Probability of success due to chance alone?

Discussion in 'Risk Management' started by anesthesiaman, Oct 4, 2010.

  1. I would like to determine the probability of my success being due to chance alone (and thus not my skill).

    I started trading futures for the first time on May 25, 2010 (although I also know how to trade options and stock also - so I'm not an absolute beginner).

    On a 100k account I'm up a little over $7500. I took the month of Sept 2010 off to study for my anesthesiology oral boards - so this month should not count in the probability determination.

    The summer months had a high level of difficulty as even many pro hedge funds were down over the same period. This difficulty level ought to be taken into account, I think.

    I use Thinkorswim.

    I do not know my win/loss ratio nor the average loss or average gain, but intuitively I know that I win more than I lose and my average wins are greater than my average losses.

    The P value should be < or = to 0.05 but I do not know how to calculate this value. Anyone know how to do this probability calculation?

    Also, anyone know of a program that is compatible with Thinkorswim that would help me keep track of my win/loss ratio, max drawdown, average gain/loss, etc.?

    Thanks in advance,
    Anesthesiaman
     
  2. Perhaps I should mention also:

    I'm a pure discretionary trader. I usually don't hold longer than a few days to weeks. My trading system is based on what I see on the charts and hear in the news - not really a "system" per se. Sorry, I do not know how to describe it but it isn't important in the probability calculation.

    Statistics aren't my forte, but I guess to determine this calculation (i.e. probability of success trading futures over a 3-4 month period with return on trading capital of about 7.5%), one would have to assume a normal distribution curve.

    The question then becomes: what is the highest part of this normal distribution curve? I think this has to be known to determine the standard deviation from this mean.

    But if futures are a "zero sum game," then doesn't that mean the average person (i.e. the highest part of the curve) has a return over same 3-4 month period equal to 0 (i.e. person just breaks even)?

    Any help appreciated in this calculation.

    Thanks,
    Anesthesiaman

    ACCOUNT SUMMARY NET LIQ VALUE: $106,630.93
    ENDING CASH BALANCE $106,630.93
    STOCK BUYING POWER $202,011.86
    OPTION BUYING POWER $101,005.93
    DAYTRADING BUYING POWER $426,523.72
    TOTAL COMMISSIONS YTD $1,378.10
    UNSETTLED CASH $0.00
    AUTOTRADE PROGRAMS None
    CASH BALANCE START OF TODAY: $106,630.93
    DATE DESCRIPTION FEES AMOUNT BALANCE
    10/1/2010 12:00:00 AM Cash balance at the start of 01.10 $0.00 $0.00 $106,630.93
    10/2/2010 12:00:00 AM Cash balance at the start of 02.10 $0.00 $0.00 $106,630.93
    10/3/2010 12:00:00 AM Cash balance at the start of yesterday $0.00 $0.00 $106,630.93
    10/4/2010 12:00:00 AM Cash balance at the start of today $0.00 $0.00 $106,630.93

    ORDER HISTORY
    TIME PLACED SPREAD SIDE QTY SYMBOL SPC EXP STRIKE TYPE PRICE RULES STATUS

    TRADE HISTORY
    EXEC TIME SPREAD SIDE QTY SYMBOL SPC EXP STRIKE TYPE PRICE NET PRICE

    TRADE ACTIVITY BY SYMBOL

    EQUITIES NET MARKET VALUE:
    SYMBOL DESCRIPTION SHARES PRICE MKT VALUE

    OPTIONS NET MARKET VALUE:
    OPRA DESCRIPTION CONTRACTS PRICE MKT VALUE

    FUTURES
    SYMBOL DESCRIPTION CONTRACTS SPC TRADE PRICE MARK

    FUNDS NET MARKET VALUE:
    SYMBOL DESCRIPTION SHARES PRICE MKT VALUE

    BONDS NET MARKET VALUE:
    SYMBOL DESCRIPTION SHARES PRICE MKT VALUE

    YEAR TO DATE PROFIT & LOSS TOTAL P/L YTD: $7,561.0654
     
  3. Let's try to help you the easiest way.

    Step 1: I think you can count how may trades you made and how many wins you had.

    Your win rate is: w = number of winning trades/total trades

    Step 2: I think you can calculate the total amount of wins and the total amount of loser. Just add all winers and all losers, separately.

    You profit factor is: PF = sum of winners/sum of losers

    Step 3: There is a neat formula to calculate your average win to average loss ratio from those two numbers

    R = PF x (1-w)/w

    After you do your homework, we can continue.
     
