Charles, thank you for this thread, its very insightful and i'm hopeful profitable going forward. http://www.stevehopwoodforex.com/phpBB3 ... =16&t=5238#
A question I have had is that what do you do when after a price pull back that takes out some of your trades at breakeven, Do you re-set the trades again effectively in the same place as price moves forward again? Or do you wait and reassess?
Apologies if this has been covered previously and I've missed it.
rgds
Peter
Gravy Trains & Weather Vanes
-
RisklessPips
- Trader
- Posts: 246
- Joined: Mon May 09, 2016 2:24 pm
- Location: Nairobi, Kenya
Gravy Trains & Weather Vanes
Hi Peterpeterke » Tue Aug 22, 2017 8:22 am wrote:Charles, thank you for this thread, its very insightful and i'm hopeful profitable going forward. http://www.stevehopwoodforex.com/phpBB3 ... =16&t=5238#
A question I have had is that what do you do when after a price pull back that takes out some of your trades at breakeven, Do you re-set the trades again effectively in the same place as price moves forward again? Or do you wait and reassess?
Apologies if this has been covered previously and I've missed it.
rgds
Peter
Good question.
If your trade(s) are taken out at BE and there are no other trades open in the same direction you have a choice
a) Look at the EMA's and decide if another trade in the same direction is appropriate (remember seeding is the ONLY time we use historical evidence (an indicator) to suggest future movement !!)
b) If the reversal looks strong stay out and wait for the opposing trade to kick back in on the other side of the channel
In either case with the GTWV EA it doesn't really matter because this little gem will now add new open trades each time the last one goes to BE. This was not the case in the fully manual world and I had to constantly monitor for exactly this situation.
While on this let me also say on a GT move periodically review your support resistance zones to make sure they are still accurate for when (not if) prices reverse and you have to input your contra WV trade.
There is an important element of interaction between you, the EA and price action in GTWV
Charles
Trading is a mind game - good job I have a brain
-
RisklessPips
- Trader
- Posts: 246
- Joined: Mon May 09, 2016 2:24 pm
- Location: Nairobi, Kenya
Gravy Trains & Weather Vanes
Week 1 -Test Results Analysis
I know the week is not yet out but I am posting these results because I am about to change the test bed in order to incorporate changes brought on by GTWV EA.
The two result files are for H4 and W1.
The former is for a two week period, the latter for the 4 days this week.
I will over the weekend (time permitting) discuss the test bed and ideas behind it.
Summary
H4 is showing an expectancy of 6% (i.e for every dollar at risk, it will return 6%) My aim is a 4% return (post 1)
Seen in the traditional way - every time I trade H4 I can expect 38 cents. Given that my daily target is 20USD - I need to trade just over 50 times per day to hit that target - Difficult but doable !
W1 is showing an expectancy of 12.5% ! Seen another way every trade is returning 1.15USD !!
The win rate is v high on GTWV as you would expect with trades taken only after previous ones are at BE - but you will also know a high win rate is not what defines a winning strategy ! A positive expectancy for whatever strategy you use is more important.
Even as I now start to tweak the model to find the sweetspots for the different TF's I am encouraged we are off to a very good start.
Charles
I know the week is not yet out but I am posting these results because I am about to change the test bed in order to incorporate changes brought on by GTWV EA.
The two result files are for H4 and W1.
The former is for a two week period, the latter for the 4 days this week.
I will over the weekend (time permitting) discuss the test bed and ideas behind it.
Summary
H4 is showing an expectancy of 6% (i.e for every dollar at risk, it will return 6%) My aim is a 4% return (post 1)
Seen in the traditional way - every time I trade H4 I can expect 38 cents. Given that my daily target is 20USD - I need to trade just over 50 times per day to hit that target - Difficult but doable !
W1 is showing an expectancy of 12.5% ! Seen another way every trade is returning 1.15USD !!
The win rate is v high on GTWV as you would expect with trades taken only after previous ones are at BE - but you will also know a high win rate is not what defines a winning strategy ! A positive expectancy for whatever strategy you use is more important.
