How to make a living with trading? | Page 15 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

Do you make a living trading small timeframes? 58 replies

Do people actually make a living off of trading? 38 replies

Living in China, trading for a living 18 replies

(can) small accounts make a living ?! 611 replies

How can I make a living trading forex? 25 replies

Attachments: How to make a living with trading?
Exit Attachments

How to make a living with trading?

  • #281
  • Mar 28, 2017 9:56am Mar 28, 2017 9:56am
Quoting ensale
{quote} The one sided view is strong in this one too!

My view is based on things I can apply the Scientific Method to, any testing of feelings have been found no better than random choices, while it is possible that your feelings are some kind of subconscious decision making, it is pretty much impossible to test that and as such in the absence of any facts I reject the idea that they work.

I cannot prove that they don't work, but that is not the same as you proving that they do. My methods are soundly rooted in testable probability, that works for me, I wish you luck of course with your particular trading style.
I live entirely from the growth of my personal trading account
  • #282
  • Mar 28, 2017 10:50am Mar 28, 2017 10:50am
Quoting guybaker
{quote} My view is based on things I can apply the Scientific Method to, any testing of feelings have been found no better than random choices, while it is possible that your feelings are some kind of subconscious decision making, it is pretty much impossible to test that and as such in the absence of any facts I reject the idea that they work. I cannot prove that they don't work, but that is not the same as you proving that they do. My methods are soundly rooted in testable probability, that works for me, I wish you luck of course with your particular...
you can simple backtest the strategy manuelly. its fast if you know how to to and save when you know how to do it correct without cheating you.

a program is normally slower programmed and backtested then done manuelly backtested. a program is bether when you dont have a real big edge and need to optimize your program. if this is not neccessary then manuel is faster ( if you dont use very complex strategies, but i have never seen a complex strategiy superior to an easy).

so the Scientific reason is normally no very important reason for a strategy programmed (if you dont need optimization ), there are some good reasons, but this is often 1 of the smaller. if you start trading then this reason is more important for you. because i dont know if you are beginner or medium trader, i cant evalute so how usefull it is.
  • #283
  • Mar 28, 2017 10:53am Mar 28, 2017 10:53am
Quoting Artcool
All strategies, whether manual or fundamental, can be replicated by a bot. If a trader doesn't try to code his strategy, that could be : He doesn't know how to code and doesn't feel comfortable in giving the rules to some coder He is afraid to see that his strategy is not profitable in the long term which will destroy his confidence and bring him back to step 0. He is bored in his life and unconsciously wants to keep himself busy every day. His ego wants him to believe that he has some special power that will keep him out of bad trades.
Lol okay the most untrue statement I've ever read. Not all strategies can be replicated by a bot I guarantee you that
  • #284
  • Mar 28, 2017 11:35am Mar 28, 2017 11:35am
Quoting guybaker
{quote} IMHO Discretion is either 1) subconscious logic which could in principal be understood and coded, or 2) Luck. There is no magic in trading or anywhere else. My broker is IG.com, I doubt even if I were making 1m a month they will care, they are primarily interested in volume they make their money from the spread or in the case of Binaries the margin between buy and sell prices.
Hi Guy,

Well said, both statements are as true as can be, to me discretion is equal to intuition which in turn is subconscious logic in terms of my brain recognizing a chart pattern or setup it has analyzed a thousand times before and learned that the probability of the price moving in one particular direction is greater than the probability of movement in the opposite direction.

Anyway, trading with IG, do you use PRT and if so are you active on the prorealcode forums?
  • #285
  • Mar 28, 2017 11:39am Mar 28, 2017 11:39am
Quoting guybaker
{quote} My view is based on things I can apply the Scientific Method to, any testing of feelings have been found no better than random choices, while it is possible that your feelings are some kind of subconscious decision making, it is pretty much impossible to test that and as such in the absence of any facts I reject the idea that they work. I cannot prove that they don't work, but that is not the same as you proving that they do. My methods are soundly rooted in testable probability, that works for me, I wish you luck of course with your particular...
I agree with that.. the sub mind defintley needs more research and we don't know everything about our brain yet. but agree.
It's in fear when people show what they truly value.
  • #286
  • Mar 28, 2017 1:04pm Mar 28, 2017 1:04pm
Quoting dukas_trader
{quote} you can simple backtest the strategy manuelly. its fast if you know how to to and save when you know how to do it correct without cheating you. a program is normally slower programmed and backtested then done manuelly backtested. a program is bether when you dont have a real big edge and need to optimize your program. if this is not neccessary then manuel is faster ( if you dont use very complex strategies, but i have never seen a complex strategiy superior to an easy). so the Scientific reason is normally no very important reason for a...
In my experience it is pointless back testing manually because of what is called confirmation bias. Also it is not possible to test how one would have felt in the past
I live entirely from the growth of my personal trading account
  • #287
  • Mar 28, 2017 1:06pm Mar 28, 2017 1:06pm
Quoting guybaker
All wise words. an additional option is if possible code your method, let it run with minimal effort and go and work a job to build up your account.
@PiranoTrader

Pretty much what I said about working at the zoo
Demo only
  • #288
  • Mar 28, 2017 1:10pm Mar 28, 2017 1:10pm
Quoting cfta
{quote} Hi Guy, Well said, both statements are as true as can be, to me discretion is equal to intuition which in turn is subconscious logic in terms of my brain recognizing a chart pattern or setup it has analyzed a thousand times before and learned that the probability of the price moving in one particular direction is greater than the probability of movement in the opposite direction. Anyway, trading with IG, do you use PRT and if so are you active on the prorealcode forums?



