Becoming competent at forex trading

[skyQuake]

If you have correct money management, positions sizing etc, then you have an ege, a 50/50 chance is all you require to profit.

Have a mechanical/discretionary strategy which gives a greater expectancy then 50% is an extra edge.

How so? You can't manage a random number system to profitability.. Not with brokerage and slippage

 

[Mr J]

If you have correct money management, positions sizing etc, then you have an ege, a 50/50 chance is all you require to profit.

Have a mechanical/discretionary strategy which gives a greater expectancy then 50% is an extra edge.

No amount of money management will turn a strategy with negative expectancy into one with positive expectancy. A 50% chance in trading is not necessarily a coinflip. If we hit 50% and have a 2:1 RR, we're not making our money with money management. It may seem that we are, but rather than it being a "coinflip", that 50% chance requires price to move twice as far in one direction as it does the other. Therefore the edge comes from forecasting movement, not money management. Money management does not provide an edge, it controls the level of risk.

 

[lukeaye]

No amount of money management will turn a strategy with negative expectancy into one with positive expectancy. A 50% chance in trading is not necessarily a coinflip. If we hit 50% and have a 2:1 RR, we're not making our money with money management. It may seem that we are, but rather than it being a "coinflip", that 50% chance requires price to move twice as far in one direction as it does the other. Therefore the edge comes from forecasting movement, not money management. Money management does not provide an edge, it controls the level of risk.

Thats the thing, your capping it to 2:1 RR. The only constant is your Risk, The profit is random and uncapped in theory. Whether you have an expectancy based on historical performance, you still don't have a constant foward expectancy. Price movement is completey random, expectancy is subjective. Methods which historically had say a 87% expectancy, can revert to a 22% expectancy foward.

Money management is a far greater edge, then an expectancy model, thats all im trying to say. I was guilty of thinking that i had to come up with the perfect trade set-up, and in reality neglected the greatest edge I could have. Correct money management.

I agree that predicting price movements foward, based on historical prices can give you an additional edge, but i feel money management is a greater edge, thats all.

 

[Mr J]

Price movement is completey random.

No, it's not. If you really think this, you shouldn't be trading.

expectancy is subjective

Expectancy is not subjective at all. It has a precise value, we just do not know what it is.

Money management is a far greater edge, then an expectancy mode

Expectancy is edge.

I understand why you think what you do, but the problem is that taking 2:1 is no longer a 50-50 proposition. By altering the R:R, you're now requiring price to move further in one direction than the other, or in other words you now require an uneven distribution in price movement. If price movement was random as you suggest, then requiring a 2:1 R:R would simply lower the winrate to compensate, and end up in the same place. If price movement isn't random, then setting a 2:1 R:R might produce a small edge, but this would be due to the behaviour of price movement, and not money management. In this case, the trader would be exploiting inherent characteristics of the market. I doubt it could be randomly applied though and provide a worthwhile edge (if any edge).

 

[>Apocalypto<]

Thats the thing, your capping it to 2:1 RR. The only constant is your Risk, The profit is random and uncapped in theory. Whether you have an expectancy based on historical performance, you still don't have a constant foward expectancy. Price movement is completey random, expectancy is subjective. Methods which historically had say a 87% expectancy, can revert to a 22% expectancy foward.

Money management is a far greater edge, then an expectancy model, thats all im trying to say. I was guilty of thinking that i had to come up with the perfect trade set-up, and in reality neglected the greatest edge I could have. Correct money management.

I agree that predicting price movements foward, based on historical prices can give you an additional edge, but i feel money management is a greater edge, thats all.

1:1 wont help you in any market if you have 7 losers to 3 wins in 10 trades....

trust me I know from personal experience!

 

[>Apocalypto<]

Developing a system as mentioned before is key.

One really important thing for me was becoming comfortable with taking small losses. Second: The ability to stay with trends and not close out winning positions while they are still running.

Many beginning traders cut their winners and let their losses run.

This could apply to any market/

Good post!

 

[lukeaye]

No, it's not. If you really think this, you shouldn't be trading

Well then what is random. Historically it has been defined with a probabilty, but foward anything can revert to random, or unexpected.

Expectancy is not subjective at all. It has a precise value, we just do not know what it is.

Expectancy on historical results, yes. Expectancy is subjective. My example. A system has an expectancy of 70% over 30 years of data. Over the last 5 years it has an expectancy of 20%, and this year 55%. Which do you take as your expectancy?