Even as I now start to tweak the model to find the sweetspots for the different TF's I am encouraged we are off to a very good start.
Charles
You do not have the required permissions to view the files attached to this post.
Trading is a mind game - good job I have a brain
-
RisklessPips
- Trader
- Posts: 246
- Joined: Mon May 09, 2016 2:24 pm
- Location: Nairobi, Kenya
Gravy Trains & Weather Vanes
Test Strategy for coming week.
Ahead of next week I want to lay down a test strategy in public - it keeps me focussed and honest
As discussed previously we are trying to establish a sweetspot for each TF we trade with GTWV.
The sweetspot is a combination of BE/SL/TP/Trailing stops AND lots traded that raises the expectancy of the system to a point where you make your daily income target.
In determining expectancy there are 4 factors that come into play
a) Dollar average win
b) Dollar average loss
c) Win rate percentage
d) Loss rate percentage
GTWV has a high historically provable win rate which equals an historically (provable) low loss rate due to ONLY taking trades once the previous has broken even. (80/20)
History, as we know on this thread, points to the future BUT is not guaranteed to be followed. Therefore these win / loss rate elements may change - we do not have control over them.
What we do have control over are the average wins and losses.
This is where my tests will focus effort towards raising the model's expectancy and therefore long term profitability and consistency.
How will we lower average losses
Instead of risking 2% of initial balance (10 USD) on each 0.01 lots traded at 100 pips SL, we will apply a 1.5% risk at 75 pips SL per 0.01 lots. This means the loss rate may rise (and the win rate fall) but the loss average will fall. Taking losses may not be a great feeling - however realists accept (and welcome) them. If the average loss can be reduced we can positively affect expectancy and therefore long term profitability.
How will we raise average wins
1) Set BE and trailing at a level that gives price "room to work" away from entry point.
2) Set TP at a level that "realistically challenges" price to do more in order to hit that level
3) Stack more trades on winners (GTWV EA is indispensable for this, but manual additions also come into play)
Finally, GTWV is continuously in the trenches. That means there is no cherry picking of trades - we just do what price tells us when it tells us. However, the different sessions seem to have different personalities in terms of how price communicates.
Asia is quiet, crossovers and openings are quick and mid sessions are also quiet.
I therefore aim to test different settings for these price communication styles which I will call peak and off-peak.
This is a WIP - more on it as I report on progress starting next Friday.
Charles
EDIT - Light bulb moment
GTWV is in the trenches continuously. Therefore the requirement that the gap between a seed's SL and the S/R zone is filled with mitigating WV trades is NOT a necessary condition for profit - ne c'est pas ?
We have to make average losses smaller !.
The decision to reduce SL from 100 to 75 pips per 0.01 lots above can be extended further to test the limits.
You don't want to have so many losses your loss rate approaches 100%, but a higher loss rate (perhaps up to 60% ?) is not a bad thing as long as your dollar win average is steady or rising and your dollar loss average is falling.
Testing !
Ahead of next week I want to lay down a test strategy in public - it keeps me focussed and honest
As discussed previously we are trying to establish a sweetspot for each TF we trade with GTWV.
The sweetspot is a combination of BE/SL/TP/Trailing stops AND lots traded that raises the expectancy of the system to a point where you make your daily income target.
In determining expectancy there are 4 factors that come into play
a) Dollar average win
b) Dollar average loss
c) Win rate percentage
d) Loss rate percentage
GTWV has a high historically provable win rate which equals an historically (provable) low loss rate due to ONLY taking trades once the previous has broken even. (80/20)
History, as we know on this thread, points to the future BUT is not guaranteed to be followed. Therefore these win / loss rate elements may change - we do not have control over them.
What we do have control over are the average wins and losses.
This is where my tests will focus effort towards raising the model's expectancy and therefore long term profitability and consistency.