All of my execution is done via my proprietary software running on servers in London. I use the IG web platform for account monitoring and administration only at this point.

I looked at PRT in the past and for manual trading it looks quite good, it is also programmable, but it has some limitations in that regard and also the code is stored and executed server side which in a way is good, but in another way it means you are giving up your method to the service provider.
I live entirely from the growth of my personal trading account
  • #289
  • Mar 28, 2017 1:15pm Mar 28, 2017 1:15pm
Quoting Artcool
All strategies, whether manual or fundamental, can be replicated by a bot. If a trader doesn't try to code his strategy, that could be : He doesn't know how to code and doesn't feel comfortable in giving the rules to some coder He is afraid to see that his strategy is not profitable in the long term which will destroy his confidence and bring him back to step 0. He is bored in his life and unconsciously wants to keep himself busy every day. His ego wants him to believe that he has some special power that will keep him out of bad trades.
Very true. Describes me less than a year ago. Recovery is long from these symptoms.

EDIT: You are not seriously trading from North Korea, arent ya
Demo only
  • #290
  • Mar 28, 2017 1:21pm Mar 28, 2017 1:21pm
Quoting guybaker
{quote} In my experience it is pointless back testing manually because of what is called confirmation bias. Also it is not possible to test how one would have felt in the past
oh, you are very new in trading?

its very easy to test without cheating yourself, if you know some basics. but this you will learn over the time. in metatrader you have some possible ways, in your frontend i dont know (ninjatrader or selfcoded .net??).

if someone cant backtest manuell, you will not be able sometimes develop a good working strategy, you need to be stick often with some working or medium working strategies, i did see this kind of arguments to often from programmers, after i showed them the proper way they chnaged all their arguments forever. how you will find possible strategies that give you a drawdown at max of montly win, when you cant backtest manuelly fast in proper way? maybe possible but needs long time (if all need be programmed), but difficult without math and basic knowledge in proper manuell backtest.

or are you only a quant statistical edge trader? thats different, then of course because of your optimization you need a program, but this programs have less wins/drawdown ratio in most time.

so all together: its very easy to test in proper way, even for the most unexperienced or most emotional self cheating person, if you learn the basics how to backtest manuell in proper way. its worth to train this, it helps you a lot, more important then improve your programming skill. its 1 of the basics someone needs to know for create a stratgy in fast time and most save.
  • #291
  • Mar 28, 2017 1:28pm Mar 28, 2017 1:28pm
Quoting dukas_trader
{quote} oh, you are very new in trading? its very easy to test without cheating yourself, if you know some basics. but this you will learn over the time. in metatrader you have some possible ways, in your frontend i dont know (ninjatrader or selfcoded .net??). if someone cant backtest manuell, you will not be able sometimes develop a good working strategy, you need to be stick often with some working or medium working strategies, i did see this kind of arguments to often from programmers, after i showed them the proper way they chnaged all their...

Well I have been trading for 10 years plus. I have done a lot on back testing manually, and then run the same test with my computer, it is always a long way out.

If you can back test without confirmation bias, well done, but for me I prefer to use a computer because as well as being more accurate it is much much faster, indeed I like to test 10,000+ setups, doing that manually is an insurmountable task. Yes I am Quant style trader.
I live entirely from the growth of my personal trading account
  • #292
  • Mar 28, 2017 1:48pm Mar 28, 2017 1:48pm
Quoting guybaker
{quote} Well I have been trading for 10 years plus. I have done a lot on back testing manually, and then run the same test with my computer, it is always a long way out. If you can back test without confirmation bias, well done, but for me I prefer to use a computer because as well as being more accurate it is much much faster, indeed I like to test 10,000+ setups, doing that manually is an insurmountable task. Yes I am Quant style trader.
ok, when you need optimization cycles for get a good result you need a program, thats not possible in a good time in manuell testing.
but more then 1 setup is crazy, you lose many time when not use includes for backtest setups (optimizations) automatic result printouts. this optimization inside the frontends is there for beginner programmer, that they can play with it, but put all inside the program is a must for saving years and have much time.

