04 $0.00
    10/01 $405.20

    Have fun! :)
     
  12. Impressive! Thank you.
     
  13. Me too :D I don't know where I came up with the 3.8 max loss at...damn I gotta quit writing these early a.m. posts! :p Results look real good for October, thanks for the info!
     
  14. Hey guys,

    PLEASE post questions and comments in the main thread otherwise this thread becomes very long and anyone wanting to track the monthly results gets bogged down reading alot of questions and answers. Thanks.
     
  15. Deducting $5.00 commission per trade and no slippage.
     
  16. hey ! this trading system/rule is beautifully simple !

    can anyone who has back-tested it advise as to where and how often it could go wrong ?

    and why is the entry stop placed 0.33% above or below the open price ? is there anything particular about 0.33% ?

    tks n rgds
     
  17. Check the post two above yours...

    Hey guys,

    PLEASE post questions and comments in the main thread otherwise this thread becomes very long and anyone wanting to track the monthly results gets bogged down reading alot of questions and answers. Thanks.
     
  18. enjoy...
     
  19. Mogul,
    Could you post the TS code used to produce those results, I'd like to play around with it. If I find anything of interest, I'll let you know. Most of the postings of code on these two threads don't have it quite right it seems.

    Thanks very much
     
  20. I used DarthSidious original code, and modified to fit TS8.

    Might also be interesting to note that a basic stop-loss of say 3pts enhances the results and caps the risk (per trade)...trailing stop not as useful. True, since the system inherently brackets the price range, a stop loss is built in, BUT those days when the market trades above the initial stop-entry point, you could have a larger gap down to your exit...

    let's see if it stops working in 2005....

    Code:
    {********************************************************************************
    * Tested working with ts2000i. Modified for use with TS8
    * Use on 1 min bars (Not necessary - I just coded this on 1 min bar
    * Source credits: DarthSidious
    ********************************************************************************}
    input:
    {All times in CST}
    ETime(1300),
    CtxToTrade(1),
    MaxEntriesInADay(1), LastEntryTime(1459), EODExitTime(1500);
    
    var:
    BuyStop(0), SellStop(0);
    
    {*******************************************************************************}
    if(Date <> Date[1]) then
    begin
    BuyStop = 0;
    SellStop = 0;
    end;
    
    {***Entry***********************************************************************}
    if(Time = ETime) then
    begin
    {Round down to nearest valid tick - this needs to be changed for
    contracts that doesn't trade in 0.1 tick interval, of course}
    BuyStop = floor((OpenD(0) * 1.0033) / 0.1 + 0.5) * 0.1;
    SellStop = floor((OpenD(0) * 0.9967) / 0.1 + 0.5) * 0.1;
    end;
    
    if(MarketPosition = 0 and Time <= LastEntryTime 
    and EntriesToday(Date) < MaxEntriesInADay and BuyStop <> 0 
    and SellStop <> 0) then
    begin
    Buy ("LE") CtxToTrade contracts next bar at BuyStop stop;
    SellShort ("SE") CtxToTrade contracts next bar at SellStop stop;
    end;
    
    {***Stop************************************************************************}
    if(MarketPosition = 1) then
    Sell("LX") CtxToTrade contracts total next bar at SellStop stop
    else if(MarketPosition = -1) then
    BuyToCover("SX") CtxToTrade contracts total next bar at BuyStop stop;
    
    {***EOD Exit********************************************************************}
    if(Time = EODExitTime) then
    begin
    if(marketposition = 1) then
    Sell ("EODLX") next bar at market
    else if(marketposition = -1) then
    BuyToCover ("EODSX") next bar at market;
    end;
    
    SetExitOnClose;
    
    
     
  21. Thank you mogul! I was really wanting the TS8 code! This system really seems to work over the long run and has a lot of promise. Take care.
     
  22. Just curious on the reasons why it doesnt seem to work on any of the other emini contracts? Take care.
     
  23. Great job, mogul. I wanted to see something like that a long time ago, but somehow could not get anyone to post it and since I have no data for ER2 I could not do it myself.

    Now, have you checked how the system worked for ES, NQ2, and YM over at least the last 2 year period... Could you possibly post the results for these eminis too? That would be really appreciated.
     
  24. as requested...
     
  25. and....
     
  26. and...
     
  27. Many thanks for your work on other eminis...
     
  28. Fascinating. The clear conclusion is of course that this is a system designed for trending markets.

    The interesting question is are YM and ES simply in rangebound enviroment or are these instruments SO efficient that they retrace most of the gains one way or the other and wipe away the edge?
    Is this why so many traders lose money trading these particular instruments?

    Finally - would using a 30 pt (YM) 3 pt (ES) stop make these systems profitable? or will they just churn you?

    Let us know mogul if you can.
     
  29. Thanks Mogul.

    This is interesting.

    I was very curious to see the other instrument results because all of these instruments have similair behaviour with the post 14:00 eastern time moves.

