The Coin Trade System aka Flip The Coin | Page 3 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

Flip The Coin System 72 replies

The Coin-Flip Theory 386 replies

Coin Flip Reloaded 18 replies

Coin Flip with correct Money Management system. Will it profit ? 25 replies

Issues about Market Depth, AKA LEVEL II 1 reply

Attachments: The Coin Trade System aka Flip The Coin
Exit Attachments

The Coin Trade System aka Flip The Coin

  • #41
  • Mar 9, 2008 5:50am Mar 9, 2008 5:50am
Quote
These coin toss experiments seem to have exceedingly short lifespans. This is the third one I've seen just sort of...fizzle. I wonder why that is? Boredom?
Probably because of the remarkably low expectancy. I back-tested several coin flips approaches, both with and without a Stop&Reverse, on various time frames, with various delays between the next trade, and with various stop losses. No matter what settings I used it always drained the equity in rapid fashion. It was only a matter of how long you ran the test for (anything under 2 years isn't really reliable).

If I had to guess why... the market may go up or down. If you always went long, you'd have a 50% / 50% chance of getting it right assuming a perfectly ranging, non-trending market. If you also flip a coin for direction that's a 50/50 chance you'll guess the direction, meaning you have a 1/4 or 25% chance of getting the right direction at the right time. In order to profit from that your entry would have to net over a 4:1 RR, which for random entries is going to be problematic at best (it's problematic on non-random entries even).

I'm not sure you can optimize your SL and TP to fix that simple fact without turning this into a carry-loss system (open up the SL wide enough and I'm sure some trades will work out, but those that don't will really hurt you).

I suspect you'd do better always going long or short, or basing it on some kind of long term trend. I suspect looking for momentum clues and trading only when there's enough volatility to support you might help. But all of that is outside the scope of a simple coin flip method.

As for using a coin flip for entry... in a trending market that works. Statistically a coin flip entry in the direction of the trend (again, I've back tested this pretty extensively) is as good as any other entry. Guess that explains why the dart board method works so well in a bull market. Ranging or sideways periods tho... probably not so good .
  • #42
  • Mar 9, 2008 7:04am Mar 9, 2008 7:04am
Bullrock, I did tons of analysis on coin-flipping to work out the effect of effective SLs and TPs a year back, to work out what a dumb system looks like (to compare against as a baseline). As you know, the chance of a win assuming no SL or TP is 50% excluding the spread. On GBPUSD including the spread, it was more like 46% if you were trading daily candles).

By optimising the SL, I worked out you can get around a 1% edge on average (and again, it has to be on average, because with a random series of coin tosses, you'll get different results each time, so this is based on a distribution of 1000s of attempts).

Needless to say, a coin-flip system, if fully optimised with the right SLs (and no TPs) has a very slow downward equity curve - ie, it very slowly bleeds to death.

With TPs, nothing I did would increase the overall profits on average.

This tells you 2 things:
1) TPs are only beneficial if you can place them with knowledge of the market (ie based on S&R and in certain market conditions etc). I don't use them myself.
2) SLs can give you an edge so long as you always make sure they're wider than they have to be. The 1% edge gradually faded to 0% but never below if you widened them out. But if you closed them up, it'd drop to less than zero; hurting your trading, and eliminating any advantage your system has. If in doubt, widen them out.

PS The gambler's ruin theory is sound, but this interpretation is wrong:
"Even with better than even odds, you will eventually go bankrupt, and have to quit. But, this may take very long, indeed."
  • #43
  • Mar 9, 2008 7:33am Mar 9, 2008 7:33am
Fun stuff although things like this always give folks the impression that Forex entries are more difficult than they are.

I'm reminded of the Turtles doing random entries simply to see if they can manage a trade regardless of their entry points. THIS should be the the lessons learned with this activity if any as exits/profits are the mojo's of trading.

I also believe that a very large sample should be part of the project.

Cheers and good fun,

kt
  • #44
  • Mar 9, 2008 8:10am Mar 9, 2008 8:10am
While I admire the basic idea that the markets are random, they are not. They appear to be random, based on a quick glance however, if you really study them you will see they are not.

I know it sometimes seems that whatever entry you make the market always goes for you and kicks your stop, its going down you sell, it turns stopsyou out and carries on going down, but you bought it. So it has you again!

If I was paranoid I would think that the trading god had it in for me! Just me, but 80% of people are lossing, so the trading god is after 80% of us, or 80% of us are doing the wrong thing at the same time.

I was predicting a turn in the euro dollar, my levels were out, but upon re-analysis, I understand why, I know why I was wrong, I lost 40 points trying to get in twice, then I said ok wait for the turn then trade it, I waited it turned I got in and made 80 points.

Sell high, buy low - that is how you trade, news is designed to get you in the wrong way. Bad news from the US on Friday should have sent the dollar down to the floor, but the good old FEDS announce that they are going to buy back 100 billion US dollars and the dollar shoots up. Predict that, which is precisely what I did using pattern recognition, so lets see how we can predict the market, not just throw up our hands and say its random, money management will save your bacon, trading is about closing a loosing position first and letting a profit position roll up.