starting one time/cycle and get as result ( and all others in file) whats best (and has biggest good result area inputs combinations around) its more easy then start many time with different settings.
  • #293
  • Mar 28, 2017 1:59pm Mar 28, 2017 1:59pm
Quoting dukas_trader
{quote} ok, when you need optimization cycles for get a good result you need a program, thats not possible in a good time in manuell testing. but more then 1 setup is crazy, you lose many time when not use includes for backtest setups (optimizations) automatic result printouts. this optimization inside the frontends is there for beginner programmer, that they can play with it, but put all inside the program is a must for saving years and have much time. starting one time/cycle and get as result ( and all others in file) whats best (and has biggest...
When I say setups I mean simulated trades over as much data as I can get, I like to simulate 10,000+ trades to establish whether an edge is reliable.
I live entirely from the growth of my personal trading account
  • #294
  • Mar 28, 2017 2:17pm Mar 28, 2017 2:17pm
Quoting guybaker
{quote} When I say setups I mean simulated trades over as much data as I can get, I like to simulate 10,000+ trades to establish whether an edge is reliable.
ah ok, i do it different way. because of currencies had bigger spreads many years ago and did come smaller over the years, its impossible to test very shortterm strategies in this time. i like do backtest not more then 3 years.

because i only program what i allready used in live trading a backtest its for me more to find errors in coding or news that should be avoid.
  • #295
  • Mar 28, 2017 2:23pm Mar 28, 2017 2:23pm
Quoting dukas_trader
{quote} ah ok, i do it different way. because of currencies had bigger spreads many years ago and did come smaller over the years, its impossible to test very shortterm strategies in this time. i like do backtest not more then 3 years. because i only program what i allready used in live trading a backtest its for me more to find errors in coding or news that should be avoid.

Yeah I see your point about spread. I test on bid price data, and simulate the costs for now based on my brokers current spread, not perfect, but not bad.
I live entirely from the growth of my personal trading account
  • #296
  • Mar 28, 2017 2:57pm Mar 28, 2017 2:57pm
Quoting guybaker
{quote} Yeah I see your point about spread. I test on bid price data, and simulate the costs for now based on my brokers current spread, not perfect, but not bad.
depends on the years, if you use currencies and go in last 10 years ago, then you get bad results. if i would test scalping experts in this old years i would get 100000-10 win/ risk per year. because you could scalp with actual spread inside the spread 10 years ago. before even more worst. so i would not test it more then 5 years at max. bether maybe 3-4.

for long term trades its not that important. for this i test allways with 3 times the spread i have, so i have space for higher spread time, increased spread because of higher volume, commission, slippage, higher rollover and so on...... but for short term trades i would only have insane big wins in backtests that will kill the real result complete.

also i compensate after the results the wins from news time and wekkend gaps . i delete the wins that wasnt there in real because of the high spread and put in the losses that was there because of higher spread...
  • #297
  • Mar 28, 2017 3:00pm Mar 28, 2017 3:00pm
Quoting dukas_trader
{quote} depends on the years, if you use currencies and go in last 10 years ago, then you get bad results. if i would test scalping experts in this old years i would get 100000-10 win/ risk per year. because you could scalp with actual spread inside the spread 10 years ago. before even more worst. so i would not test it more then 5 years at max. bether maybe 3-4. for long term trades its not that important. for this i test allways with 3 times the spread i have, so i have space for higher spread time, increased spread because of higher volume, commission,...

I agree that more recent years are likely to be more reliable and I give them more weight, but I see no harm in running the test on older years.
I live entirely from the growth of my personal trading account
  • #298
  • Mar 28, 2017 3:04pm Mar 28, 2017 3:04pm
Quoting dukas_trader
{quote} depends on the years, if you use currencies and go in last 10 years ago, then you get bad results. if i would test scalping experts in this old years i would get 100000-10 win/ risk per year. because you could scalp with actual spread inside the spread 10 years ago. before even more worst. so i would not test it more then 5 years at max. bether maybe 3-4. for long term trades its not that important. for this i test allways with 3 times the spread i have, so i have space for higher spread time, increased spread because of higher volume, commission,...

This for my latest system is tested back to 2003, the results look remarkably consistent
Attached File(s)
File Type: pdf Zen Testing.pdf   62 KB | 425 downloads
I live entirely from the growth of my personal trading account
  • #299
  • Mar 28, 2017 3:16pm Mar 28, 2017 3:16pm
Quoting guybaker
{quote} This for my latest system is tested back to 2003, the results look remarkably consistent {image}
how will be the trades structure?
in what currency, how long is the average holding time, what trading amount, what fees (spread, commission, rollover) is used in what trading time, how many trades per year, trades, was there bigger wins in news or weekend gaps?

my long term trading systems i can use in 2003 too, but my scalping experts i cant, they would have 5 million win from a 10.000 account in the first year because of the spreads and real it was maybe 20-100k (depends from the years). so extrem difference.
  • #300
  • Mar 28, 2017 3:25pm Mar 28, 2017 3:25pm
Quoting andddres1
try Topsteptrader, a trading combine test costs 120usd the cheapest for one time if you pass, then they found you with 30k

just do this.

if you "have it" you'll make enough

If you don't "have it" you'll still do what you're doing today and live in the internet infamy.


/thread
0 traders viewing now