    From the results of the other instruments, on the surface level, one might think that YM,ES and NQ don't move much post 14:00 eastern time.
    There are large moves in all of these instruments post 14:00 on many occassions.

    This system is tweaked to ER2.

    Now, why the 1.0033 and .9967 on ER2 and what would work on the other instruments ?
     
  30. imo the "clear conclusion" is the coeffecients are tweaked to ER2 volatility.
     
  31. Y no doubt. "trending markets" implies ER2 which is mcuh more trendy than either ES or YM which are the retrace kings of all tradables.

    Ultimatley -= does this show yet another example of failure of systemic trading which redcuces to nothing more than an optimized algorithm on a particular data set?
     
  32. I don't know that this shows a failure of systematic trading. Actually this shows a profitable systematic trading method.

    Yes, it is an optimized algorithm of current ER2 data.

    The key is, and probably is for all systematic trading is knowing when to stop while you're ahead.

    or moving to a new system

    or combining systems

    or getting out at the first "new low" drawdown..........

    I suppose this could go on and on.

    This system is in the red as of my calculations to date. (this month)
     
  33. and just like that..........


    The system is in the black month to date.

    + 4.6 according to my fills and calculations.


    Nothing like a good trend day to make up for things.

    (or kill things, my countertrend system had it's worst day ever yesterday, 11/19)
     
  34. How is that different from discretionary trading?

    God knows I dont; want start another flame war with everyone about systems trading.

    My deeper point is that "methodic" trading ir isolating a susccessful setup vis a via a particular tradable is really what this comes down to.

    Systemic trading "mumifies" the methodic approach and therefore carries the seeds of its own dustruction
     
  35. quite a difference from my numbers...what kind of slippage are you getting\incorporating?
     
  36. hi there

    i'm still not convinced about applying this system to markets other than ER2.

    also not convinced that the signal stop (0.33% from open) at 1 pm will work in other markets. why only 0.33% ??

    but its good to know that your live trades of this system have generated profits.

    have you tweaked the system to suit your trading style ?
    if yes, then would you mind divulging the details ?

    rgds
    ----

     
  37. Hello Mogul.

    My stats are probably a bit too perfect.
    I have been accounting for .10 slippage on the in and the out about 50% of the time.

    The results that you posted with the software generated backtest are probably the worst case scenario (which is a good thing) where you had .10 slippage on every trade on the in and the out.

    What are your current results ?
     
  38. FWIW I used the same parameters on the S&P pit contract in 1989. There is no magic number of which I am aware. Varying breakout parameters yield a spectrum of results. A percentage of ATR or previous daily range also "works". Waiting for a consolidation or narrowing of the daily ranges as a filter has improved results in the past. Trailing stops tend to degrade the performance. Tight stops tend to degrade the performance. Waiting for a retrace after the breakout tends to degrade the performance. Unless you have some exceptional discretionary ability, taking early profits tends to degrade the performance. In the final analysis, if enough trending days appear and you catch and ride a significant percentage of them you may produce a profit. The ER2 is currently and recently a nicely trending market. Instruments which trend nicely come and go. In my opinion, when the market you are trading lacks trending movements you will get break even results from the trades and the loss of execution and transaction costs.
     
  39. Thanks
     
  40. November results attached
     
  41. A peculiarity of Tradestation caused the very first trade of the month to be incorrect. Thanks to Manni0805 and Macaw for pointing out the mistake. The repost is attached.
     
  42. Hi Anomaly,

    I don't get the same results in November. Have you checked the individual trades to see if they match the chart?

    For example, on November 1, I get the following:

    Opening Price: 584.80
    Buy Price: 586.80
    Sell Price: 582.90

    On a 5 min chart (as you are using for your backtester), a trade should have triggered at the 13:35 bar (CST) at the buy stop and held until the open (I'm assuming) of the 15:00 bar exiting at 587.00 for a net profit of 2 ticks minus commission. You don't show a trade for that day.

    On November 2, I get the following:

    Opening Price: 588.50
    Buy Price: 590.50
    Sell Price: 586.60

    A trade should have triggered at 14:05 when that bar dipped below the short sell trigger but at the price of 586.50 and held until the open of the 15:00 bar at 584.40 for a net profit of 21 ticks minus commission.

    These are the results I get from my backtesting software and that I have verified by hand. Can you please let me know if there is an error in my system's logic or my understanding of the plan?


    Thanks. :)
     
  43. By hand I see price opening at 590.90 on Nov. 2 at 14:00 east coast time, so it would have activated an immediate buy over your 590.5 buy price. This would have actually resulted in a nice loss for this day.
     
  44. You are right on that one. I am not sure what happened there. I think I spent too much time looking at charts that morning.

    Did you get a trade trigger on November 1 at 13:35 CST for a 2 tick profit?
     
