Hi, Thanks to everyone for contributing. I've removed one more fundamental error in my backtesting calculations as I spotted in my previous post! Note that my numbers are approximate and are meant to give me an idea of order of magnitude and comparison between the different intervals. There still could be some fundamental errors in my calculations For gbpusd 2004, 1 min data from forexite. I tested 1) Use midnight as a pivot everyday. If price goes down, buy, if price goes up, sell. 2) Use midnight as a pivot everyday. Buy and sell if price goes up or down. 3) Use start of year as pivot. Buy and sell if price goes up or down. Overall for the same interval spacing (0.0100 to 0.0020), 2) was better than 1) for profitability/Max Drawdown and 3) was better than 2). For 2004 gbpusd, method 3) I found that 0.0070 gave the best result. I also wrote a program to backtest trading at certain percentages. Excel was way too slow with having to recalculate the unrealised p/l for nav for every single trade! Trades:608 pips realised:40,390 Unrealised:-22,950 Starting with $10000 balance @ 0.10% nav per trade : 12110.69 unrealised : -1237.48 nav: 10873.21 % profit : 8.7% Max DD Bal : $ 9716.55 % DD : 2.83% Profit/DD :3.08 When scaled to 0.30% Bal: 16914.62 Unrealised:-4312.01 NAV:12602.61 % profit :26.0% Max DD Bal : $ 9135.46 % DD : 8.65% Profit / DD: 3.01 I then tested 2003 gbpusd @ 0.0070 intervals Trade:399 pips:25,340 unreal pips:-44,415 Bal:$11288.62 Unrealised:$-2293.19 NAV:$8995.43 %profit:-10.0% MaxDD Bal: $9162.61 % DD: 8.37% Profit / DD:-1.20 Profit / DD for: 0.0100 : -1.15 0.0080 : -1.00 0.0070 : -1.20 0.0060 : -0.95 0.0050 : -0.95 0.0040 : -0.91 0.0030 : -1.17 0.0020 : -0.89 So even though in the first half of 2003, gpbusd moved up and down, that last uptrend in nov/dec produced the 'nav' loss at the end of the year. Max nav for 0.0070 was at 7/11/03 with nav of 10624.41. So my conclusion is that trading one currency using a pure fixed grid is probably not a good idea I believe that other currencies & rules need to be added to a pure fixed grid to increase profitability. Best regards Alan
I have been waiting for something to hit me. So I re-read this Journal. I didn't realize how many good posts were in here. I even took some extra vitamin B12.... I think someone might have said this...but it hit me anyways.. What if there was simply an entry made at 0.02% trade ticket in all 15 pairs long and short. Two Accounts in Oanda...Forget charts and drawing lines. then... Each hour or two hours simply go to the platform and close the profitable entries (look under trade tab and sort by clicking the header at the top) and take the profit and re-enter them. Oanda makes this simple and manageable with their platform. On the pairs with no profitable trades simply enter in another trade. The side effect of this, is that you can observe which times of day/night yield the most profit.... What am I missing? I would need to wake up each hour, 24 hours a day, maybe each three or four hours at the slow times (or maybe each hour during slow times and each three or four hours at the faster times?). Wouldn't this set the trades at the actual volatility levels and no need to line them up as they are TIME INCREMENTS. The trade ticket takes care of the martingale (just transfer money back and forth between the long and short accounts to keep them balanced). Slower instruments and times would have tighter increments...and vice versa... This all seems so easy compared to what we have been discussing? Michael B. P.S. Assuming this time increment concept, is something to try...I found a neat free indicator with is own feed to tell us the 2 hour speed on the EUR/USD. Now if I can talk them into the other 14 pair? There is a chat area on the platform if anyone wants to talk live with me (I am there 24 hrs a day, and can be grouchy if you wake me). You got to download it though, and maybe some of you prefer not to. http://66.199.246.146/forexunited.htm
You know this "Time Increment Concept" could be altered to be "one-sided closer times" when the long/short account is improving its unrealized.
Added some more positions....couldn't wait for limit orders per the spreadsheet rules to fill (I do not know if the positions will give any other "average down" opportunities) I added 4000 units of EUR/HUF and 6000 units of AUD/JPY...this brought those average trade line prices closer to the current price. When this system gives the opportunity to "trade out" a profit for a loss position, then it's the problem of how to get back "in" to get the daily interest back up to the level one desires. You folks that are following this, as long as the system is trading at the prices it is now it is a no-brainer, its just to collect interest and wait. Todays Trades (Friday)
dw, Would your spreadsheet work for any pair? or does it assume anything that would be specific to the EUR/USD? And what would be you opinion using the dw start to find balance then going forward with "Time Increments" ? Then assuming you like "Time Increments", would you recommend that I post a spreadsheet here tracking the hourly range of each of the 14 instruments, to discover when to lengthen and shorten the time increments? Michael B.
You nailled it right here Electric, a LOT of labour for the amount you made. You would need to trade with 50000$ of capital to make this system worth trying. It could become good with automated trading though.
Yeah max... Sad... Oh well, I will continue to trade System #2 and post here, unless some of you brainchilds has a low labor method of trading System #1 and can convince me to forward test it here. My requirements are to make 25% a year after payment for labor. System #1 is in the can.... So in conclusion System #1 could be the proof a trader needs to deduce that trading is not gambling, but it is much easier to gamble. I will try to keep the additions to this Journal at a minimum now for those of you following System #2, so that I do not clog your inboxes. I am busy "dialing up" the gambling meter in System #2 to yield $6.36 a day to net $191.00 per month in labor, it is almost there and needs just 1000 or 2000 more EUR/HUF units. Then I will investigate daily injections of this $6.36 a day to average down on the lagging pairs in this three pair sytem.. Michael B. P.S. I hope some of you have learned one thing, if not many things from this Journal. I feel satisfied, was it good for you?