Coin-Toss Trading Strategy | Page 5 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

TEB: Random Entry - Coin Toss Great with R.R.R. 105 replies

Weighted Coin Toss Trading 118 replies

Toss a coin? 9 replies

coin toss entry question 24 replies

Attachments: Coin-Toss Trading Strategy
Exit Attachments

Coin-Toss Trading Strategy

  • #81
  • Nov 15, 2008 7:59pm Nov 15, 2008 7:59pm
LOL!

Quoting billbss
The coin toss strategy can be made simpler by just writing out a check to your broker everyday and not messing with tossing the coin.
  • #82
  • Jul 21, 2009 7:35pm Jul 21, 2009 7:35pm
If you toss a coin to choose trade direction there is possibility that you are always on the losing side of the trade. If you only trade long (or short) then it is almost impossible to lose every trade. With a tight SL, a win can make up 2 or more losses.
  • #83
  • Aug 8, 2009 12:17pm Aug 8, 2009 12:17pm
RonaldRaygun:

Did you come up with a money management scheme for this? Would like to discuss MM with you.
  • #84
  • Aug 8, 2009 3:39pm Aug 8, 2009 3:39pm
Quoting gbppip
RonaldRaygun:

Did you come up with a money management scheme for this? Would like to discuss MM with you.
I have a a method and can make money on a coin toss, but it's not mechanical.
As for talking, I don't see why not. PM me, or e-mail or skype, or open a ticket (so I can read it at work) at forex.offers4u.biz/support
  • #85
  • Nov 24, 2009 8:37pm Nov 24, 2009 8:37pm
Dear Ronald,

I know this is an old thread. If you are still intrested, I may be able to shed some light for you. I have done A LOT of work with random entries. Previous posters are right, in that no money management system can overcome negative expectancy. Over time, it is a loser. Over time.... think about that concept... what does it mean? Start wrapping your head around this concept "The Probability of Probability". When you start down this road, you will begin to discover that a random entry system is quite workable given a set number N of trades. Infinity will not work, but it is quite possible to give yourself a reasonable chance of success over a few thousand trades. I cannot adequatley express all of the mathematics or concepts here... nor do I want to. I will leave you with a simple formula which could help you down this road shoud you choose to travel it:

N= log(1-DC) / log(1-P)

N= Number of trials neccessary for an event to occur
DC= Degrees of Certainty for the event to occur (must be less than 1)
P= the probability of the event occuring

I wish you all the best in your trading.....
  • #86
  • Nov 25, 2009 1:19am Nov 25, 2009 1:19am
You are right Petra in that I cannot expect simple lot sizing MM to turn a negative expectancy system into a positive expectancy one.

My current iteration for the infinite coin toss model is to look at various price levels and determine the likelihood that the price will reverse at a specific level. If you have a look at the "Dynamic Colors?" thread, you'll see that I did pull together a heatmap of sorts with plenty of help. Interpreting that heatmap has been a little tricky to get into an EA form, but I more or less know how to trade with that sort of information.

I have since moved on from trading as simply an odds-based game.
  • #87
  • Nov 25, 2009 1:40pm Nov 25, 2009 1:40pm
This might be a stupid question but how do I use the ATR and the EMA to gauge volatility?

It says "10 day exponential moving average of the Average True Range". I applied ATR(14) to my chart but it just shows 0.4821 . Where does the EMA come into the mix?
  • #88
  • Nov 25, 2009 9:41pm Nov 25, 2009 9:41pm
regular ATR is the simple moving average of the true range.

You basically need the exponential moving average of the true range.
  • #89
  • Nov 25, 2009 9:55pm Nov 25, 2009 9:55pm
Hehe, I wouldn't be surprised if this worked.

But you need an dynamic SL and a way to trail your wins....
Chandelier exit perhaps?

There is no secret that the big dogs who make serious cash have heaps of small wins and losses and some big wins which makes the money...

I reckon the problem here comes when trying to actually do it. Follow the trades through... Mentally hard to implement....

