Learning profitable automated trading

[N Rothschild]

are you actually ****ing retarded?

 

[oildaytrader]

are you actually ****ing retarded?

Thats all?

 

[N Rothschild]

why would i want to build a mechanical system when i have spent about 40 pages slagging you off for them?

im done replying to you, i think i and many others have proved your a complete spastic and any newbies reading this thread will realise that aswell.

enjoy your $1500 a year, dont spend it all at once!

 

[oildaytrader]

why would i want to build a mechanical system when i have spent about 40 pages slagging you off for them?

im done replying to you, i think i and many others have proved your a complete spastic and any newbies reading this thread will realise that aswell.

The trolls had an agenda and had been trying hard.

Al least I learnt something useful from Meanreversion

 

[wackypete2]

why would i want to build a mechanical system when i have spent about 40 pages slagging you off for them?

im done replying to you, i think i and many others have proved your a complete spastic and any newbies reading this thread will realise that aswell.

enjoy your $1500 a year, dont spend it all at once!

Wait, don't exit this thread yet. Let me stoke the fire some...

NR, did you hear what ODT says about your discretionary trading??
ODT, you gonna take that cr*p from NR??


OK...when you hear the bell come out swinging.

DING, DING...

Peter

 

[oildaytrader]

Wait, don't exit this thread yet. Let me stoke the fire some...

NR, did you hear what ODT says about your discretionary trading??
ODT, you gonna take that cr*p from NR??


OK...when you hear the bell come out swinging.

DING, DING...

Peter

He lost the battle of friendly banter on his journal , threatened to remove the posts where he got slagged upto pig with lipstick, and reverted to 4 letter words and insults.New traders should not listen to people claiming to be experts on automated trading , when none of the so called experts have ever designed a profitable system.

O D T

 

[oildaytrader]

There are many ratios used to judge a good automated system

CAGR

http://www.investopedia.com/terms/c/cagr.asp

Sortino ratio

http://en.wikipedia.org/wiki/Sortino_ratio

Sharpe ratio

http://en.wikipedia.org/wiki/Sharpe_ratio

The system has been updated and shows the following over 5.5 years of back tests 2003 to mid 2008, after deducting 1 pip slippage.

Pips profit 214477
investment 50, 000
drawdown 10,000 /20 %
29867 trades
average 7.22 pips per trade profit after slippage 1 pip.
Cagr 77 %
cagr drawdown 3.85

 

[oildaytrader]

Well, just read this thread from the start and wanted to point a few things out. ODT was trying to show/teach people some of his systems (all of which are apparently posted to be unprofitable - don't really see the point in that).

However along the way he has shown a lot of contradicting posts. Personally I think this is just the mods messing around .


Right, apparently investment bankers are using systems like this on MT4. This should have been a warning sign right at the start he was a BSer


Oh right, so investment bankers use unprofitable trading strategies to make their money! I don't really see the point in posting unprofitable systems, no knowledge to be gained apart from what to avoid!


Sorry??? Which one is it. Will the system become a self fulfilling prophecy or will it be a fight of latency?


Oh right, so what's that? 1 1/2 years without sleeping? Hang on, you spent $200,000?


Wait a minute there chap, you spent $200,000 - why not trade that $100 per point you were on about?


Glad we cleared that one up, so they won't become a self fulfilled prophecy they'll just fail hugely if he was to post them.


Wait, you spent $200,000 on systems, and could be trading $100 a point - but infact you only have $10,000? What's the point in spending 95% of your capital on system creation?!


Oh right, so that's why you have over 50 systems. To send out emails to C2 and Zulutrade. Now I understand why you have that vendor badge and didn't tell us why.


What about that $100 ODT? Thought you were making millions!


Wait a minute, I thought you said the bigger the size the more the system fails?


Again, why? It'd be more worthwhile to point someone to a pdf in MQL than posting a non-profitable system.


I thought you said previously that only amateurs don't include slippage?


This is easily overcome.


You can't be serious?

-------------------
Just a few from the start of this thread, some other threads and the posts created today are crackers. Especially the comment about being more sophiscated IB's. Also the post about each EA costing $500 to make - (200000/500 = 400 EA's to send out false signals ).

Anyways, ODT I hope you recover from your massive 95% drawdown from trading capital due to investing in system creation. Also, good luck on getting past your pocket money phase. I hope you and your "team" bring some more entertainment to trade2win.

Phil.

I did not want to give my own strategy in the EAS , but if somebody wishes to experience automation , they were there for free.There is free data available on the internet for backtesting.


PDF's of MQL are not really useful, given the useless coders hanging around forums , coding hack jobs.You need professionals to code properly.

Whatever else is not important.

 

[meanreversion]

There are many ratios used to judge a good automated system

CAGR

http://www.investopedia.com/terms/c/cagr.asp

Sortino ratio

http://en.wikipedia.org/wiki/Sortino_ratio

Sharpe ratio

http://en.wikipedia.org/wiki/Sharpe_ratio

The system has been updated and shows the following over 5.5 years of back tests 2003 to mid 2008, after deducting 1 pip slippage.

Pips profit 214477
investment 50, 000
drawdown 10,000 /20 %
29867 trades
average 7.22 pips per trade profit after slippage 1 pip.
Cagr 77 %
cagr drawdown 3.85

The testing stops mid 2008 again, why is that? For goodness sake, if we don't know how it performed during late 2008 and 2009, what value is it?

 

[N Rothschild]

none.

 

[meanreversion]

Hmm something to do with the decimal places changing, rite? Whichever, it's not statistically rigorous or robust. My guess is there would be a sharp fall off in performance during the Lehman mess.

 

[oildaytrader]

Hmm something to do with the decimal places changing, rite? Whichever, it's not statistically rigorous or robust. My guess is there would be a sharp fall off in performance during the Lehman mess.

