Coin Toss Trading

The Lobby
1

[B]Note to all: I put this thread in this category because it is for frivolous purposes only.[/B] This is by no means a way for anyone to trade, it’s just a silly experiment we can have some fun with.

Having seen numerous comments on this, and other various sites about making a coin flip for trading, I decided to set up a demo account, and do just that.

Nightly, at 9 p.m. pst, I will flip a coin. Heads will always be buy, and tails will always be sell. I will set up the trades with 100 pip t/p s/l on either side of the entry point. The entry happens upon the coin flip, no other signals, or indicators will be employed. The entry will happen at market price immediately.

I will use the euro only, and the trade will remain in effect until the position closes. What I mean by that is, if it makes it all the way through a 24 hour period without closing, the open trade will still be in effect, and no new trade will be initiated.

As soon as the first trade is over, I will open an mt4pips account, and post the link here.

The first trade is now open, the toss called a buy, and the entry price is 1.3624.

It’s a 50k account, and I set it to .10 for profit per pip.

We should have a nice long data base to draw from. It will likely take forever to get through it.

Feel free to add thoughts, or suggestions, or most importantly, humor.

Maybe we can have guest coin tossers time to time:D

2

Rethinking the equal s/l t/p targets.

Maybe I should go to a more standard r/r of 2:1.

Any thoughts?

3

Not so frivolous as you may think!!

You may be pleasantly surprised!! :smiley: :smiley:

A trader has tried exactly the same thing here on this forum some years ago

http://forums.babypips.com/newbie-island/2946-let-do-experiment.html?highlight=experiment

I will be interested to see how you go!!

4

Are you open to suggestion? I read about rather than heads/long and tails/short, use heads/no trade and tails/short. The former could have 100% failure but 100% failure for the latter is less likely. A side by side running could be conducted. Testing holding till close without a stop loss versus stop loss / take profit can be explored.

5

I thought about that to be honest, but went for the equal sides proposition. My thinking was taking one way trades only could really backfire in a longer term trend. For instance, the euro is in a slow slide right now. Based on that fact alone, and selling only, we could have an edge. But if that trend goes bullish, and all we do are sells or no trades, we limit the true 50/50 possibilities.

I do think I need to go for a 2:1 risk reward scenario though.

Approaching it from a true money management perspective may be more beneficial that just an equal split. Slowly we would lose it all based on the spread alone.

6

I would like to see the resuts of something like that purely from a statistics standpoint. Much appreciate the suggestion!

7

:slight_smile: Please do keep tabs. I think it will be a fun experiment.

Thanks for the link, I will read through it this weekend.

8

I did something similar just looking at a year’s worth of data on excel of course. Results were about even except that one buys or sells I forget which came out a bit ahead and looking back at a chart zoomed out I could see an obvious long term trend which would account for that. So assuming no long term trends you should come out about even.

9

It will take a long time to get the results there is another way though, I understand you can download tick quality chart history data from dukascopy which will give you backtest results at 99% quality. A simple EA could run a backtest to see what would happen.

10

I tried coin flipping several months ago, it’s a waste of time…

No offense intended =)

11

Thanks for the replies:D

Forward idiocy is the only way, Especially since I don’t know how to program a damn thing on MT4.

This is intended to be purely entertainment, but I think there will be some real surprises.

Last night’s trade was a loser.

I’m going to change the s/l t/p to a 2:1 r/r ratio.

I’m also going to increase the lost size a bit, and lessen the s/l t/p to 30 and 60 pips respectively. Half the daily range should be enough to make a trade happen nightly.

And also, once the trade is initiated, I won’t meddle with it.

It’s all or nuttin’ hunny.

This won’t take much time each night to do, and I will post the mt4pips link shortly.

Now here’s a question. Theoretically, if the 2/1 ratio pays off at a 50/50 win rate based on the coin toss, this should make money right?

10 wins is 600 pips, and 10 losses 300. Subtract the spread, and we should be up 280 pips a month on the euro. Of course in this case, math isn’t a 2=2=4 proposition. The odds change that.

We shall see:D

Cheers!

12

This is intended as entertaining frivolity only:D

Please someone start with the jokes:p

13

I found the whole exercise of developing strategies based on random entries extremely useful. I’d go as far as to say it was amongst the most beneficial research that I’ve ever undertaken.

14

Tails tonight.

Sold in at 1.3506

s/l at 1.3536, t/p at 1.3446. We shall see:D

15

This should be a good one I’ll be interested to see what happens especially as I also have a sell in at 1.3514 lol

16

S/L at 1.3536 is to close to call…haha, may the force be with this trade :smiley:

17

Dead soldier.

LOL even if I had widened the stop, the dip down never made it to the t/p point.

Stop losses are the devil:p

18

I don’t think its as simple and as lucrative as you have proposed. The chances of price moving 30 pips is much greater than it moving 60 pips. The chances of it moving 30 pips against you and then 60 pips in your favour would reduce the odds to your detriment. It certainly isn’t a 50/50 proposition for this reason. I wish I could give you the exact mathmatical formulae, but my skills aren’t as strong as they used to be.

Strictly off the top my head, I would put your chances over the long run at a loss equal to the number of trades multiplied by the spread or slightly less. But I’m sure there is somebody who could model this more accurately and I would enjoy seeing an attempt to do this.

19

This really is a very interesting excercise at first it seemed a simple and obvious outcome that statisticly you should win 50% of the time but, and I have no idea how to apply this mathematically, there might be another factor involved, in that the price doesnt just go up or go down, during one 24hr period it often does both, and goes up and down in waves so even though you could open a trade that runs into profit, during the 24 hr period it may run into loss and still be at a loss at the end of the 24hr period but then had it run longer or shorter it may have been in profit when you closed it out so I was wondering how much does this wave action work against us to reduce our odds of a succesful trade

20

I think if your take profit and stop loss are equal then your odds are 50 / 50 disregarding certain other factors.