Money management - What is the secret? -
1 2 3 4

thread: Money management - What is the secret?

  1. #21
    95070Can you post a user's guide, please. The Help menu does not work. Even after google translation it is fairly perplexing.

    What are quote iniziali uguali (initial shares equal) and quote iniziali diverse (several initial shares??) ?
    Why Can I input the winrate twice? (attesi is WR correct?) Also it is not clear I will enter 50 instead of 0.5

    At some point I could select the dropdown win/loss. You can't change As soon as you decide on the value. It is fairly an issue although I guess it is a feature to never alter the past.

    When an input signal is wrong you show a popup I valori at ingresso non sono corretti (The input values #8203;#8203;are wrong ) but that does not tell which is wrong and why it is wrong. (btw people hate popups)

    what's to be set at the pillar Quota (stocks )?

    I set 50 percent and RR =1, I get 0 bet dimensions...

    You are using a proprietary format for saving the result. Is it possible to import from a csv file?

  2. #22
    The results are fully in line with my previous articles. With a winrate=50% and RR=1 that your MM starts gambling 7.37% of the starting capital I set to 1000.

    If we bet 7.37% each time and get 10 losers the last result is 464.96. With your MM the result is 317.56. The risk increases to 14.68% =gt; worse DD

    If we bet 7.37% each time and get 10 winners that the last result is 2036.66. With your MM the result is 1624.59. The risk decreases to 2.84% =gt; significantly less profits

    Yet if we alternate triumph and loss 5 times the result is better: 1003.97 rather than 973.

    Clearly the Gambler's Fallacy in the office. Such MM can't work as long as there isn't any serial correlation involving the transaction results. In the long term you will get the same overall result but with a higher volatility of the equity curve (=higher risk of ruin)

  3. #23
    The results are fully in accord with my prior articles. With a winrate=50 percent and RR=1 your MM starts gambling 7.37percent of their starting capital that I set to 1000. If we bet 7.37% each time and receive 10 losers the last result is 464.96. With your MM the result is 317.56. The risk increases to 14.68percent =gt; worse DD If we bet 7.37% each time and get 10 winners the last result is 2036.66. With your MM the result is 1624.59. The risk decreases to 2.84percent =gt; less profits Nevertheless when we alternate win and loss 5 times the result is better: 1003.97 rather than 973. Certainly...
    Great thank you

  4. #24
    Thank you but you wrong.... Youdont need to consider only the first bet....in the worse scenarioyou lose 1000....
    I just posted the consequence of your own software...


  5. #25
    This is the Gambler's Fallacy: from 100 trades there is no reason that there'll be 50 winners and 50 winners.

    No I didn't attempt to input more than 10 trades because it is a pain in the... neck to input the values. It if was possible to upload an excel document it would make sense see the difference with a regular MM and to create a lot of random datasets. To me the outcome is apparent the same result with volatility in the equity curve.

    Since you have the algorithm, why don't you generate a Monté Carlo simulation and also post the graph?

  6. #26
    Senior Member Tataylo's Avatar
    435
    quote I think it's more simple than what it sounds, just I would like to make a question: Assuming you have a egy with R1:1 and that following long backtest you supposed to possess a winning ratio around 50 percent .ok? Now you would like to use it risking all the capital ( 1000 euro) having an aggresive money management that you would like to accomplish the max profits. Can you tell me exactly what is going to be the results following 100 trades, supposing that you were right so that you will have 50 failure and 50 winner? This is only math, we are not including all the others factors (disperse,...
    If you've got a win rate of 50% in 1:1 RR, then your profit factor is 1. Hence in the event that you start with 1000 EUR, and we ignore costs, you will (generally ) end with 1000 EUR, no matter how many transactions you make, and no matter what place sizing MM you use. No difference is made by aggressive or conservative MM.

    Win rate = 50% and RR = 1:1 is effectively the same as a (fair) coin toss.

    @PipMeUp: please correct me if I am wrong.

  7. #27
    Is this egy base on'alternative' condition?
    Still attempting to know this egy.
    Thanks for sharing, anyhow

  8. #28
    You want to use it risking all the capital ( 1000 euro) having an aggresive money management that you want to accomplish the max profits.
    No. You don't wish to maximize your profits. If it were the case you'd bet 100 because this is the wager which maximize your profit and a series of winner maximizes the amount of their profits. Of course the first loser loses it all. Everything you need maximizing is the expectancy of growth of your account.

    Can you tell me what is going to be the outcomes after 100 trades, assuming that you were right so that you will have 50 failure and 50 winner?
    No. The maximum likelihood is 50 winners really but you could have any amount of winners between 0 and 100. The probabilities will distribute around the gaussian bell curve.

    Do generate a simulation of your MM. You can accomplish this with excel and compare it to a predetermined fractional MM. You will see on your own. Your MM driven equity will always ramble then return and cross the others. The only way to get your MM function is to statistically confirm that the trading egy's outcomes are correlated.

  9. #29
    Senior Member Tataylo's Avatar
    435
    quote Thank you Hanover. If we ignore, you are completely right. Now, did you download the app I posted? Please can you attempt to set a egy with a win rate 50% at 1:1 RR?? Would you tell me please what will be the result? If you want to put a 1:1 RR, since I explained, Bear in mind you have to put 2 in Quote iniziali uguali's box. Thank you
    Sorry, no I have not downloaded your software.

    I gave my views earlier in the thread:

    1. To be profitable you need positive expectancy, i.e. a profit variable gt; 1. No MM (position sizing/staking system) can change expectancy; I don't think there's any sort of secret. Hence it's not a priority of mine to come up with a'superior' MM.

    2. Risk management is of paramount importance, to preserve capital. The deeper an account goes into drawdown, the more difficult it becomes to recuperate (explained here). More competitive MMs increase risk of drawdown.

    3. Everything else being equal, RR and win rate operate in approximate reverse proportion to each other (such as in blackjack: greater payoff stakes = commensurately less frequent wins). I believe that markets can offer the possibility of'long tails', and there are exceptions that are possible, as I tried to explain here.

    But IMHO the very best chance of finding any sort of robust advantage is to gain an understanding of the hows and whys of price movement. Hence that is my focus. I replied to your question only because I was pretty confident that I knew the response.

  10. #30
    Junior Member elenalopwz_'s Avatar
    19
    This is exactely the Gambler's Fallacy: from 100 trades there's absolutely no reason that there will be 50 winners and 50 winners. No I did not attempt to input over 10 trades since it is a pain in the... neck to input the values. It was possible to upload an excel document it would make sense to create a bunch of random datasets and see the difference using a regular MM. The outcome is clear anyway: the same effect with much more volatility in the equity curve. Since you've got the algorithm, why not generate a Monté Carlo simulation...
    This is precisely perfect. If you do a Monte Carlo simulation of 10,000 trades using a 50% win rate, you will find a supply of win% approximately 50%, but every individual simulation run will give a response different than 5000 wins, using a small number of exceptions. Any trading system that adjusts trade size assuming that a return to average or expected probability ignores the property of this machine. Of the 10,000 trades you might have 9000 losers and 1000 winners, but the likelihood of the trade being a winner is 50%.

    If you want to check the likelihood of a positive or adverse results from a particular win% and R(and other variables ), there are calculation tools on the internet that function as a sanity test.

    This http://www.forexscamalerts.com/risk-...own-calculator tool uses Monte Carlo simulation.

    This one relies on a http://2ndskiesforex.com/risk-of-ruin-calculator/.

  •