This is why most people fail, as soon as they see a small profit they close and they let their loosers run, the loosers continue to grow and their winners are painfully small - hence failure.

Someone needs to design a trading platform that shows your positions in reverse ie looser = winners and winners equal loosers.

Give that to any new trader and I guarentee that they are consistantly making money in week one what ever system they are using.

This system will work if the following is applied, close the loss trades quickly, give the winning trades enough room to develop, and hold the profit in them using trailing stops.

The direction trade is not important, long short its all the same it is how you manage that trade that counts.


Smalldog2


I look for butterflies, bats and crabs, and I call that working!
  • #45
  • Edited 3:28pm Jun 1, 2008 3:03pm | Edited 3:28pm
Interesting...just stumbled upon this thread, but from what i know:

coin tosses follow a discrete binomial probability distribution, with the random variable having a countably finite number of possible values (H or T).

i'm not sure price can be accurately modeled after a binomial distribution even if at first it may seem correct because the random variable is continuous(at least from what i think) and (un)countably infinite.

but maybe i understood the purpose of the thread wrong.

this is something that definitely deserves a closer analysis.
  • #46
  • Jun 1, 2008 3:12pm Jun 1, 2008 3:12pm
At least keep a journal to back up your statements.... Everything points to them being random (yawn)

Quoting smalldog2
While I admire the basic idea that the markets are random, they are not. They appear to be random, based on a quick glance however, if you really study them you will see they are not.
  • #47
  • Jun 1, 2008 3:23pm Jun 1, 2008 3:23pm
Quoting tdion
At least keep a journal to back up your statements.... Everything points to them being random (yawn)
What would you accept as proof, tdion?

Thx,

PC
  • #48
  • Jun 1, 2008 3:35pm Jun 1, 2008 3:35pm
This Thread


Where is Bullrock? Haven't heard from him in awhile.
Just Trade It
  • #49
  • Jun 1, 2008 3:44pm Jun 1, 2008 3:44pm
One flip of the coin a few months back and the system is history.....
Same Whore .... Different Dress
  • #50
  • Jun 1, 2008 3:48pm Jun 1, 2008 3:48pm
Quoting tdion
At least keep a journal to back up your statements.... Everything points to them being random (yawn)
I'm not sure if I should even bother starting to explain why everything does not "point to them being random" but FYI, here are is a paper that provides a model which outperforms random walk models.

I have plenty of similar stuff in my extensive book collection, but obviously I have no way of sending you those and even if I did, I don't think it's worth it.It probably won't even make a difference, because you've clinched to "randomness" and you just won't let go of it, no matter how much rational evidence someone, anyone would provide to you.

As with many other people and in many other different situations, "randomness" just feels right to you, so you decide to automatically disregard any other contrary evidence.

http://papers.ssrn.com/sol3/papers.c...ract_id=308586

Paper Title:

The Out-of-Sample Success of Term Structure Models as Exchange Rate Predictors: A Step Beyond

Abstract:
A large literature suggests that standard exchange rate models cannot outperform a random walk forecast and that the forward rate is not an optimal predictor of the spot rate. There is evidence, however, that the term structure of forward premia contains valuable information for forecasting future spot exchange rates and that exchange rate dynamics display non-linearities. This Paper proposes a term-structure forecasting model of exchange rates based on a regime-switching vector equilibrium correction model which is novel in this context. Our model significantly outperforms both a random walk and, to a lesser extent, a linear term-structure vector equilibrium correction model for four major dollar rates across a range of horizons.
  • #51
  • Jun 1, 2008 4:00pm Jun 1, 2008 4:00pm
Quoting seeit
then using a coin toss..


Its not 50% ...it is 25 % ....here is why....

The market is a 50/50 % ....up or down

Your choice ( analysis , coin toss, etc.)...50 /50 %

The mistake is that people think that there coin toss is a 50/50 chance of being right....IT IS 25%

Because it is .50X .50 = .25 %

If you were to stick with on direction for the rest of your life...lets say Long , Everytrade you made the rest of your life would have a 50/50 chance.

But if you go back and forth in your mind between long and short....and the market is going back and forth... guess what...it is not 50/50 its 25%

If in vegas , you sit a table that has a red or black choice...baring luck over x trials...you should be @ 25 % right.

If you sit at that same table ..and pick just one color, lets say red...over the same period as above , you should be closer to 50%

Vegas puts out a contract on anyone how brings up this topic
Vegas makes money because they don't pay true odds. For example there are 6 ways to make a 7 on the dice, so the odds are that a 7 will come out 1 time for every 6 rolls. But they only pay you 4 to 1 odds so over time, they have to win.