How will we lower average losses
Instead of risking 2% of initial balance (10 USD) on each 0.01 lots traded at 100 pips SL, we will apply a 1.5% risk at 75 pips SL per 0.01 lots. This means the loss rate may rise (and the win rate fall) but the loss average will fall. Taking losses may not be a great feeling - however realists accept (and welcome) them. If the average loss can be reduced we can positively affect expectancy and therefore long term profitability.
How will we raise average wins
1) Set BE and trailing at a level that gives price "room to work" away from entry point.
2) Set TP at a level that "realistically challenges" price to do more in order to hit that level
3) Stack more trades on winners (GTWV EA is indispensable for this, but manual additions also come into play)
Finally, GTWV is continuously in the trenches. That means there is no cherry picking of trades - we just do what price tells us when it tells us. However, the different sessions seem to have different personalities in terms of how price communicates.
Asia is quiet, crossovers and openings are quick and mid sessions are also quiet.
I therefore aim to test different settings for these price communication styles which I will call peak and off-peak.
This is a WIP - more on it as I report on progress starting next Friday.
Charles
EDIT - Light bulb moment
GTWV is in the trenches continuously. Therefore the requirement that the gap between a seed's SL and the S/R zone is filled with mitigating WV trades is NOT a necessary condition for profit - ne c'est pas ?
We have to make average losses smaller !.
The decision to reduce SL from 100 to 75 pips per 0.01 lots above can be extended further to test the limits.
You don't want to have so many losses your loss rate approaches 100%, but a higher loss rate (perhaps up to 60% ?) is not a bad thing as long as your dollar win average is steady or rising and your dollar loss average is falling.
Testing !
Trading is a mind game - good job I have a brain
-
RisklessPips
- Trader
- Posts: 246
- Joined: Mon May 09, 2016 2:24 pm
- Location: Nairobi, Kenya
Gravy Trains & Weather Vanes
Emerging Thoughts For GTWV Operation
This post addresses points 1 and 2 in post 1 and has crystallized as a result of testing thus far:
GTWV is rarely out of the market. We are at times :
1) Riding the GT moves
2) On the WV
3) In the channel (Most often here)
This method takes trades when and only when price indicates it is time to do so. This is enforced by the EA but is also a primary and key rule for manual trading.
As a consequence of always being in the market three key points emerge
1) We remove a major decision making burden from ourselves.
Whereas a "dip in and out" strategy places the burden of being right about price entry (level and direction) on the trader and their indicators, GTWV places that burden squarely back on price itself. Price tells you when you are right or wrong. You don't have to make that decision.
We still have to formulate and use a few rules but these are the 'protocol of how we listen to price' - I will expand on that further below.
2) We need not be concerned with S/R zones
S/R zones are important if you are dipping in and out of the market. You must take cognizance of these zones so that you are entering trades at the best level in the direction you have determined.
If you are in the market continuously, as we are, the zones are of no consequence. Price will :
a) Take heed of the zones and therefore tell us not to do anything yet or
b) Breach them and thereby tell us to buy or sell if it "accepts" a previous trade and makes it BE
We just listen to price and do its bidding !
3) Each individual trade becomes insignificant as we become unconsciously competent
Being "always in the market", GTWV enables a mindset to develop where the trader can focus on 'trading in the zone' as each trade is of increasingly small significance because another will be taken at a point in level, direction and time as determined by price without the trader having to think about it. This is the start of unconscious competence.
This mindset is very different for a "dip in and out" trader who has to start the entire decision making process all over again each time a trade closes out for profit (or loss).
Protocol for listening to Price
Price is the sole determinant of trading activity in GTWV however, we need a set of rules about how to listen to price.
This is how I have chosen to do it - there may be other (better) ways - post your ideas.
1) Add a seed at market price (This is the only time I use history in the EMA's to predict future price movement)
2) Using a predefined risk of 1% of initial balance, I place my SL at 50 pips = 5 USD
3) I place a stop order contra trade as follows :
a) If the seed is long, the short stop order contra is placed at SL + 2.5 times the BE pips set in GTWV EA
b) if the seed is short, the long stop order contra is placed at SL - 2.5 times the BE pips set in GTWV EA
Why do I listen this way ?