But as the above poster said, maybe it is better to go in one direction all the time... Or just look for higher highs and higher lows... oh hold on that's a system.... forget it..
  • #90
  • Nov 28, 2009 11:11pm Nov 28, 2009 11:11pm
I wouldn't either. Use your best strat and see how far it gets you long term. Or use a random strat enough times till you get a positive equity curve and before you know it you will have something you can sell to the general public for a profit.
  • #91
  • Jul 13, 2010 9:21am Jul 13, 2010 9:21am
Quoting Ronald Raygun
IMPROVE money management. Not to replace a trading system.
Holy bump batman I seem to be bumping old threads of late.

I never really looked into this on a figures basis but I did today.

I have a system I have used for the last year or so, it has a 35% win rate with a RR just under 1:5 (spread eats into what would have been a 1:5 without spread), it has over the last year achieved an unaltered expectancy of 2.19 .

However I have a MM system which has increased the expectancy of the system significantly and also has the desirable effect smoothing out the equity curve.

I'll point out that it's not martingale, and a it is a safe MM strategy to use because the maximum loss it can take at any time is 1.56%


I'll demonstrate the results.

Assume we have a system similar my original one, 33% win rate (easier to use this) with a RR of 1:5. It has an expectancy 2.46.


I rolled a 3 sided dice a hundred times, and I got:

Landed on one 31 times.

Landed on two 33 times.

Landed on three 36 times.



Assume normal MM:

1) Assume that rolling a one meant winning and rolling a two or three meant a loss.

31wins, 69losses.

5/1*31/69 = expectancy 2.24

2) Assume that rolling a two meant winning and rolling a one or three meant a loss.

33wins, 67losses.

5/1*33/67 = expectancy 2.46

3) Assume that rolling a three meant winning and rolling a one or two meant a loss.

36 wins, 64 losses.

5/1*36/64 = expectancy 2.81

Average expectancy was as expected 2.50 (orignal before results was around that)



HOWEVER, by applying different MM by altering %R on each trade, and only "accepting" losses after a certain sequence is not achieved and maximum loss achieved (By hedging or otherwise)


Assume altered MM:

1) 33wins, 3losses: 42.5/1.56*31/3 = 2.81

2) 33wins, 2 losses: 42.5/1.56*33/2 = 4.49

3) 36 wins, 1loss: 42.5/1.56*36/1 = 9.80



Change of

1) 2.24 to 2.81

2) 2.46 to 4.49

3) 2.81 to 9.80

Average expectancy from 2.50 to 5.70, +3.20 in expectancy.

My own increase in expectancy is not as large but is pretty considerable and worth doing.This MM idea can be used with practically any high RR system but works well with system designed specifically for it.



Purpose of my post is this, there are a lot smarter people than me here in FF so I ask them this : Are my wrong here ? Are my being fooled by randomness ?



As far as I can see the usage of a different MM system can increase expectancy.

I wonder can it turn negative expectancy systems into positive expectancy . @Ronald Raygun I assume you have already achieved this ?

I'll have a look at that tomorrow.

Regards,
Jig
  • #92
  • Jul 13, 2010 10:09am Jul 13, 2010 10:09am
You mention 'sequence' so I'm guessing multiple trades open. So I must ask, what's your floating loss?

As for achieving positive expectancy through MM, no not even close. I eventually expanded to trade management.
  • #93
  • Jul 17, 2010 11:11am Jul 17, 2010 11:11am
Quoting Ronald Raygun
You mention 'sequence' so I'm guessing multiple trades open. So I must ask, what's your floating loss?
Nope, no multiple trades open. Nothing special you have already explained the logic yourself here below, it's just the opposite:

Quoting Ronald Raygun
Well, consider this.

If you know that the likelihood of a trade is a winner, then it's probably better to size your lots smaller. Depending on how often the loser happens, will allows you to at least 'take the hit' and survive to hit another streak of wins.

