Is my trading system (long-term) profitable?

Discussion in 'Journals' started by PlannedTrades, Jun 23, 2013.

Is this a long-term profitable system with a +EV ?

  1. Yes, I would keep trading it!

    0 vote(s)
  2. No, you're gambling!

    1 vote(s)
  1. Do you folks think my trading results suggest that my strategy is long-term profitable (taking EV, and risk management into account - I never risk more than 2% per trade, and don't use any Martingale nonsense, but it DOES result in a large drawdown because of negative consecutive runs). You can copy/paste my results into Excel to run formulas on it.

    Here are the results from approximately 9 months of REAL money trading in the currency market. I only traded the EUR/USD and EUR/GBP (this system can be traded on a smaller time frame as well - provided you are an active trader, but it works better for someone who has a full-time job, and doesn't need to log in very often. Otherwise it will results in way too many trading alerts / signals throughout the year... though I'm sure it can be programmed using algorithms too, but I am not a programmer):

    01] profit = 53.30 pips
    02] loss = -5.20 pips
    03] loss = -9.70 pips
    04] loss = -4.10 pips
    05] loss = -2.90 pips
    06] loss = -6.10 pips
    07] loss = -6.20 pips
    08] loss = -15.10 pips
    09] profit = 10.10 pips
    10] loss = -8.90 pips
    11] loss = -6.50 pips
    12] loss = -12.20 pips
    13] loss = -5.60 pips
    14] profit = 16.10 pips
    15] loss = -6.10 pips
    16] loss = -3.10 pips
    17] profit = 3.20 pips
    18] loss = -14.30 pips
    19] loss = -7.80 pips
    20] profit = 152.50 pips
    21] loss = -11.20 pips
    22] profit = 84.60 pips
    23] loss = -16.10 pips
    24] loss = -19.80 pips
    25] loss = -16.30 pips
    26] profit = 147.50 pips
    27] profit = 20.30 pips
    28] profit = 69.20 pips
    29] loss = -26.20 pips
    30] profit = 34.20 pips
    31] profit = 4.20 pips
    32] profit = 83.90 pips
    33] loss = -49.80 pips
    34] profit = 15.60 pips
    35] loss = -24.40 pips
    36] profit = 12.20 pips
    37] profit = 4.10 pips
    38] loss = -5.90 pips
    39] profit = 17.80 pips
    40] loss = -26.10 pips
    41] loss = -0.80 pips
    42] loss = -7.30 pips
    43] loss = -17.30 pips
    44] loss = -20.40 pips
    45] loss = -1.70 pips
  2. is this a joke or.....?

    How in the world can you expect someone ELSE to do the WORK for you? I like how you add, you can copy this to excel to run you are giving us all the privilege of confirming if your system is worth a damn or not.

    The only tip I can give you, 45 is not a large enough sample. While its true that samples of 30 or greater do resemble a normal distribution, its my rule of thumb that for trading samples of less than a 100 are useless.

    Wait wait, i mean

    your system is super profitable! Go trade live now !

  3. I've already run the numbers. They show me an +8.26 pip EV on each trade, but "negative runs" can, and will happen and I needed some insight into the long-term viability of my system. Also, I'm not a math major.

    trading samples of less than a 100 are useless.

    Oh no... not that age ol' "sample size" excuse again! Warren Buffett has a sample size of less than 100 investments, btw.

    Math major? To be clear - I learned that crap in school. Not university. While this is not high school level, it was required knowledge to pass university permission where I lived (German Abitur).

    You will live and die with statistics when you do trading. Saying "hey, I am clueless" will not help.

    There are some good books to read.

    "Fooled by Randomness" and "Trading Systems - a new approach to system development and portfolio optimizations". Both deal with the statistical side and how to analyze results data and make sense out of it.

    These are investments, not trades. And they cover a great many years.

    Another hint - regardless of sample size, your TIMEFRAME is too small to make long term conclusions. You should at least cover 3-5 years to know how stable your results are. Maybe you just got a lucky year? The behavior of various markets changes in cycles, and 9 months is not enough to even cover a yearly cycle (and yes, in many markets those exist). For example I can make a 9 month sample without a winter - and a great many things are different in winter (energy consumption and payments for example for those countries that have winter at the moment).

    I could show you quite some diagrams of trading approaches that fail to work or work very well in a 1-2 year timeframe. 9 Months now is full in the QE cycle - no real bust. VERY distorted market. No long term viability possible. Backtest on 5 - 10 years. Results should be similar to indicate stability.
  5. Guys, I hear you about 9 months possibly being a luck of the draw... THAT'S why I was asking for input (to see if I just got lucky, or what). Also, to establish a 3 to 5 year track record one must extrapolate from real market results from day 1 anyway - not to mention back-testing. So technically speaking... day 1 should provide as much relevant information as day 1,000 - on an individual basis - i.e., Buffett's first couple of successful investments are really what hooked his initial investors on, but they would hardly be regarded as gamblers (even back in the day).

    For the record, I DID perform back-testing on the two currency pairs as well (including commissions), and it yielded similar results as the ones with real money, but of course running some algorithms would provide a more speedy, and accurate analysis because there's only so many calculations I can do by hand. So I haven't looked at other currency pairs yet because these two do the job, and have tight spreads. If I had an algorithm scanning 22 currency pairs in the fx market on a 24/5 basis then my system would generate > 1,000 trade entry signals (on the time frame I am currently trading) - though more than half would be "duplicates" in a sense because they cover similar base currencies. Taking this down to a lower time frame would mean > 30,000+ sample trades... there's no way I can backtest that without an algorithm. I'm primarily a swing trader anyway.

    Anyway, I guess I'll just keep trading this. The way I see it I never lose more than 2% (and in most cases less than this) so if I don't make anything going forward, no biggy... least I'm not losing much. My system reinvests profits to compound results as it consistently risks a fixed % of capital on every trade.

    Btw, in case it isn't obvious, this is a trend-following system, and I am convinced it works in the stock, and futures markets too - probably even BETTER since individual stocks trend better, and not as much leverage available so one is less capable of trading too large.

    I think I might hire a programmer to do some back-testing, on all of these markets, but the usual fear lurks... that if this system truly delivers the types of results I have been getting then it'll stop working after the programmer steals it, lol. He could sell it to a hedge fund, and after multiple sales everyone will have the source code, and it will become inefficient. Probably just an irrational fear, but I'm human.
  6. What are you talking about ?

    Do you mean his portfolio is less than 100 investments? Or in his career he made less than 100?

    The first point is true, he probably has 40-50 stocks in his portfolio. But his investment strategy (even tho its fundamentally focused instead of TA) has been tested over thousands and thousands of trades--thats why he is worth what he is.

    If you think your 9 month TA system is on the same level as Warren Buffet, then you dont need our need a psychiatrists help...