In order to answer this we need to understand what price will do when you add your seed.
a) Price may "reject" your seed and reverse towards SL where it will be taken out.
In this case your contra stop trade will trigger and as it is 2.5 times BE pips away from SL, 2 new trades should be opened by the EA before the seed gets taken out, and both should be in profit. (I use 2.5 times BE pips to account for spread)
b) Price may "reject" your seed and trigger the contra trade without hitting SL
In this case "the price channel" has been opened up. At some point price will decide to make a break and as GTWV is listening to price on both sides of the channel the EA will act accordingly.
c) Price may "accept" your seed and NOT trigger the contra trade
In this case you are on a GT and GTWV will roll out new trades each time the previous one hits BE. The pending contra stop order will not affect processing unless of course price decides to reverses mid flow
Questions ?
Charles
This post addresses points 1 and 2 in post 1 and has crystallized as a result of testing thus far:
GTWV is rarely out of the market. We are at times :
1) Riding the GT moves
2) On the WV
3) In the channel (Most often here)
This method takes trades when and only when price indicates it is time to do so. This is enforced by the EA but is also a primary and key rule for manual trading.
As a consequence of always being in the market three key points emerge
1) We remove a major decision making burden from ourselves.
Whereas a "dip in and out" strategy places the burden of being right about price entry (level and direction) on the trader and their indicators, GTWV places that burden squarely back on price itself. Price tells you when you are right or wrong. You don't have to make that decision.
We still have to formulate and use a few rules but these are the 'protocol of how we listen to price' - I will expand on that further below.
2) We need not be concerned with S/R zones
S/R zones are important if you are dipping in and out of the market. You must take cognizance of these zones so that you are entering trades at the best level in the direction you have determined.
If you are in the market continuously, as we are, the zones are of no consequence. Price will :
a) Take heed of the zones and therefore tell us not to do anything yet or
b) Breach them and thereby tell us to buy or sell if it "accepts" a previous trade and makes it BE
We just listen to price and do its bidding !
3) Each individual trade becomes insignificant as we become unconsciously competent
Being "always in the market", GTWV enables a mindset to develop where the trader can focus on 'trading in the zone' as each trade is of increasingly small significance because another will be taken at a point in level, direction and time as determined by price without the trader having to think about it. This is the start of unconscious competence.
This mindset is very different for a "dip in and out" trader who has to start the entire decision making process all over again each time a trade closes out for profit (or loss).
Protocol for listening to Price
Price is the sole determinant of trading activity in GTWV however, we need a set of rules about how to listen to price.
This is how I have chosen to do it - there may be other (better) ways - post your ideas.
1) Add a seed at market price (This is the only time I use history in the EMA's to predict future price movement)
2) Using a predefined risk of 1% of initial balance, I place my SL at 50 pips = 5 USD
3) I place a stop order contra trade as follows :
a) If the seed is long, the short stop order contra is placed at SL + 2.5 times the BE pips set in GTWV EA
b) if the seed is short, the long stop order contra is placed at SL - 2.5 times the BE pips set in GTWV EA
Why do I listen this way ?
In order to answer this we need to understand what price will do when you add your seed.
a) Price may "reject" your seed and reverse towards SL where it will be taken out.
In this case your contra stop trade will trigger and as it is 2.5 times BE pips away from SL, 2 new trades should be opened by the EA before the seed gets taken out, and both should be in profit. (I use 2.5 times BE pips to account for spread)
b) Price may "reject" your seed and trigger the contra trade without hitting SL
In this case "the price channel" has been opened up. At some point price will decide to make a break and as GTWV is listening to price on both sides of the channel the EA will act accordingly.
c) Price may "accept" your seed and NOT trigger the contra trade
In this case you are on a GT and GTWV will roll out new trades each time the previous one hits BE. The pending contra stop order will not affect processing unless of course price decides to reverses mid flow
Questions ?
Charles
Trading is a mind game - good job I have a brain
-
peterke
- Trader
- Posts: 56
- Joined: Wed Nov 16, 2011 9:17 am
Gravy Trains & Weather Vanes
Charles, thanks for your continued notes - very insightful.
