Is Forex reliable?

I think he's referring to John Forman's research, but I'm not sure.

//----

hey hanover.... thanks..... john forman, the guy with the 19.95$ ebook.....

john forman's trade2win membership predates mine by 2 months..... it stands to reason 99.6% viewing his bio/profile vs some hayseed's would surely choses his views over mine....

so let it be..... my thoughts on his research will remain unheard.....
//----

fwiw, my profile will never included a download link to a 19.95$ ebook..... trading pays all the bills with room to spare......h

that forman dude is whack. There is enough research on the net that shows about 2% of daytraders make money with less than 1% making enough to justify trading.

I have never seen research done for swing and trendfollowing traders.

Well, you're partly right. I don't make currently make a living from trading forex. My goal is to contribute to the on-going development of an arsenal of profitable EAs, that complement each other as market conditions fluctuate. Writing 'dumb' EAs that are capable of outperforming 99.6% of human traders is indeed a challenging project, and it's one of several that my programmer/trader buddy and I are jointly working on, forex and otherwise.

Now to set the record straight: less than 5% of my income comes from writing software that's forex related. I do get a number of people offering to pay me to write indicators and EAs for them, but I turn them all away, because I find it much more interesting to develop my own stuff, which I also believe (whether rightly or wrongly) has greater long term potential. The bulk of my income happens to come from investments (some active, some passive) in other markets. So your veiled accusation that I make a living out of scamming people, by selling worthless EAs, is completely wrong.

In one of your posts you say that you don't trade forex yourself. Instead, all you seem to do is repeatedly quote one person's research into trader profitability. I guess that puts your views into some kind of perspective.
______________

@hayseed: I would value the views of a real hands-on trader, especially one as capable as yourself, over those of a 'passive' academic like Mr Forman, any day of the week.

@pemully: thanks for the info. Whether the failure rate is 99.6%, 98%, 95% or whatever, the stats show that making money from forex is indeed challenging.

Now I must bow out of this thread, and get back to my projects.

I think the essential question is "Am I reliable?".

A few facts should answer your question:
1. Spot FX is seen as an entry market for beginners by trading firms.
2. There is a reason why they choose eg. futures which can be same volatile and there is plenty of opportunity there - Due to its decentralization and lack of regulation at source is pron to scamming brokers, who do very well out of 95+% loosers.
4. Most of FX brokers will switch you to direct market access only at extremely volatile periods eg. news, speeches, so they lower the risk of loosing money IN CASE you win and earn measly spread. Otherwise they take all transactions onto their own books - which means: they are quite confident you will loose money.
5. Why would brokers lend you 500 times more money than you have /high leverage/ while credit crunch? It doesn't happen on any other market!
6. Most popular soft given for free by brokers: Metatrader had been created by Soviets as a casino software initially - easy to manipulate with appropriate add ons.

...now...

good luck with your trading :-)

I completely agree with this.

Everyone keeps blaming external factors for their own lack of success.

As long as you believe that you're losing because "everyone loses", or "the game is rigged against you" and "the brokers set it up so that blah blah blah is against you" or [insert excuse here] - you will continue to lose.

Why not start asking what makes successful traders successful and focusing on that? That is if you actually want to be successful...

You get what you focus on - if you want arguments and drama and ego battles and to blame [insert external factors here - brokers/big money/'manipulators etc], you can certainly get that here.

If you want reliable consistent success it's there if you seek it out and stick at it.

Your choice, it's your money, time and life...

The answer is very simple

1) if you gamble it is the most riskiest
2) if you trade it is safe

In case anybody was reading it, I just deleted my last post, after reading that the EA I was commenting on is actually a kind of martingale. Hence I would not run it on a live account, under any circumstances.

hanover Hi,
need your opinion and guidance if you spare some time.


This is the EAs im testing from 20/8. They are index plays, which I developed myself with a good programmer friend. It made 35.000 pippers and DD is relatively stable. Im reading a lot of probability and im digging into statistics. Have several accounts and all look delightfull.


On this account I have 5 EAs which each has 6 rules to trade. I am doing an error rates on all of them inside the EAs. Have found out that a lot of rules inside have very poor decisions. Under 45% percent. Primary R:R is 1:2,25, but with overall 50% chance of winings my True R:R is at 1:1,125.


With your experiance, can you tell me whats your take on this.


What statistically can you recommend additional to play around error rate?
I can not backtest, cus Im grabbing all the market and backtesting would be really painfull if I decide to work it in excel. I trade 21 pairs. 7 Currencies.


Can you say something about accounts which have 50% rate of winnings. Of course, primary R:R is over 1:2 too 1:3. How they transform in the longterm.
Im beeing told, that I need to find only rules inside this system with % winings over 65%. I have a range of decisions in this traded history of less than one month that are sitting on 92-75 % of winnings. Im gonna test forward to see where they gonna land.


What can you say about a trade count and when one can say the system is relatively good. I read lots of things about your reading the summary of trade report. With profit factor over 1,5 and DD..

EDit: as its index trading its definately strongly correlated system.


Thanks for any direction.

21vs7,

OK, I will do my best, for whatever my opinion is worth.