Every game is rigged that way.
  • #52
  • Jun 2, 2008 4:54am Jun 2, 2008 4:54am
Quoting seeit
then using a coin toss..
Its not 50% ...it is 25 % ....here is why....
The market is a 50/50 % ....up or down
Your choice ( analysis , coin toss, etc.)...50 /50 %
The mistake is that people think that there coin toss is a 50/50 chance of being right....IT IS 25%
Because it is .50X .50 = .25 %
If you were to stick with on direction for the rest of your life...lets say Long , Everytrade you made the rest of your life would have a 50/50 chance.
But if you go back and forth in your mind between long and short....and the market is going back and forth... guess what...it is not 50/50 its 25%
If in vegas , you sit a table that has a red or black choice...baring luck over x trials...you should be @ 25 % right.
If you sit at that same table ..and pick just one color, lets say red...over the same period as above , you should be closer to 50%
Vegas puts out a contract on anyone how brings up this topic

I didnt bother reading the rest of the thread but i couldnt let this go unanswered. That statement is wrong. The odds are 50/50.

there are four possible outcomes

You go long and are right.
You go long and are wrong.
You go short and are right.
You go short and are wrong.

each one has a 25% chance of happening, since two of them lead to being right and two to being wrong your chances of being correct on any one trade are 50/50
  • #53
  • Jun 2, 2008 6:11am Jun 2, 2008 6:11am
nope, odds arent 50% - they arent 25% either

if you were playing with zero spread and with an equal stop and target then you'd be 50 / 50

but you're not - you are instantly wrong whichever way because of spread, so if the criteria was 1 pip tp vs 1 pip stop then you would have a negative expectation of 100% because the spread would instantly take your stop.

then of course there is the 'wrong when you are right' scenario - if you are right at the wrong time, then you are wrong.

what if your flip goes heads long, price rtetraces 20 pips taking your stop but then goes north 100 pips - you were right but at the wrong time.

these reasons and so many more not mentioned give a coin flip a <50% probability of success, therefore the only way to win with this is through stop and tp management.

https://linktr.ee/Soultrader
  • #54
  • Jul 21, 2009 7:50pm Jul 21, 2009 7:50pm
What is "random"?

Quote
random: without definite aim, direction, rule, or method
Let's say the market is not "random".

No one knows which way the next tick will go.

No one can control the outcome of their trade. Once you enter, it is "wait and see".

That means the outcome of any given trade is "random".

Quote
Luck is a belief in good or bad fortune in life caused by accident or chance which happens beyond a person's control.

Since you can not control the outcome of any particular trade, there is an element of "luck".

That fact rubs many traders the wrong way.
  • #55
  • Jul 21, 2009 8:42pm Jul 21, 2009 8:42pm
Random is an outcome function that accepts no inputs. For that reason, there's no memory between the outcomes. There's no correlation between what happened last time and what will happen next. Ex: A true coin flip.

Non-random is an outcome function that accepts at least 1 input from the prior outcome. That means no matter how complex the function is, there is always some measure of predictability by looking at the past. Ex: Poker.

Psuedo-random is an outcome function that accepts at least 1 input, but that input is not from a prior outcome, it's from an external factor. That means no matter how complex the function is, the measure of predictability is based entirely on your ability to determine the external factor. Ex: Randomized computer games.
  • #56
  • Jul 24, 2009 5:08am Jul 24, 2009 5:08am
1/ Hold for a week.
2/ Stop loss about 20% of TP.
3/ Also use a break even stop.
Back tests on 10 years of 5 min data show a profit. Once a day is pointless, no chance to get the risk reward high enough. Good way to test money management techniques in my opinion. Also to show the true value of an indicator.
  • #57
  • Jul 25, 2009 5:39pm Jul 25, 2009 5:39pm
Quoting Rabid
Random is an outcome function that accepts no inputs. For that reason, there's no memory between the outcomes. There's no correlation between what happened last time and what will happen next....
I understand what you are saying. Are these your definitions? If not, where did you copy them from?
  • #58
  • Jul 25, 2009 6:24pm Jul 25, 2009 6:24pm
Quote
I understand what you are saying. Are these your definitions? If not, where did you copy them from?

They're mine, mostly. But I probably copied the idea from somewhere, no idea where.
  • #59
  • Jul 25, 2009 11:29pm Jul 25, 2009 11:29pm
For those who want to try the coin flip experiment... I suggest using a TRUE random number generator (Just set possible outcomes to 0 and 1)

http://www.random.org/

there are others out there too.
  • #60
  • Edited 2:41pm Jul 30, 2009 2:28pm | Edited 2:41pm
I have a different Idea.
Take a dart board to someone who doesnt know anything about what your doing, and have them Put at random Ls and Ss in all the different areas on the board. Dont tell them what its for.

Now throw a dart everyday aat a set time nd what it lands on take your favorite pair long or short.

Now you have to come up with TP and SL

suggest 115 tp and 55 sl

What do you think of that one?

PS If your are god at darts this may not work to well!
0 traders viewing now