  4. Foremost, you should be able to explain your edge(s) concisely, and, these strategies should reflect a known/established market behavior.

    Second, stats alone do not qualify whether or not your success is chance or not. Of course a few years of succesful trading matter, however, consistent and repeatable action is what constitutes a valid approach.

    Third, do you know your worst case risk scenario? Given you apply a consistent action whenever a suitable opportunity presents itself, what is your worst case drawdown?

    Mike
     
  5. Hello crank

    No, he is not requred to explain any edge to anyone. Just from the markets he trades, his profit factor, and win rate, a skilled trader (that excludes you) can decide whether he has an edge or not.
     
  6. Intradaybill,

    Thanks. That is extremely helpful. But sometimes I scale in/out of my trades - though not all the time. How do I account for this regarding calculating wins and losses? Also sometimes I hedge my positions using options.

    Anyway, perhaps I can still make this calculation on a rudimentary level since thinkorswim clearly organizes the futures I've traded and their total gain (or loss): if I trade copper 10 times and I'm down, I'll call that a losing trade. Vice versa if I'm up.

    I will do this calculation now and post results.

    Mike805,

    My edge is just skill - I wish I had an "edge," but the only edge I can think of is experience. However, I'd argue that experience - i.e. real experience over the years - is a definitive "edge" esp. for a discretionary trader who has no system of green/red arrows to follow.

    Also, I always know my worse case drawdown in every trade - that is what I think about throughout the trade duration.

    Anyone know of a program that could automate calculating my win/loss record (I assume by downloading trade statements from Thinkorswim and importing into this program)?

    Thank you,
    Anesthesiaman
     
  7. Welcome back intradaydouche. So am I off ignore, or on ignore? Since you can't seem to stick to something as simple as ignoring someone, and, since you read at a 3rd grade level, I'll re-write for you what I said: I didn't ask him to explain his edge. I said he should be able to. There is a difference.

    What's "crank"? Isn't that a drug?

    FYI, I hear lithium treats bi-polar disorder. Check it out, it might help your problem. Don't know what would help that megalomania though...

    In general intradaydouche, I keep being pleasantly suprised by how insecure you are. It's not much fun making fun of someone who is genuinely afraid of their own shortcomings, both physical and mental.

    Step one towards recovery:

    - Admit your shortcomings.

    Good luck douche,

    Please put me on ignore (again).

    Mike
     
  8. You may be doing something that is a classical discretionary edge. Several come to mind.

    - Dip buying, fading highs. Both a form of mean-reversion.
    - Using some form to tech. analysis to get a read on direction.
    - Looking for stops getting hit (a mild form of tape reading)
    - Fundamental/News catalysts.

    These are some discretionary edges. There are many more and in this case experience always plays an important role.

    Most systemic "edges" simply quantify behavior to a point where the rules are well defined and a computer can implement them. Just about anything falls under this category. I'm sure you could eventually program your buy/sell rules. Statistical arbitrage is a classical one that by most definitions trades the spread between two products. Basically mean reversion on a spread rather than on the security.

    Risk management is the most important though. Your approach, which needs to have positive expectancy regardless, may not give you good results without proper risk management.

    Excel is a teriffic program. I suggest you learn to work with Excel as it can do everything you're asking and quite a bit more.

    Mike
     
  9. Thank you Mike for explaining this to me. I always felt a little inferior for not being able to describe my "edge" as other than experience (basically failing the old adage: if you have to think about it, you don't have one). I didn't know that what you describe above would be considered "edges."

    Well you've really made my day, because I do all 4 that you describe above but I never knew they were considered "edges" - just something I learned to do. I am always trying to improve all of these edges you describe.

    Intradaybill,

    I've performed the calculations that you mentioned in your post. They are probably oversimplified since Thinkorswim just shows the futures or stock/option symbol I've traded and if that future or stock/option is a win or a loss - and from this info I calculated with Win Rate and PF. I included all trades even the hedges (I've actually significantly slowed down the use of hedges since most of the time it seems I'm right and don't need the hedge):

    Win rate = 16 / 25 = 0.64
    Win = $12558.60
    Loss = $4993.86
    Profit Factor = 12558.60 / 4993.86 = 2.51
    R = 2.51 x (1 - 0.64) / 0.64 = 1.41

    Now what do I do with R? What does this mean?

    Thanks,
    Anesthesiaman
     
    #10     Oct 4, 2010