IMO its important to know the strengths and weaknesses of a strategy: under what market conditions it will prosper, and under what market conditions it will struggle. (For example, with the EA I described in post #34, it will make money if EU 'trends' far enough, often enough, during a certain period of time; it will lose if there are too many prolonged sideways periods. However, there are also edges with the entries that help to alleviate this). Hence you would need to evaluate your strategy, and make a similar 'robustness' analysis. Also, make sure that your strategy is not falling into any of these traps.

You haven't said a lot about your MM. Consistent position sizing is important IMO, as MM in itself can not provide a lasting edge.

Statistically, your results are excellent. You have a PF of 1.82 over 2600+ trades. When you say 20/8, do you mean 20 Aug 2013? If so, it has made 2600 trades in about 3 weeks? Wow, that is pretty crazy! If you've read my thoughts here, you'll know that I like to see an EA perform well for several months, preferably years, to gain a feel for possible best and worst case scenarios, and also to feel comfortable that it's on-balance profitable across the widest possible range of market conditions. Personally, I like the fact that your average win (5.70) > your average loss (3.32). IMO this means that the win rate of ~ 50% is perfectly OK. The PF is the product of win rate and RR (ave win/ave loss), and takes all of this into consideration. A PF of 1.82 over 2625 trades is very good. The worst case drawdown of only 6% is excellent also.

I've read some of Daniel Fernandez's books and blogs. IMO he is as good as any, when it comes to evaluating automated strategies. One interesting metric he uses is dividing the annualized return by the worst case drawdown. (I call this the 'Fernandez ratio', or F-ratio ). For example, if your automated system is returning 30% p.a. and the worst DD is 15%, then it has a F-ratio of 30/15 = 2. Anything above 1.5 is very good; virtually no EAs have a F-ratio > 3, if run over a long enough period. (Note to everybody: remember, we are talking about automated systems here, not discretionary traders, who should be able to easily outperform 'dumb' EAs).

Hope that gives you some ideas. I don't know the specifics of how your strategy operates, hence I can't really comment further.

Good luck,
David

Thanks mate,
Ill reply with more data td when I digest all you have written and linked. I appreciate

EDIT: I was writting answer for half an hour and when I tried to post it, it was canceled. Damn it.

IMO its important to know the strengths and weaknesses of a strategy: under what market conditions it will prosper, and under what market conditions it will struggle.
Hence you would need to evaluate your strategy, and make a similar 'robustness' analysis. Also, make sure that your strategy is not falling into any of these traps.


I think I need trends and volatility. I need 2 index trends (300 pips and more direction). 2 Index trends meaning: 2 currencies are gaining. 2 currencies are loosing. 1 is strong and one is weak. Something out of this...AND ON THE BOARD.


About the traps. Yes. What it still does is open sell and buy. So, will limit it to close a buy, when it gets a sell.
Second trap you talk about is: that it is definately opening a lot of positions on the same cross and around the same entry. That needs to be optimized also.

Floating PL. Yes, for sure. Im seeing a lot of accounts which have a lot of floating equity. But dont manage to get out and cap them. Im seeing/in a way calculating the maximum limits of the range where signals could cap the move. Im seeing that those who cant cap the move TPs are wider of course..But in a way with all the strategies I could find the optimum levels of pips where TP should occur. High probability. Same goes for SL.





here are some trades I put on when it was +20k pips. Maybe you see something more...

http://www.forexfactory.com/showthre...77#post6926777



You haven't said a lot about your MM. Consistent position sizing is important IMO, as MM in itself can not provide a lasting edge.

Yes, I just wanted to grab as many data as possible so this issue Is gonna come at the end. Surely I will always use consistent size. All setups the same. ATM using 0,01 as said. But as it has turned out it was pretty much valid to the account base. Will surely get more into this when I decide it forward what to do.


Statistically, your results are excellent. You have a PF of 1.82 over 2600+ trades. When you say 20/8, do you mean 20 Aug 2013? If so, it has made 2600 trades in about 3 weeks? Wow, that is pretty crazy!

Three weeks only yes. Trading like crazy It actually has 6 rules inside the EAs. And five EAs. So, one EA can have max 252 positions which makes 1260 for 5 EAs. Im collecting data, surely It wont be that much at the end.


If you've read my thoughts here, you'll know that I like to see an EA perform well for several months, preferably years, to gain a feel for possible best and worst case scenarios, and also to feel comfortable that it's on-balance profitable across the widest possible range of market conditions. Personally, I like the fact that your average win (5.70) > your average loss (3.32). IMO this means that the win rate of ~ 50% is perfectly OK. The PF is the product of win rate and RR (ave win/ave loss), and takes all of this into consideration. A PF of 1.82 over 2625 trades is very good. The worst case drawdown of only 6% is excellent also.


As you can see in the first attachement I calculate every rule for its % winings, plus I calculate the true R:R per risk. At least I see it this way...

Lets say: primary R:R is 1:3 and we are 50% victorious on deals. I calculate ((wins)50*3)/100=1,5. So the true R:R is 1:1,5 when one is 50% right.