I explained the data situation, unavailability of data after mid 2009 and changes to 5 decimal.If you are on a live account , you don't need back testing .Anyway remainder of 2008 was good as I was partly on live and witnessed the performance, 2009 I was on live account and witnessed the market conditions and performance.

I stopped trading nfp days with triple directions of price , which were the biggest losers of money, so actual can only look better

 

[meanreversion]

I explained the data situation, unavailability of data after mid 2009 and changes to 5 decimal.If you are on a live account , you don't need back testing .Anyway remainder of 2008 was good as I was partly on live and witnessed the performance, 2009 I was on live account and witnessed the market conditions and performance.

I stopped trading nfp days with triple directions of price , which were the biggest losers of money, so actual can only look better

Unavailability of data after 2009? Changes to 5 decimal? What ARE you talking about?

Find a way to get round it, otherwise it negates all your results. Besides which, five years data isn't enough anyway, you should make it 10.

 

[oildaytrader]

Unavailability of data after 2009? Changes to 5 decimal? What ARE you talking about?

Find a way to get round it, otherwise it negates all your results. Besides which, five years data isn't enough anyway, you should make it 10.

Alpari used to provide data but stopped providing it in 2009. Originally metatrader formatted data used to be available in four digits, brokers introduced 5 digits in 2008/9.Very little data was available in 5 digits.

http://thetrademachine.com/blog/200...ader-history-data-get-90-backtesting-quality/

We also back tested two more years from 2001 to 2002 , plus the the 5.5 years on the above tests equals 7.5 years back testing and 1.5 years of live accounts which overrides any backtesting.

What is more important than quantity is quality of testing.It is more important to test in periods of adverse market conditions, which the period tested covered,

 

[meanreversion]

Hmm, so in a couple of years time you will still be showing charts which end in May 2008? Won't that seem a little odd?

Get a different data provider, or re-code to allow for the change in decimal places (why this makes any difference I have no idea, but whatever..).

OR stop referring to out-of-date, redundant charts.

 

[oildaytrader]

Hmm, so in a couple of years time you will still be showing charts which end in May 2008? Won't that seem a little odd?

Get a different data provider, or re-code to allow for the change in decimal places (why this makes any difference I have no idea, but whatever..).

OR stop referring to out-of-date, redundant charts.

From 2010 I started collecting my own data from live feeds.

Does 2.5 years of data /backtesting really matter , when there have been forward testing and live accounts trading?On backtests you get instant fills , whereas on real markets there is liquidity issue , non fills and spread widening .None of this reflected in back testing.

I design my systems for future trading and trading in difficult adverse market conditions.

Here are two more years testing from 2001 to 2002.

 

[meanreversion]

You're giving mechanical trading a bad name with this half-baked approach.

Get a new data provider, or re-program.

YES the most recent data DOES matter, OF COURSE it does. In fact, it's MORE relevant than what happened 8-9 years ago.

So you started collecting your own data from 2010? So there's a gap from mid-2008 to start of 2010? Ridiculous.

I have two different data suppliers, so I am able to check the system using two different sets of inputs.. this gives me more confidence in the data.

If you're using only one set of data points which stopped 2 years ago, how can you possibly have confidence in the results you're getting??

 

[oildaytrader]

YES the most recent data DOES matter, OF COURSE it does. In fact, it's MORE relevant than what happened 8-9 years ago.

If you're using only one set of data points which stopped 2 years ago, how can you possibly have confidence in the results you're getting??

Confidence is gained from trading Live accounts and forward testing on small stakes.Live trading is far superior to back testing, and if system designer knows the logic works consistently , and logic is based on a winning formula of not trading in bad markets and trading with the edge, there is no need for more than a few years of back testing.Only a few months of back testing is required to tell whether your program can perform robustly or not, provided the developer has the knowledge to input the correct parameters.The problems arise with clueless amateurs designing mechanical systems , and then seeking data and back testing for 100 years to confirm whether they are doing the right thing or not.If the designer is a professional currency trader , the output will be professional.

 

[meanreversion]

There is so much wrong with your response, I don't know where to start.

1. "Live trading is superior to backtesting" Ok, why have you bothered to show any graphs if you think it's not really that important? You contradict yourself. (By the way, do you even know what "forward testing" means?)

2. "Only a few months of backtesting is required" Utter garbage. If your system worked well during the first 3 months of 2007 when vols were at an all time low, would you expect it to perform in the same manner during the last 3 months of 2008, when vols were extremely high?

3. "The problem arises with cluless amateurs designing mechanical systems" -- buddy, you're talking about yourself.

I truly hope that no-one takes anything you say seriously.

 

[oildaytrader]

There is so much wrong with your response, I don't know where to start.

1. "Live trading is superior to backtesting" Ok, why have you bothered to show any graphs if you think it's not really that important? You contradict yourself. (By the way, do you even know what "forward testing" means?)

2. "Only a few months of backtesting is required" Utter garbage. If your system worked well during the first 3 months of 2007 when vols were at an all time low, would you expect it to perform in the same manner during the last 3 months of 2008, when vols were extremely high?

3. "The problem arises with cluless amateurs designing mechanical systems" -- buddy, you're talking about yourself.

I truly hope that no-one takes anything you say seriously.

Here is the equity curve since Jan 1 2010, the market conditions have not been favourable but adverse and it has made a profit of 700 points Here are today's actual live trades .

If mechanical system designer is knowledgeable , he should be able to use the 3 months low volatility data and devise a profitable system coding to trade in low volatility and high volatility markets.He also needs knowledge of a fundamental currency trader.

I have seen plenty of amateurish systems on forex factory and other forums , rated 5 stars and million views .