  45. hallo @ all,

    hope that AR will post the results for dezember.

    best regardsd Manfred
     
  46. hallo again,

    why does AR not post the results anymore?

    rg M.
     
  47. By my calculation this system made about 7 points in December.

    January has been a pretty wild ride so far.
     
  48. $700 per emini russell a month is nice system. This is one of the very few I have seen that can actually work over time.
     
  49. word on street is AR began fading all system signals once he found out every guy on ET was trading it. What ya think of that?

    Mr. Zen
     
  50. LOL!

    He would have done all right this week. :D
     
  51. you would be very brave to fade this

    also that means you are fading the market mechanism that makes a system like this work...again, very brave

    I know your joking, just wanted to throw that in
     
  52. See early in the thread for the EL code used to calculate this, and the commission / slippage assumptions:

    For the last 500 days.. (to recap)
    ------------------------------------------
    Period Net Profit
    1/05 $743.60
    12/04 $74.00
    11/04 ($836.00)
    10/04 $2,608.80
    9/04 $754.00
    8/04 $2,644.40
    7/04 ($566.40)
    6/04 $1,249.20
    5/04 ($1,256.00)
    4/04 $1,754.00
    3/04 $2,509.60
    2/04 $2,153.60
    1/04 ($751.20)
    12/03 $2,564.00
    11/03 $1,663.60
    10/03 $1,219.60
    9/03 $3,078.80
    8/03 $469.20
    7/03 $909.20
    6/03 $1,569.20
    5/03 $553.60
    4/03 $1,024.00
    3/03 $219.20

    Current Month & last month daily
    ------------------------------------------
    1/27 $115.20
    1/26 $255.20
    1/25 ($74.80)
    1/24 $415.20
    1/21 $5.20
    1/20 ($414.80)
    1/19 $485.20
    1/18 $185.20
    1/14 $265.20
    1/13 ($524.80)
    1/12 ($704.80)
    1/11 ($14.80)
    1/10 ($514.80)
    1/07 $425.20
    1/06 ($394.80)
    1/05 $645.20
    1/04 $465.20
    1/03 $125.20
    12/31 ($454.80)
    12/30 ($44.80)
    12/29 $0.00
    12/28 $185.20
    12/27 $135.20
    12/23 $0.00
    12/22 $45.20
    12/21 $215.20
    12/20 ($134.80)
    12/17 ($114.80)
    12/16 $85.20
    12/15 $185.20
    12/14 $185.20
    12/13 $115.20
    12/10 ($234.80)
    12/09 ($434.80)
    12/08 ($14.80)
    12/07 $615.20
    12/06 $75.20
    12/03 ($184.80)
    12/02 ($294.80)
    12/01 $145.20

    Nothing to sneeze at, I would say. It's holding up after taking some beating early this month.
     
  53. thank you for posting the results...

    best regards
     
  54. Yes, thanks for posting the results. I wonder what happened to AR?:confused:
     
  55. Just wondering if anyone has the results for Feb?
     
  56. Me too, out of curiousity...
     
  57. Actually till today, for the last 250 days.
     
  58. Good job, thanks a lot...
     
  59. March 05 results:

    -$1320
     
  60. April 05 results:

    + $ 1180
     
  61. May 05 results

    - $60
     
  62. June 2005 results

    -$1055
     
  63. July 2005 results

    + $2800


    <img src="http://www.elitetrader.com/vb/attachment.php?s=&postid=803288">
     
  64. August 2005 results

    -$860


    <img src="http://www.elitetrader.com/vb/attachment.php?s=&postid=832241">
     
  65. Good job, AR. Thanks for keeping us informed.
     
  66. September 2005 results

    + $1645

    <img src="http://www.elitetrader.com/vb/attachment.php?s=&postid=856596">
     
  67. This will be the last post to this thread.
    The graph shows the monthly profit or loss since the system was posted September 2004.

    <img src="http://www.elitetrader.com/vb/attachment.php?s=&postid=856601">
     
  68. AnomalyResearch,

    Thanks for the peak into your system. Impressive results, indeed. I am looking forward to your straight-forward and simple methodologies in the future, if you choose to post them again.

    Thanks again.

    Michael B.
     
  69. Does anyone have the results of this system for 2007 that they'd be willing to post here (or PM)? I am curious as to how it performed during this time. I would appreciate any feedback.

    These are the rules: http://www.elitetrader.com/vb/showthread.php?s=&threadid=37520

    Some TS code:
    http://www.elitetrader.com/vb/showthread.php?s=&threadid=37520&perpage=6&pagenumber=27

    http://www.elitetrader.com/vb/showthread.php?s=&threadid=39031&perpage=6&pagenumber=1

    I know that the rules as posted by AnomalyResearch are correct, but since I'm not a programmer, I can't vouch entirely that the code is as it should be. I'm not saying that it isn't but rather I do not know.

    Thanks for any help, and thanks again to AnomalyResearch for the contribution.
     
  70. Bump.

    Anyone?