Coin Flip Trading

trendie

Legendary member
Messages
6,875
Likes
1,433
So, on a purely random trade, and excluding expenses, commission,s (for the time being), is there a 50/50 probability of a trade winning or losing?

For the purpose of examples, a 100 pip win or a 100 pip loss.

Would you expect an account to neither win nor lose?

(I have a cunning plan, more cunning than a cunning than a cunning thing fron the planet cunning, and doesn't require the usage of pencils up nostrils nor underpants.)
 
Does this cunning plan involve flipping a coin then becoming a signal provider ? If so, I'm in, that is surely more cunning than a fox that has just graduated and become professor of cunning at oxford university.
 
Oh no trendie, you appear to be yet another victim of the 'so' epidemic that's sweeping the land - where every sentence and new conversation starts with 'so'.

Oh well, never mind. Yes, in answer to your question - and all other things being equal - it's a 50/50 probability over time. I imagine the equity curve would resemble a momentum oscillator like RSI where there are profit peeks and loss troughs above and below the break even centre line. So, at any one time, the account might be in profit or loss. I'll let the probability experts and maffs geeks suggest how far above or below it might go.
Tim.
 
timsk,

I can only apologise for my error. I am normally fairly pedantic about such things.

It must be the company I keep here on T2W! :)
 
What if, you went for 100 pips.
Suppose you placed 5 lots.
However, you hold on for the whole 100 pips when it goes in your favour, and you take a loss at every 20 pips if it goes against you.

For example, the market is at X.X500.
You get a BUY signal.
You now wait for X.X600 to close out all your 5 lots.

But, should the market go against you, you close out a lot at
X.X480,
X.X460,
X.X440,
X.X420
X.X400.

Your gains would be 500 pips if you win.
(5 lots closed out at 100 pips)

Your losses would be 300 pips if you lost.
(1 lot closed at -20,
1 lot closed at -40,
1 lot closed at -60,
1 lot closed at -80,
1 lot closed at -100)

However, if the market wobbled;
if the market wobbled back 20 pips before hitting 100, you would get
(4 lots closed at 100 (ie, +400), less 1 lot closed at -20; nett= +380)

if the market wobbled back 40 pips before hitting 100, you would get
(3 lots closed at 100 (+300), less lot closed at -20, and another at -40; nett= +240)

if the market wobbled back 60 pips before hitting 100, you would get
(+200, -20, -40, -60; nett= +80)

if the market wobbled 80 pips,
(+100, -20, -40, -60, -80; nett= -100)

I am wondering if this can be seen as an asymmetric risk profile?
That is, a way of getting turning a 100 pip win or a 100 pip loss into a way of skewing amounts won or lost toward one direction.

Obviously, the coin flip was just a way of setting a baseline.
If you can gauge the direction, and wanted to make more of the 100 win/100 loss, would this be a way?

I suspect this has been done to death elsewhere!

EDIT: I was playing with asymmetric risks with options, and thought this would be a proxy using bog standard trading tools.
 
Last edited:
So, on a purely random trade, and excluding expenses, commission,s (for the time being), is there a 50/50 probability of a trade winning or losing?

For the purpose of examples, a 100 pip win or a 100 pip loss.

Would you expect an account to neither win nor lose?

(I have a cunning plan, more cunning than a cunning than a cunning thing fron the planet cunning, and doesn't require the usage of pencils up nostrils nor underpants.)

Trendie, you are getting desperate! :D
 
So, on a purely random trade, and excluding expenses, commission,s (for the time being), is there a 50/50 probability of a trade winning or losing?

For the purpose of examples, a 100 pip win or a 100 pip loss.

Would you expect an account to neither win nor lose?

(I have a cunning plan, more cunning than a cunning than a cunning thing fron the planet cunning, and doesn't require the usage of pencils up nostrils nor underpants.)

In the long run it will be about 50/50, and you will have no edge,

With your alternative system for stops, the profitable trade % will change, as will the profit amounts, but you'll still be left with no edge.
 
In the long run it will be about 50/50, and you will have no edge,

With your alternative system for stops, the profitable trade % will change, as will the profit amounts, but you'll still be left with no edge.

If there are 6 possible outcomes, ie, +500, +380, +240, +80, -100, and -300, then the total possbile is (1200 - 400) = nett of 800 pips over 6 possible events.

Isn't that a positive outcome?
4 scenarios out of 6 result in a profit, accounting for wobble.

Even if we consider that if the market goes 60 and 80 pips against you, chances are they will result in a full 100 pip loss, (300 pips across 5 lots)

thats still, +500, + 380, +240, -300, -300, -300.
which is +1120 against -900, which is +220 over 6 events.

I concede the spread of events is random.
 
So, on a purely random trade, and excluding expenses, commission,s (for the time being), is there a 50/50 probability of a trade winning or losing?

For the purpose of examples, a 100 pip win or a 100 pip loss.

Would you expect an account to neither win nor lose?

(I have a cunning plan, more cunning than a cunning than a cunning thing fron the planet cunning, and doesn't require the usage of pencils up nostrils nor underpants.)

I would expect to lose the spread.
 
In the long run it will be about 50/50, and you will have no edge,

With your alternative system for stops, the profitable trade % will change, as will the profit amounts, but you'll still be left with no edge.

It won't.

