system

[crewefan9]

hi guys, i thought up this system trading forex using spread betting, its probably been done before because if it was this simple everyone would do it right? In theory it should work but i know it probably wont but would like someone who knows there maths to tell me why, here goes

do say 10 trades on forex flipping a coin to decide wether to go long or sell short.
enter the trade with a say 25 point stop loss and a 70 limit and leave it until it hits either one.

my theory being, stocks can only go 2 ways i.e 50-50 up or down ( i know its not that simple) flicking the coin is to make it so your choice wether to buy or sell is completley random. As everything is 50-50 apart from the stop and limit which is in my favour i should win in the long run.

I know the actually chances of a currency going up or down are not 50 50 as there are other factors but as you are choosing whether to buy or sell totally randomly surely it will even itself out evently?

thoughts?

 

[tomorton]

Hard to say this will definitely not work because of the good risk:reward.

However, it's not 50-50 whether price goes up or down after any individual entry. As a rough rule of thumb, price from any moment will continue in the direction of it's previous trend 2 times out of 3. Therefore, if you use a coin toss to determine direction, which really is 50-50, you are going to throw away half of your with-trend opportunities, as your coin toss will force you to enter counter-trend. Of course, much of the time, price isn't in a trend, so a coin toss would not be so damaging: but the point is that any entry during that ranging period in either direction would be equally likely to be damaging, perhaps getting back to your 50-50 estimate.

So you end up taking trades when you shouldn't take any, and directions which your eye will tell you are wrong. The system might pay long-term but who has the stomach to keep at it?

 

[Chartsy]

the coin flip/random entry thing has been done to death and it doesn't work

 

[crewefan9]

yeh i presume it probably has but i want a mathmatical answer why

 

[tar]

hi guys, i thought up this system trading forex using spread betting, its probably been done before because if it was this simple everyone would do it right? In theory it should work but i know it probably wont but would like someone who knows there maths to tell me why, here goes

do say 10 trades on forex flipping a coin to decide wether to go long or sell short.
enter the trade with a say 25 point stop loss and a 70 limit and leave it until it hits either one.

my theory being, stocks can only go 2 ways i.e 50-50 up or down ( i know its not that simple) flicking the coin is to make it so your choice wether to buy or sell is completley random. As everything is 50-50 apart from the stop and limit which is in my favour i should win in the long run.

I know the actually chances of a currency going up or down are not 50 50 as there are other factors but as you are choosing whether to buy or sell totally randomly surely it will even itself out evently?

thoughts?

If the SL=25 and your target =70 pips , then the odds is not 50-50 ...

 

[crewefan9]

yes ofcourse i know that hence that is why they are set like that so theoretically my wins will be bigger than my losses. i.e flipping a coin but the odds being 70-25 in my favour

 

[speedracer1996]

I always run a regression analysis on my trading ideas/trade triggers. If it will work then the historical data should prove you right or wrong...

 

[tomorton]

Your idea is really gnawing at me, so here I'd like to expand a bit on probability. Let's assume the market ranges twice as much as it trends, but when trending, continues in the trend direction twice as often as it reverses. Let's also assume that, when price is ranging, it is 2.8 times as easy for it to make 25 pips as 70, so 74% (2.8/2.8+1) of all range entries hit the stop, 26% hit the target.

So you have 1,000 potential entry points. 667 entries will be in ranges and 333 in trends. Of the range entries 74% (494) are stopped at -25 = -12,350: but 26% (173) hit target, making +12,110.

Of the 333 trend entries, let's say 167 are with-trend, 166 are counter-trend: assuming 2 out of 3 with-trend entries hit target, that makes 111 winners at +70 = +7,770 pips, but 56 losers at -25 = -1,400. Of the 166 counter-trend entries, let's say 2 out of 3 are stopped out, that makes 111 losers at -25 each = -2,775 pips: but 1 out of 3 are winners, making 55 wins at +70 = +3,850 pips.

Overall net result is
-12,350
+12,110
+7,770
-1,400
-2,775
+3,850

= +7,205, which is indeed positive.

But look how many assumptions.

 

[tar]

yes ofcourse i know that hence that is why they are set like that so theoretically my wins will be bigger than my losses. i.e flipping a coin but the odds being 70-25 in my favour

who said it is in your favor ? in theory your stops will be hit more often ...

 

[tomorton]

who said it is in your favor ? in theory your stops will be hit more often ...