An alternative is to start 3rd from highest risk, then keep decreasing risk until you hit a loss, then go back to 3rd highest risk.
Quoting Ronald Raygun
As for achieving positive expectancy through MM, no not even close. I eventually expanded to trade management.
MM can increase already positive expectancy dramatically. As for positive expectancy through MM alone, I have not achieved it either. Despite numerous tries/tests over the last two days over a long enough time period you always lose everything unless you have positive expectancy to begin with. Now it seems obvious why this is so, but it didn't seem so at the time

Regards,
Jig
  • #94
  • Edited 12:29pm Jul 17, 2010 12:13pm | Edited 12:29pm
Not to belabor the obvious, dollars aren't in the formula for expectancy anyway.

expectancy = (win % * average win) - (lose % * average loss)

where average win and average loss is expressed in pips, or ticks, or cents (per share of stock)

In our case it's pips and it's up to the trader how many dollars/pip to trade. That's where your MM strategy comes into play.
You are in a maze of twisty little passages, all alike.
  • #95
  • Nov 15, 2010 9:18am Nov 15, 2010 9:18am
...but this thread looks like a good place

On your profile you say

=============================
1) Toss a coin: Heads long, Tails short
2) Manage the trade as you see fit. If you think the trade is bad, exit.
3) When you exit a trade, go back to step 1.

If you can double your account in less than 50 flips, I consider you a good enough trader to go live. I personally average between 20 and 30 flips, and I do it once a week, on the weekends, trading simulated data.
=============================================

would you be able to post a trade history of doing this on a demo account (or real account)

This question is with all sincerity.....not suggesting at all any doubt....actually recently I directed someone to your profile.

I think actually seeing this would be very helpful for lots of people.

Thks!!

PGPB
  • #96
  • Edited Nov 16, 2010 2:03pm Nov 15, 2010 3:26pm | Edited Nov 16, 2010 2:03pm
IF event A=prob of next single toss (always=0.5),
and event B= prob of 2 consecutive heads in 2tosses = 0.25,
then if my first throw is a head, which event (A=independent single event or B=2tosses as a single event) is considered as the overall event prob for my next throw on head?
  • #97
  • Edited 1:29am Nov 18, 2010 1:18am | Edited 1:29am
this isn't directly related to ronald's bio challenge which as i read it means "open a position based on a 50/50 and then deal with whatever you are given and massage your sequence of trades into profit... however you can".

i wanted to play around with a / the basic coin toss 'game', and as i have been doing a bit of php i produced this last night. if you have wampserver or xamp etc i'm sure you will be able to run it fine. this could likely be done in javascript, but js pisses me off and i don't like it so php and a php localhost server is what i have done here.

if you can, and want, to play with this then have at it. it's functional (as i see it) over form. don't think there are any bugs.

Inserted Code
<html><head><title>Nubcake's Coin-flip</title></head><body text=#ffffff bgcolor=#000000>

<?
  $totalflips = 0;

  $balance = 0;
  $betsize = 1;
  $spread = 1;
  $positiontype = 'None';
  $price = 0;
  $openprice = 0;
  
  $action = 0;  //0 = none, 1 = flip, 2 = close open position

  if ( isset ( $_GET['balance'] ) ) {
    $totalflips = $_GET['totalflips'];
    $balance = $_GET['balance'];
    $betsize = $_GET['betsize'];
    $spread = $_GET['spread'];
    $positiontype = $_GET['positiontype'];
    $price = $_GET['price'];
    $openprice = $_GET['openprice'];    
    $action = $_GET['action'];
  }
  
  function f_writelink($actiontype, $actiondescription) {
    global $totalflips, $balance, $betsize, $spread, $positiontype, $price, $openprice, $action;    
    echo '<a href="coinflip.php?totalflips=' . $totalflips . '&balance=' . $balance . '&betsize=' . $betsize . '&spread=' . $spread . '&positiontype=' . $positiontype . '&price=' . $price . '&openprice=' . $openprice . '&action=' . $actiontype . '">' . $actiondescription . '</a>';
  }
  
  if ($action == 1) {  //flip coin
    $totalflips++;
    if ( rand(0, 1) == 0 ) {
      //up
      $price++;
      echo 'HEADS<p>';
    } else {
      //down
      $price--;
      echo 'TAILS<p>';
    }
  }
  echo 'Total Flips = ' . $totalflips . '<p>';
  