The market has an inbuilt negative expectancy in the spread. Plus you will stop out when your stop is TOUCHED but you can't limit out when your target is touched. You have to presume you only get filled on your target when you move past that price.

Consider ES - with 5 tick stop & 5 tick target.

Your coin tells you to BUY. You buy 1500.00. That price is the inside offer. The current inside bid is 1499.75.

Your stop loss is at 1498.75 - only 4 ticks below the inside bid at the point of entry. So if the market ticks down 4 ticks and trades 1 contract there -- you are done.

Your target is 1500.25 - 5 ticks above the inside offer at the point of entry. The market needs to tick up 5 ticks to get there. Once there, you will be at some position or other in the queue and you need the people in front of you to get filled before you get filled.

So your coin toss absolutely had negative expectancy.
 
If there are 6 possible outcomes, ie, +500, +380, +240, +80, -100, and -300, then the total possbile is (1200 - 400) = nett of 800 pips over 6 possible events.

Isn't that a positive outcome?
4 scenarios out of 6 result in a profit, accounting for wobble.

Even if we consider that if the market goes 60 and 80 pips against you, chances are they will result in a full 100 pip loss, (300 pips across 5 lots)

thats still, +500, + 380, +240, -300, -300, -300.
which is +1120 against -900, which is +220 over 6 events.

I concede the spread of events is random.

In what way positive? The initial coin flip you had a 50% of getting 500. In the second system, your chance of getting 500 is a lot less than 50%. The probability adjusts downwards. I believe there is a less than 50% chance of any positive outcome
 
Last edited:
In what way positive? The initial coin flip you had a 50% of getting 500. In the second system, your chance of getting 500 is a lot less than 50%. The probability adjusts downwards. I believe there is a less than 50% chance of any positive outcome

The chances of getting to 100 pips is 50/50.

If you get to +100 without a wobble you collect +500 from 5 lots.

If you get to +100 with a -20 wobble you collect +380 from 4 winning lots and a losing lot.

If you get to +100 with a -40 wobble, you collect +240 from 3 winning lots and 2 losing lots.

Your chances of hitting -100 is also 50/50.
However, lets say you lose 5 lots, staged, resulting in a -300 loss on all 3 scenarios, from 5 lots.

+100 = 50/50.
-100 = 50/50.

But the manner of staging your ,losses, but holding your winning lots, results in:
+100 = +500, + 380, +240. nett=1120.
-100 = -300, -300, -300. nett = -900.

EDIT: we are not talking about getting to 500 pips, we are talking about getting to either +100 pips or -100 pips.
The changeable amounts reflect the pullback or wobble before it gets there.
In the 3 losing scenarios, and 3 winning scenarios, the nett gains are skewed against the nett losses if you stage the losers.

EDIT: I am willing to concede I am missing some fundamental probability concept.
But hope I am open-minded enough to learn what that is.
 
Last edited:
Trendie, you said there are 6 possible outcomes +500, +380, +240, +80, -100, and -300.

Now we get -300 if it hits -100 before it hits +100. The chance of this happening is 50% (coming from your initial post). IS that correct or have I misunderstood? So we have 6 possible outcomes, but one of them is 50% chance and it's negative. So what are the % chances of the other 5 outcomes? Unless I've completely misunderstood, the other 5 outcomes cumulative come to 50%. So for example, +500 might be 17% chance of occurring...or whatever. When you calculte the expectancy, it will be zero.
 
I think there are only 2 outcomes.
-100 or +100.
Maybe I got ahead of myself with the various outcomes what with the wobbles.

The outcomes are 50/50.
 
So if you buy the whole Bayesian even stevens thing with regards to up/down moves then you're less than 50/50 due to what Dionysus Taoist said then when you get to -20 you're only 4 lots and you need to move up 120 (+fill) to hit you target and yet you're gonna concede further losses if the market moves down only 20 mores and so on. Everything that happens after the initial less that 50% event -ve so the net result canny be positive cannit?
 
I think there's a difference in the way I'm looking at this from you guys... you're all looking at the trade being placed initially where I'm looking at a series of independent events when each stop gets hit. does the 50/50 start again when the first target is hit or are we saying that the markets more likely to move up once that happens?
 
Hi scose

from my perspective its just one trade.
The trade closes at +100 as a win or at -100 as a loss.

For the purposes of this exercise we have no expectations of what the market may or may not do.
 
I'd have thought it would be like this:

If you multiply out all of the loss probs you get 50%, do the same with the targets and you get mother funking nada.

See why I asked about the strat assuming that the market is more likely to go up once the loss band are hit?
 

Attachments

  • trendie.xlsx
    9.2 KB · Views: 233
Last edited:
If there are 6 possible outcomes, ie, +500, +380, +240, +80, -100, and -300, then the total possbile is (1200 - 400) = nett of 800 pips over 6 possible events.

Isn't that a positive outcome?
4 scenarios out of 6 result in a profit, accounting for wobble.

I think there are only 2 outcomes.
-100 or +100.
Maybe I got ahead of myself with the various outcomes what with the wobbles.

The outcomes are 50/50.

Make your mind up Trendie! :LOL:

Yes price goes +100 or -100, but the outcomes of the trade with your management in terms of P&L are +500 or +380 or ... -300 (you said these numbers in an earlier post). So we consider this as 6 possible outcomes. The probability of -300 is ...?
 
Top