2009.09.10 DRAIN THE BANKS - LIKE A RAT

free & uncensored discussion arena for TheRumpledOne

Moderator: moderators

GeorgeCh
rank: <50 posts
rank: <50 posts
Posts: 10
Joined: Fri Dec 31, 2010 10:59 am
Reputation: 0
Location: Geneva
Gender: None specified
Contact:

Postby GeorgeCh » Mon Jan 03, 2011 11:01 pm

So, I just took the system on a two-day backtest with a twist.

I manually scrolled bar-by-bar on 5M TF through EUR/USD on Dec 30 and Dec 31 (and yes, I realize this is not an exhaustive sample size by any stretch of imagination).

The idea was simple - by scrolling through the charts and making decisions on the basis of the information available to me at the time - rather than retroactively eyeballing through the available data - I could put myself in the shoes of a trader using the system with the element of uncertainty thrown it.

I used the following rules:

1. Wait for the price to make a new high.
2. On 5M TF, wait for a red candle to close.
3. On 5M TF, enter on the next candle as soon as the price reaches the previous candle's low.
4. SL: 10 pips; TP: fixed 5 pips (to eliminate the discretionary decision-making element).

System performance (1 is win, 0 is loss):


1
1
0
1
0
1
1
1
1
1
1
1
1
1

In short, the system produced 12 wins and 2 losses over the two-day period.

Depending on how aggressive you were with risk management, you would make at least 4.05% return over the two-day period - assuming that your SL is equal to 1% of account, and your TP is equal to 0.5% of account.

Should those values be increased to the recommended 2% SL and 1% TP, the return would be 8.22%.

And, just to throw it out there, were you crazy enough to do 5%/2.5%, you'd end up with a whooping 21.37% gain.

Now, obviously, this sample size cannot be representative of the population; more to the point, the winning streak probably skews the results somewhat.

On the other hand, it should be noticed that one of the losers could have been turned a winner by exiting at 5 pip profit, and that at least two winners could have been turned into even greater winners by following the subsequent downtrend.

Before I take this system for a live spin, I'll run more such tests and revert back to you.

Please add www.kreslik.com to your ad blocker white list.
Thank you for your support.

User avatar
PTG
rank: 500+ posts
rank: 500+ posts
Posts: 645
Joined: Thu Jan 08, 2009 2:24 pm
Reputation: 2
Gender: Male
Contact:

Postby PTG » Mon Jan 03, 2011 11:44 pm

GeorgeCh wrote:So, I just took the system on a two-day backtest with a twist.

I manually scrolled bar-by-bar on 5M TF through EUR/USD on Dec 30 and Dec 31 (and yes, I realize this is not an exhaustive sample size by any stretch of imagination).

The idea was simple - by scrolling through the charts and making decisions on the basis of the information available to me at the time - rather than retroactively eyeballing through the available data - I could put myself in the shoes of a trader using the system with the element of uncertainty thrown it.

I used the following rules:

1. Wait for the price to make a new high.
2. On 5M TF, wait for a red candle to close.
3. On 5M TF, enter on the next candle as soon as the price reaches the previous candle's low.
4. SL: 10 pips; TP: fixed 5 pips (to eliminate the discretionary decision-making element).

System performance (1 is win, 0 is loss):


1
1
0
1
0
1
1
1
1
1
1
1
1
1

In short, the system produced 12 wins and 2 losses over the two-day period.

Depending on how aggressive you were with risk management, you would make at least 4.05% return over the two-day period - assuming that your SL is equal to 1% of account, and your TP is equal to 0.5% of account.

Should those values be increased to the recommended 2% SL and 1% TP, the return would be 8.22%.

And, just to throw it out there, were you crazy enough to do 5%/2.5%, you'd end up with a whooping 21.37% gain.

Now, obviously, this sample size cannot be representative of the population; more to the point, the winning streak probably skews the results somewhat.

On the other hand, it should be noticed that one of the losers could have been turned a winner by exiting at 5 pip profit, and that at least two winners could have been turned into even greater winners by following the subsequent downtrend.

Before I take this system for a live spin, I'll run more such tests and revert back to you.


George, have you taken the spread (bid/ask differential) into account ?
This is my new signature: "new signature".

GeorgeCh
rank: <50 posts
rank: <50 posts
Posts: 10
Joined: Fri Dec 31, 2010 10:59 am
Reputation: 0
Location: Geneva
Gender: None specified
Contact:

Postby GeorgeCh » Mon Jan 03, 2011 11:56 pm

PTG wrote:George, have you taken the spread (bid/ask differential) into account ?


Empirically, no; visually, in all of the above cases, it would not have made a difference, in that a winner would still be a winner (in other words, the difference wasn't substantial).

In the test that I am doing right now, however, I treat winners that win by one pip as losers, simply because 1 pip is the minimum spread one can expect on this pair.

I can also tell you already that a fixed 5 pip TP is NOT profitable; it worked fine in the above scenario, but the system is getting slaughtered in early September. To avoid this, you really need to use TRO's "take what you can get" approach to exits - trying to let winners run and cut losers asap.

Unfortunately, in the testing, I have seen several instances where a trade went dangerously close to TP only to revert in your favour (sometimes even after making a new high, as long as that new high fell into the 10 pip TP allowed). This would call for more discretion in cutting losers prematurely.

Alternatively, settling for three pips per trade would have turned at least a few of the losers into winners; however, this would have to be weighted against the ensuing greater impact that each losing trade will have on your performance.