  echo 'Price = ' . $price . '<p>';
  
  if ($action == 2) {  //close open position
    if ($positiontype == 'Long') {  //close long
      $balance += (($price - $openprice) - $spread) * $betsize;
    }
    if ($positiontype == 'Short') {  //close short
      $balance += (($openprice - $price) - $spread) * $betsize;
    }
    $positiontype = 'None';
  }

  echo 'Balance = ' . $balance . '<p>';
  $t = ($price - $openprice) * $betsize;
  if ($positiontype == 'Short') { $t *= -1; }
  $t -= ($spread * $betsize);
  if ($positiontype != 'None') {
    echo 'Open Price = ' . $openprice . '<p>';
    if ($positiontype == 'Long') { echo 'Long '; }
    if ($positiontype == 'Short') { echo 'Short '; }
    echo 'floating p/l = ';
    if ($t < 0) {
      echo '<font color=#ff9999>' . $t . '</font><p>';
    } else {
      echo $t . '<p>';
    }
    $t += $balance;
    echo 'Equity = ' . $t;    
  }

  echo '<hr><p>';
  
  if ($positiontype == 'None') { 
    echo '<form action="coinflip.php" method="GET">'
    . '<input type=text name="betsize" value="' . $betsize . '" /> bet size<p>'
    . '<input type=text name="spread" value="' . $spread . '" /> spread<p>'
    . '<select name="positiontype">'
    . '<option value="Long">Heads</option>'
    . '<option value="Short">Tails</option>'
    . '</select><p>'
    . '<input type="submit" value="Open Position" />'
    . '<input type=hidden name="totalflips" value="' . $totalflips . '" />'
    . '<input type=hidden name="balance" value="' . $balance . '" />'
    . '<input type=hidden name="price" value="' . $price . '" />'
    . '<input type=hidden name="openprice" value="' . $price . '" />'
    . '<input type=hidden name="action" value="0" />'
    . '</form>';
    
    echo '<p><hr><p>';
    f_writelink(1, 'Flip Coin');
  } else {
    f_writelink(1, 'Flip Coin');
    echo ' / ';
    f_writelink(2, 'Close Position');
  }
  
?>
<hr>
<p>
Heads = Long<p>
Tails = Short<p>
<hr>
<p>
<a href="coinflip.php">Start Over!</a>

</body></html>

edit : oh, and yes it's extremely basic to cheat this... but what's the point if you do? it's just as easy to flip a real coin and write your results and cheat yourself that way. same difference. you are the only one keeping score in the end.

:nerd:
Congratulations ScalaFX for coming out of the closet!
  • #98
  • Nov 18, 2010 1:34am Nov 18, 2010 1:34am
Totally off on my own tangent on this one:

I look at the Coin Flip scenario as a paratrooper being dropped off in No Man's Land.

Either you know how to trade the market you are dropped off in, or you don't.

Bottom line: assess the market conditions, and exit immediately if the trade isn't one YOU would have been in.

What I'm saying is, just because you are put into a trade that looks bad from the start, doesn't mean that you can "manage your way into it being profitable." Exit immediately, if that's the best course of action in the moment.
  • #99
  • Nov 18, 2010 1:43am Nov 18, 2010 1:43am
Quoting tdion
Totally off on my own tangent on this one:

I look at the Coin Flip scenario as a paratrooper being dropped off in No Man's Land.

Either you know how to trade the market you are dropped off in, or you don't.

Bottom line:...
yeh yeh man, same dif. i wasn't implying that my understanding of rr's challenge was to have each trade a home-run. "manage as you see fit" is ambiguous enough to mean anything at all.

that being said, he does clearly say that when flat on positions (no open trades) that you would then immediately open a new random long or short. so one way or another you will be in a trade and don't get to pick and choose the battle which is what we are often told to do by 'experts'.
Congratulations ScalaFX for coming out of the closet!
  • #100
  • Nov 18, 2010 1:46am Nov 18, 2010 1:46am
I didn't even read what you wrote. This was just something that popped in my head earlier...
0 traders viewing now