"The decision to reduce SL from 100 to 75 pips per 0.01 lots above can be extended further to test the limits."
I think the placement of the stop loss is possibly not so much about the total length but more driven by the placement of the first of series of trades in the opposite direction.
If you only have 20 pip distance to recover before the first of series of trade in the opposite direction then the stop loss can be significantly closer than if it was a 40 pip gap.
The impact of this is that the stop loss should be calculated reasonably consitently by a ratio of that distance (based on a % of the ATR?)
In theory that could generate a minimum breakeven distance for the stop loss then any profit could be an extension from that.
It also means that different timeframes and different pairs will have different stop loss settings as price bars are longer at higher time frames and for different pairs.
I'll see what I can come up with using my rudimentary maths skills
rgds
Peter
"The decision to reduce SL from 100 to 75 pips per 0.01 lots above can be extended further to test the limits."
I think the placement of the stop loss is possibly not so much about the total length but more driven by the placement of the first of series of trades in the opposite direction.
If you only have 20 pip distance to recover before the first of series of trade in the opposite direction then the stop loss can be significantly closer than if it was a 40 pip gap.
The impact of this is that the stop loss should be calculated reasonably consitently by a ratio of that distance (based on a % of the ATR?)
In theory that could generate a minimum breakeven distance for the stop loss then any profit could be an extension from that.
It also means that different timeframes and different pairs will have different stop loss settings as price bars are longer at higher time frames and for different pairs.
I'll see what I can come up with using my rudimentary maths skills
rgds
Peter
-
RisklessPips
- Trader
- Posts: 246
- Joined: Mon May 09, 2016 2:24 pm
- Location: Nairobi, Kenya
Gravy Trains & Weather Vanes
Hey Peter.peterke » Sat Aug 26, 2017 9:54 am wrote: If you only have 20 pip distance to recover before the first of series of trade in the opposite direction then the stop loss can be significantly closer than if it was a 40 pip gap.
The impact of this is that the stop loss should be calculated reasonably consitently by a ratio of that distance (based on a % of the ATR?)
It also means that different timeframes and different pairs will have different stop loss settings as price bars are longer at higher time frames and for different pairs.
I'll see what I can come up with using my rudimentary maths skills
rgds
Peter
I use a fixed SL of 1% of initial balance to fix my loss at a known number across all pair tf combinations. My average loss becomes easier to manage.
If it is mathematically possible to show use of an atr based method will lower average losses I'm in
Charles
Edit I have made a change to the EA. If you choose to use the SL Atr it will apply the lower of that ATR calculation and any fixed number of pips chosen to the trade.
Pairs with low volatility will therefore have a tighter SL than the fixed (50 pips) once again reducing the average loss.
I will post it after testing at market open.
Trading is a mind game - good job I have a brain
-
RisklessPips
- Trader
- Posts: 246
- Joined: Mon May 09, 2016 2:24 pm
- Location: Nairobi, Kenya
Gravy Trains & Weather Vanes
GTWV EA Updated in the post where it first appeared - Changes are dated 28/8/17
If you set SLAtr to true the EA now adds a SL to trades that is the tighter to entry price between
a) a missing stop loss fixed number of pips and
b) a calculation based on ATR.
The settings pdf is also updated with relevant comments of SL ATR settings.
Previously I had not updated the available code for a missing conditional to ensure proper addition of new trades by the EA. The updated code has this modification.
If interested the changes are between lines 1136 and 1156
Charles
If you set SLAtr to true the EA now adds a SL to trades that is the tighter to entry price between
a) a missing stop loss fixed number of pips and
b) a calculation based on ATR.
The settings pdf is also updated with relevant comments of SL ATR settings.
Previously I had not updated the available code for a missing conditional to ensure proper addition of new trades by the EA. The updated code has this modification.