Expectancy is edge.

Agreed, not disputing that, but so is money MM because you have the constant of risk. Without this constant, your expectancy is really nothing, because it is dynamic.

I understand why you think what you do, but the problem is that taking 2:1 is no longer a 50-50 proposition. By altering the R:R, you're now requiring price to move further in one direction than the other, or in other words you now require an uneven distribution in price movement. If price movement was random as you suggest, then requiring a 2:1 R:R would simply lower the winrate to compensate, and end up in the same place. If price movement isn't random, then setting a 2:1 R:R might produce a small edge, but this would be due to the behaviour of price movement, and not money management. In this case, the trader would be exploiting inherent characteristics of the market. I doubt it could be randomly applied though and provide a worthwhile edge (if any edge).

So can we exploit this randomness in price movement, with MM, by allowing it to randomly distribute?

enjoying this discussion!

 

[lukeaye]

1:1 wont help you in any market if you have 7 losers to 3 wins in 10 trades....

trust me I know from personal experience!

See your making the exact same assumption, that they will all be 1:1.

Only one part is constant, the other part is random. And winning 7 trades to 3trades is also result of randomness.. If you assumed 1:1 as you return/risk then why not assume 5 wins 5 losses?

But yes your right if you lost 7 and won 3 with a constant of 1:1 of course you wont make money.

 

[>Apocalypto<]

See your making the exact same assumption, that they will all be 1:1.

Only one part is constant, the other part is random. And winning 7 trades to 3trades is also result of randomness.. If you assumed 1:1 as you return/risk then why not assume 5 wins 5 losses?

But yes your right if you lost 7 and won 3 with a constant of 1:1 of course you wont make money.

fair point i did assume!

50/50 would give you a profit as long as your winners more more then your losers.

 

[lukeaye]

fair point i did assume!

50/50 would give you a profit as long as your winners more more then your losers.

But even that isnt a given! hahaha

So i guess that brings us back to the same point, Its all about a statistical edge, and any edge is good edge. I just think that mm is an easier edge to control. IMO.

 

[Mr J]

Well then what is random. Historically it has been defined with a probabilty, but foward anything can revert to random, or unexpected.

I would say that random, when people apply it to the markets, means that there is no bias, that at any point price is just as likely to go up as down, to do this or that. This would be treating it like a coinflip (let's assume coinflips are 50-50), where the only way to profit would be to get better than 1:1 on your money. However, the markets are driven by human behaviour, and that is often predictable. There are patterns, bias etc in the market, and therefore I don't think any definition of random could apply to the markets.

As for the future, I wouldn't say it is random, but unknown. Not the same.

Expectancy on historical results, yes. Expectancy is subjective. My example. A system has an expectancy of 70% over 30 years of data. Over the last 5 years it has an expectancy of 20%, and this year 55%. Which do you take as your expectancy?

I'm just saying that our edge is our expectancy. Without a positive expectancy, we have no edge. Yes, it is always changing (because conditions change, each trade is different etc) and varies from person to person.

So can we exploit this randomness in price movement, with MM, by allowing it to randomly distribute?

No. If it was truly random, then no amount of MM would give us an edge. If we consider price movement to be random, then each movement would be as likely to go up as to go down. This would be like flipping a coin, so do you think you could use MM to win at a coinflipping game?

I am also going back to saying that MM can't produce a profit. I allowed the possibility of it in my last post, but when I try to think of a profit produced by MM, it's always caused by another factor.

enjoying this discussion!

As long as we get ourselves thinking, any discussion is a good discussion. Not everyone agrees though!

But yes your right if you lost 7 and won 3 with a constant of 1:1 of course you wont make money.

In a random system such as a coinflipping game, any strategy that would give you 70% wins would lead to the losses being greater than the wins, and the net result still being 0.

 

[lukeaye]

No. If it was truly random, then no amount of MM would give us an edge. If we consider price movement to be random, then each movement would be as likely to go up as to go down. This would be like flipping a coin, so do you think you could use MM to win at a coinflipping game?

I am also going back to saying that MM can't produce a profit. I allowed the possibility of it in my last post, but when I try to think of a profit produced by MM, it's always caused by another factor.

I disagree, you are thinking in the wrong context.

I can use money MM to win a coin flipping game. Let me explain how.
Each time you take a position with the market you are making a deal?
Each position represents a deal. In this deal you will give up money (risk) if the deal goes against you. If the deal goes for you, you can close the deal whenever you like (reward).