True: you can't make the market swallow your risk just by demanding a bigger reward.

 

[crewefan9]

thanks tomorton but what you just wrote did not make any sense to me lol im not that clever thanks for your time tho

 

[crewefan9]

what im trying to get at is if i bet a pound on a coin landing heads or tails that odds are 50 50 but if i get 2 pounds every time it is heads in the long run i will win. i know thats very simple but thats my basic idea and i know it probably doesnt work but was really looking for a simple reason why it wont work

 

[tomorton]

what im trying to get at is if i bet a pound on a coin landing heads or tails that odds are 50 50 but if i get 2 pounds every time it is heads in the long run i will win. i know thats very simple but thats my basic idea and i know it probably doesnt work but was really looking for a simple reason why it wont work

Don't do this, the odds aren't 50:50 that you will win some money or break even, they're xx:yy that you will win some money or lose all your money.

 

[speedracer1996]

Fundamentally, the tighter your stops, the more you get stopped out. Therefore, the most likely outcome is that you will have more loosing trades than winning trades, but this is still ok if your gross return from your winning trades trumps your higher number of losing trades. Your return will be directly dependant on market direction and volatility. If the market goes up 50 percent of the time, then that gives you 50% odds (you’re using a coin to guarantee 50%), your odds are then reduced by volatility because it can cause you to get stopped out on an up bet in an up market. I believe you would be positive sometime and negative sometimes, but statistical probability is what you’re looking for and the market is just too dynamic for any degree of confidence unless you take a look at it empirically. You can do this with paper trading, but historical data provides a much higher sample size, and therefore is more dependable. I would not try this with real money until I had proven to my self it is dependable (no mulligans in trading)...

 

[crewefan9]

Fundamentally, the tighter your stops, the more you get stopped out. Therefore, the most likely outcome is that you will have more loosing trades than winning trades, but this is still ok if your gross return from your winning trades trumps your higher number of losing trades. Your return will be directly dependant on market direction and volatility. If the market goes up 50 percent of the time, then that gives you 50% odds (you’re using a coin to guarantee 50%), your odds are then reduced by volatility because it can cause you to get stopped out on an up bet in an up market. I believe you would be positive sometime and negative sometimes, but statistical probability is what you’re looking for and the market is just too dynamic for any degree of confidence unless you take a look at it empirically. You can do this with paper trading, but historical data provides a much higher sample size, and therefore is more dependable. I would not try this with real money until I had proven to my self it is dependable (no mulligans in trading)...

no i havnt even started trading with real money yet just looking at ideas. Im doing it on a demo account to see what happens. i did wonder wether 20 was a to tighter stop. i started off doing 100 stop 200 limit

 

[kalott]

What are the odds that after a buy the stop gets hit and then it reverse up 95?

 

[tomorton]

What are the odds that after a buy the stop gets hit and then it reverse up 95?

Very low, otherwise it would be more successful to enter 25pts lower and have a 95pt target from the very start.

The thing is, you can't just decide on a nominal stop and target which give a juicy r:r and expect the market to oblige with profits.

 

[Pi3141]

Very low, otherwise it would be more successful to enter 25pts lower and have a 95pt target from the very start.

The thing is, you can't just decide on a nominal stop and target which give a juicy r:r and expect the market to oblige with profits.

Exactly, you can only take what the market gives you.

 

[traderman1]

Fundamentally, the tighter your stops, the more you get stopped out. Therefore, the most likely outcome is that you will have more loosing trades than winning trades, but this is still ok if your gross return from your winning trades trumps your higher number of losing trades. Your return will be directly dependant on market direction and volatility. If the market goes up 50 percent of the time, then that gives you 50% odds (you’re using a coin to guarantee 50%), your odds are then reduced by volatility because it can cause you to get stopped out on an up bet in an up market. I believe you would be positive sometime and negative sometimes, but statistical probability is what you’re looking for and the market is just too dynamic for any degree of confidence unless you take a look at it empirically. You can do this with paper trading, but historical data provides a much higher sample size, and therefore is more dependable. I would not try this with real money until I had proven to my self it is dependable (no mulligans in trading)...

Good answer. The 25 pip stop will be hit more times than the 70 pip target. Simple.

 

[Hotch]

Were it so easy. This sort of thing depends a fair amount on the volatility of the instrument. If we assume a normal distribution, then the strategy has no edge.

My honest opinion (from experience) is that it's not worth the effort. Sure you can make money, but you're better off going through more conventional means.