If interested the changes are between lines 1136 and 1156
Charles
Trading is a mind game - good job I have a brain
-
RisklessPips
- Trader
- Posts: 246
- Joined: Mon May 09, 2016 2:24 pm
- Location: Nairobi, Kenya
Gravy Trains & Weather Vanes
NFP & GTWV
I was waiting to see how GTWV handled this event before posting this weeks baseline results which will appear shortly with my analysis
Charles
I was waiting to see how GTWV handled this event before posting this weeks baseline results which will appear shortly with my analysis
Charles
You do not have the required permissions to view the files attached to this post.
Trading is a mind game - good job I have a brain
-
RisklessPips
- Trader
- Posts: 246
- Joined: Mon May 09, 2016 2:24 pm
- Location: Nairobi, Kenya
Gravy Trains & Weather Vanes
Ist Sept 2017 Weekly results and brief analysis
H4 - Winner
This week H4 has increased starting balance by 71 USD (Target is 20 USD per day i.e 100 USD per week!)
The expectancy (from a statistically insignificant dataset) is positive returning over 7% per dollar traded or over 35 cents per trade.
Next week a rise in average wins of 60 cents, all else being constant
should raise expectancy above 15%. (My target as set out in post 1 is 4% of initial balance per day.)
D1 - Good Results - encouraging :youknow:
A small negative expectancy (again from a statistically insignificant dataset) was recorded losing 17 cents per trade or 3.5% per dollar traded.
A rise of 80 cents on the average win next week may propel this TF into positive expectancy country. Stay tuned.
W1 - More work required
Large negative expectancy of 23% loss per dollar traded (or over 1 dollar lost per trade)
Work to be done on finding the sweetspot next week. This will entail changing the SL/BE/TP and trails.
Other Observations
TF's H1 and below are v challenging. They make profit but give it back all too easily. For this reason I have decided to concentrate on where highest probability of success arises and that is in H4 and above.
In this regard I am also now testing a custom TF (H6) which smooths out fluctuations as does H4 and allows the logic and EA behind GTWV to operate in a manageable price environment.
It has been brought to my attention in PM's that some people thought the PDF with settings I posted earlier was to be applied to GTWV EA. This is not so ! That file merely described what the various settings do. If you choose to use GTWV EA you must play around with the setfile as found on the mq4 to find the sweetspot for the TF you are trading.
Having said that I attach the setfile for the winning TF (H4) this week below. It is as I will use it next week - so no guarantees - test, test, test.
Charles
H4 - Winner
This week H4 has increased starting balance by 71 USD (Target is 20 USD per day i.e 100 USD per week!)
The expectancy (from a statistically insignificant dataset) is positive returning over 7% per dollar traded or over 35 cents per trade.
Next week a rise in average wins of 60 cents, all else being constant
should raise expectancy above 15%. (My target as set out in post 1 is 4% of initial balance per day.)D1 - Good Results - encouraging :youknow:
A small negative expectancy (again from a statistically insignificant dataset) was recorded losing 17 cents per trade or 3.5% per dollar traded.
A rise of 80 cents on the average win next week may propel this TF into positive expectancy country. Stay tuned.
W1 - More work required
Large negative expectancy of 23% loss per dollar traded (or over 1 dollar lost per trade)
Work to be done on finding the sweetspot next week. This will entail changing the SL/BE/TP and trails.
Other Observations
TF's H1 and below are v challenging. They make profit but give it back all too easily. For this reason I have decided to concentrate on where highest probability of success arises and that is in H4 and above.
In this regard I am also now testing a custom TF (H6) which smooths out fluctuations as does H4 and allows the logic and EA behind GTWV to operate in a manageable price environment.
It has been brought to my attention in PM's that some people thought the PDF with settings I posted earlier was to be applied to GTWV EA. This is not so ! That file merely described what the various settings do. If you choose to use GTWV EA you must play around with the setfile as found on the mq4 to find the sweetspot for the TF you are trading.
Having said that I attach the setfile for the winning TF (H4) this week below. It is as I will use it next week - so no guarantees - test, test, test.
Charles
You do not have the required permissions to view the files attached to this post.
Trading is a mind game - good job I have a brain