Now you are allowed to have 100 seperate attempts at a coin flip. (The same as 100 separate stock positions).

You pick heads, and its tails. You lose -1R.

Next flip you pick correct, you take 1R. you flip again, another 1R. Flip again another 1R. Flip again, loss.

Next flip, you win 1R, next flip you lose -1R

Now you see what is happening? You are profiting from randomness by keeping your risk constant but having your profit uncapped, and at the hands of randomness. Sure you can still lose, but you have a statistical edge.

 

[Mr J]

I read the Jack Berstein thread, where you said this:

I also noticed that his buy on oct 27 sell nov 11 strategy had an expectancy over the last 30 years of 87%

Just to clarify, expectancy isn't the return we achieve (which are subject to variance), but our true edge (which is unknown to us - makes for interesting discussion).

I can use money MM to win a coin flipping game.

What if I suggested you try that in a casino?

Now you see what is happening? You are profiting from randomness by keeping your risk constant but having your profit uncapped, and at the hands of randomness. Sure you can still lose, but you have a statistical edge.

Would it be fair to say a simple version of this is to quit when we're -1, or +2? If so, consider that if we flip two coins, we have a 25% chance of winning +2, a 50% chance of losing -1, and a 25% chance of being even. The 2-1 doesn't get you anywhere. This can be applied to any R:R, and any number of flips. It's impossible to gain an edge here.

 

[Wysiwyg]

I use the word "probability". There is no guarantee that any expectancy will happen. Only probability based on repetitive human behaviour. Which is simply saying a majority agree on something.

 

[Mr J]

Two different things though, expectancy being the mathematical result when all factors are considered, and probability being one of those factors. I might be misunderstanding your post though?

 

[caribean]

Maybe aside from all the cold hard statistical edges, it is just an art?
Or, is art just an accumulation of past experiences providing subtle hints?
:silly:

 

[Wysiwyg]

Two different things though, expectancy being the mathematical result when all factors are considered, and probability being one of those factors. I might be misunderstanding your post though?

Yes a very valid point and well explained.

Any data tested in and out of sample is what happened in the past.This testing can prove up a positive or negative expectancy. We "work" the data and "the method" to attain a positive expectancy. We apply this positive expectancy model to present time trading and expect similar results to the in/out sample data testing.
Fair enough but if this were all it takes to be profitable within an expected time frame then everyone who has a positive expectancy model will be a winner.
This clearly isn`t so and that is because past performance is not indicative of future performance, but simply a general road map.

Check.

 

[Wysiwyg]

If trading was as simple as the alignment of x-y-z to profit then it would be a piece of cake but this is clearly not so. I expect this, I expect that. The markets don`t care what anyone expects and arguably can be spiteful by denying any "expectation".

I trade with probability and expect nothing. This below from the Amibroker walk forward test instruction.

We hope is that the parameter values chosen on the optimization segment will be well suited to the market conditions that immediately follow. This may or may not be the case as markets goes through bear/bull cycle, so care should be taken when choosing the length of in-sample period.

 

[lukeaye]

If trading was as simple as the alignment of x-y-z to profit then it would be a piece of cake but this is clearly not so. I expect this, I expect that. The markets don`t care what anyone expects and arguably can be spiteful by denying any "expectation".

I trade with probability and expect nothing. This below from the Amibroker walk forward test instruction.

Bingo.

Exactly what i am saying, expectancy can revert at any moment, it is dynamic and imossible to predict.

But what you can control is your mm!

My trading still revolves around probabilities though, but in a different way. I accept that given all my analysis i am still subject to the seemingly unexpected moves of the market.
I can identify points where i believe price will behave in a certain way given the statistics of past results, and hope that it behaves in the "expected" way. So i position myself at the lowest risk point. By positioning i still understand that in reality i have a 50/50 shot the position will come off. But if im wrong i will only lose a fraction of what i will let myself win. If a break does occur against me, then i will lose a very small percentage, but if i win i will let it run as long as i possibly can.

That there is my "mathematical" or "statistical" edge. Not my analysis, all im identifying is a point where i beleive a larege move will occur, for or against me i can never 100% say. But my MM allows me to be wrong more then im right and still turn a profit.

 

[Wysiwyg]

Bingo.

But what you can control is your mm!

Yes that is the only certainty we have in any trading event. In keeping an open mind --- there are trading systems that perform well in certain types of markets. Neville No Idea could buy stock in a bull market and make a decent profit.