None of this should be interpreted as a statement that the TRO method is in any way deficient; this is simply an attempt to better understand the exit strategy to optimize results.

GeorgeCh
rank: <50 posts
rank: <50 posts
Posts: 10
Joined: Fri Dec 31, 2010 10:59 am
Reputation: 0
Location: Geneva
Gender: None specified
Contact:

Postby GeorgeCh » Tue Jan 04, 2011 12:00 am

Actually, over the course of this week, I will try to compile an Excel spreadsheet detailing which entries I took and what the outcome was. It will also enumerate how far the winner would have run had it not been closed at 5 pips and how far into profit the loser went before reversing and turning into a loss.

User avatar
TheRumpledOne
rank: 10000+ posts
rank: 10000+ posts
Posts: 11440
Joined: Sun May 14, 2006 9:31 pm
Reputation: 45
Location: Oregon
Real name: Avery T. Horton, Jr.
Gender: None specified
Contact:

Postby TheRumpledOne » Tue Jan 04, 2011 2:15 pm

Image

Are any of you GREEN RATS feasting??
IT'S NOT WHAT YOU TRADE, IT'S HOW YOU TRADE IT!

Please do NOT PM me with trading or coding questions, post them in a thread.

Please add www.kreslik.com to your ad blocker white list.
Thank you for your support.

GeorgeCh
rank: <50 posts
rank: <50 posts
Posts: 10
Joined: Fri Dec 31, 2010 10:59 am
Reputation: 0
Location: Geneva
Gender: None specified
Contact:

Postby GeorgeCh » Tue Jan 04, 2011 3:13 pm

Had a great time on EUR/USD today as a red rat - took three profitable trades.

The two key lessons of the day are:

1. Let the winners run - really, I mean it. Each and every time I closed a trade at 4-5 pip profit after spread, I could've made thrice as much had I held on to it for longer.

2. Stick to the rules - it becomes incredibly tempting to bend the rules and start trading swing lows (since, as a green rat, you're already trading swing highs). The problem is that, while you can usually tell that a swing high is a swing high (it is within 20 pips of the daily high AND it is exhibiting reversal signs), swing lows are more discretionary in that they haven't gotten the upper barrier of daily high to hold them back or to function as an expected reversal area.

User avatar
tmanbone
rank: 500+ posts
rank: 500+ posts
Posts: 723
Joined: Sun Nov 15, 2009 3:52 am
Reputation: 0
Gender: Male

Postby tmanbone » Wed Jan 05, 2011 8:34 am

Woke up at the right time.

Image
"The simplicity of the markets is it's greatest disguise"

T

cosmoe1
rank: 50+ posts
rank: 50+ posts
Posts: 118
Joined: Wed Jul 29, 2009 3:22 am
Reputation: 0
Gender: Male

Postby cosmoe1 » Wed Jan 05, 2011 4:37 pm



lost my first trade of the day. made up for it on my second.

cos

cosmoe1
rank: 50+ posts
rank: 50+ posts
Posts: 118
Joined: Wed Jul 29, 2009 3:22 am
Reputation: 0
Gender: Male

Postby cosmoe1 » Thu Jan 06, 2011 5:28 am

GeorgeCh wrote:So, I just took the system on a two-day backtest with a twist.

I manually scrolled bar-by-bar on 5M TF through EUR/USD on Dec 30 and Dec 31 (and yes, I realize this is not an exhaustive sample size by any stretch of imagination).

The idea was simple - by scrolling through the charts and making decisions on the basis of the information available to me at the time - rather than retroactively eyeballing through the available data - I could put myself in the shoes of a trader using the system with the element of uncertainty thrown it.

I used the following rules:

1. Wait for the price to make a new high.
2. On 5M TF, wait for a red candle to close.
3. On 5M TF, enter on the next candle as soon as the price reaches the previous candle's low.
4. SL: 10 pips; TP: fixed 5 pips (to eliminate the discretionary decision-making element).

System performance (1 is win, 0 is loss):


1
1
0
1
0
1
1
1
1
1
1
1
1
1

In short, the system produced 12 wins and 2 losses over the two-day period.

Depending on how aggressive you were with risk management, you would make at least 4.05% return over the two-day period - assuming that your SL is equal to 1% of account, and your TP is equal to 0.5% of account.

Should those values be increased to the recommended 2% SL and 1% TP, the return would be 8.22%.

And, just to throw it out there, were you crazy enough to do 5%/2.5%, you'd end up with a whooping 21.37% gain.

Now, obviously, this sample size cannot be representative of the population; more to the point, the winning streak probably skews the results somewhat.

On the other hand, it should be noticed that one of the losers could have been turned a winner by exiting at 5 pip profit, and that at least two winners could have been turned into even greater winners by following the subsequent downtrend.

Before I take this system for a live spin, I'll run more such tests and revert back to you.


Hey George thanks for the "light bulb moment" on risk management. So many people have tried to explain risk management with so many different formulas to me. I do not know why I could not grasp it, but when I read it the way you wrote it, the light bulb came on. Thank you.

Cosmoe1

eaforex
rank: <50 posts
rank: <50 posts
Posts: 1
Joined: Thu Jan 06, 2011 7:25 am
Reputation: 0
Gender: None specified

Postby eaforex » Thu Jan 06, 2011 10:14 am

:-)

Please add www.kreslik.com to your ad blocker white list.
Thank you for your support.


Return to “TheRumpledOne”