Where is the edge can one calculate with this and still be profitable. A lot of rules end up going over 1,3 true R:R and are only over 42% accurate. Still, they make a lot of pips.


And you can see the numbers in the attachment. first is % winings = 0,55/1,20 the second (1,20) is true R:R


I attached also one other account with 1:3 primary R:R. it has 40% chance of win..


And attached the best winings account with primary 1:1 R:R. Which takes the actual down of course.


virtually no EAs have a F-ratio > 3, if run over a long enough period.

Im gonna try and beat that

thanks mate

You can make money on forex market if you manage your risks like with any other business. I am trading part time so that helps me a lot to keep me less emotional when I have a loss or losing streak.

Not much more that I can add, without knowing a lot more about your system. IMO it's very important to have a good feel for system strengths and weaknesses, and to feel comfortable trading it (even if it's automated).

IMO the key performance metrics are:

Number of trades/number of months testing: important for statistical validity, i.e. high enough to prove that the results are not 'luck', and can be maintained over a long period of time.

Profit factor: obviously the higher the better. 1.0 is breakeven. Above 1.5 with an automated system is excellent. PF by dividing the winning pips by the losing pips is even better IMO, because it tests the efficacy of your 'raw' system (entries+exits), by taking MM out of the equation. PF takes both win rate and RR (ave win/ave loss) into consideration.

Worst case drawdown: a low WDD allows you to safely increase the risk per trade, e.g. from 1% to 2% (fixed frac sizing).

F-ratio: for example, at 1% risk per trade, annualized return (AR) is 12%, WDD is 5%. Suppose you're personally comfortable with 20% WDD, then you can safely increase your pos size to 4% risk, thus likewise multiplying AR and WDD by a factor of approximately 4. However, consider leaving a margin for error, just to be safe, in case the system doesn't perform as well as it did historically (also, you may get more slippage on a live account).

Note: AR = (((deposit+netpl)/deposit)^(1/years)-1)*100
where
deposit = the initial deposit;
netpl = the net profit/loss generated by trades;
years = the number of years over which performance has been measured, e.g. 5 years 7 months would be 5+7/12 = 5.583 years

R:R: in general, and everything else being equal, the higher the RR (ave win/ave loss), the better, i.e. I regard RR as being more important than win rate. But this also depends on the nature of the system (for example, your screenshots show that the 1:1 method actually has a higher PF than the 1:3 method).

Shape of the equity curve: a picture paints a thousand words! When examining the curve, the obvious question is: Are you personally happy with the consistency?

It looks like you have an excellent system. IMO all that's needed is for it to be tested over a period of several months (backtesting is sufficient, if you have good quality data); then, when you are satisfied that it's capable of performing consistently under all market conditions, it's time to go live, with whatever size account you feel comfortable with.

Hanover, thanks for your time. I really appreciate. Im gonna again try to digest all this and report later on..

It's reliable when you have some exercise as well as know the market industry, also an excellent broker if you do not want to do it alone. In this market you need more than money, you will need to like, use a plan, contain the right equipment, etc.

Proper knowledge and experience is necessary for the trader and it helps traders to achieve their target of earning good and consistent profit. Forex market is risky also but by making and following some good strategies we can minimize the loss factors.

Need to update the stuff I started.

Im not gonna calculate error rate of different scenarios inside the Each EA im running here but decided im gonna do that at the end of the year.


I had to restart the DEMO account cus it was showing only few orders back in history so could do error calculations anymore. Seems this accounts will also have a problem. Im using FXCM. And yes, I picked acc history, all history.

Well, to comment.

First picture is of one EA only. Have 6 rules. interesting graph. 2 months of play.

Cant really tell if this is good line up or not . EA made +5k pips and is relatively stable.

Second has 8 EAs for 1 month.

+8k pippers. also relatively stable. need to calculate error rate from all of them scenarios.
Third has 5 EAs for 1 month play.

+3k pippers in less than a month. also need to calculate the error rates.


From all in all I am super satisfied with all the results. Still, on the end of the day I still see it really trading too much and I need to limit the trades somehow. Reading a lot this days and trying to figure out what next steps are.

Planning in this direction:

1. Filter trades
Too much risk on the table. Need to filter that somehow..lots of risk also on one currency index. Need to figure which,how,why,..etc.

2. Dynamic SL and TPs.
Trying to let runners win moo. Also combined with some index calculations. Also dont know how im gonna fix that.

So, looots of questions still in place, even thou the green is the marketplace.

Losers will always say forex is not reliable but if you know what you're doing it's very reliable. Learn how to use a position calculator and never risk more than 10%. If you are beginner never risk more than 5% or use demo.

As long as support and resistance will work in forex it will be possible to be a winner. Not all S/R are the same, some are very minor some are very major. Trade only the really major ones when starting. Trade into space not into traffic. Don't forget the more price tries to break a S/R the weaker that S/R becomes. Wait for price to break it and then after it makes a rounded retest trade it (either directly with touch trading, either wait for confirmation). Put more emphasis on horizontal levels. Trend lines are not so good like horizontal levels. Don't trade before major news. Don't use any indicator (most of them are totally useless) only S/R that can be seen from a naked chart.