daily range Strategy | Page 38 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

Strategy: Daily Range Spike and Pull Back 161 replies

Daily Range Strategy 19 replies

Range/Last = range in %... help find the bug... 2 replies

How to get range of h/l on previous day's range? 1 reply

Another Daily Range Strategy GBP/USD 23 replies

Attachments: daily range Strategy
Exit Attachments

daily range Strategy

  • #741
  • Jan 16, 2007 9:13am Jan 16, 2007 9:13am
I suppose it's the overall your after anyway - Maybe highlight days you should and shouldnt trade, then you can start to see the % of wins increase.

Did you have a look at my Simple Strategy?

I'm just working on Range from 00.00GMT, Stop loss and Take Profit to see which one works best and then i'll start testing it from tonight
  • #742
  • Jan 16, 2007 11:46am Jan 16, 2007 11:46am
Quoting Rama
Hi Kwackers,
I'm following this strategy in a real account since january 2 2007 but with diferent parameters. And since last friday i lost 27 pips per day (friday, monday and today).
It gives me net 53 pips loss from the begining (01/02/2007).

The reason of this consecutive loses is that the strategy doesn't work well in days without important news over de USD.

The trend's range is tighty, and it works better with volatility. So here is something very important: don't trade if market doesn't wait important news.

Regards,
Here's something to try. Just place your straddles at the end of Monday and Wednesday and Thursday. (That way we would trad only Tues, Thurs, and Frid). I have some stats that show those are the historically most volatile days of the week. Some of my backtesting shows that it doesn't always work that way, but it might be a method to increase probability for success.

Hey Kwackers! I'm down over 1500 bucks today. Both lots on the buy went bad and now I've triggered the short and gone the wrong direction there too! After two months I'm now about break even. A month ago I was well ahead in the game. Sigh.
Don Life is expensive, but includes a free trip around the sun.
  • #743
  • Jan 16, 2007 11:52am Jan 16, 2007 11:52am
Quoting Kwackers
I suppose it's the overall your after anyway - Maybe highlight days you should and shouldnt trade, then you can start to see the % of wins increase.

Did you have a look at my Simple Strategy?

I'm just working on Range from 00.00GMT, Stop loss and Take Profit to see which one works best and then i'll start testing it from tonight
Kwackers,
I looked at your strategy. Basically it's the same i'm using. The diference are the parameters used.

This are mine:
- Each day i put two STOP ORDERS based on 00:00 GMT OPEN trade of GBPUSD pair.

- Enter 58 pips above (buy) or below (sell) OPEN GBPUSD.
- Limit 15 pips from enter.
- Stop 27 pips from enter.

I backtested this parameters from 29/05/2006 to 29/12/2006. And this are de results:
1% double loss (sell and buy).
2% double profit (sell and buy).
62% profit (sell or buy).
15% loss (sell or buy).
19% didn't trade.

In 155 trading days the systems returns 834 pips of profit. It represents 27 pips per week.

It is not so profitable in pips, buy with a good combination of money management rules, it gives 4756% in the period analized. Starting with an hipotetical balance of USD 250 on may 29, at december 29 you would get 7317 USD.

In addition, the sistem is very simple, and you don't need to be in front the charts to much time.

Thats are the reasons why i decided to start using this method in my real account: simple and no time demand.

The backtest was made without considering fundamental information. But analizing the worst weeks i realized that there isn't a real trend. What i mean it was very tighty. The same that happened today and yesterday with de GBPUSD pair.

This kind of days it is better don't trade.

I think tomorrow, with USD PPI report, will be a nice day for this method.

If you want, i can explain the reasons of the parameters used.
I also backtested this parameters:
ENTER LIMIT STOP
58 22 25
58 27 27
56 60 38
56 40 30
56 50 100
56 110 100

I hope this help all of us.

Regards,
  • #744
  • Jan 16, 2007 11:53am Jan 16, 2007 11:53am
Quoting Kwackers
I suppose it's the overall your after anyway - Maybe highlight days you should and shouldnt trade, then you can start to see the % of wins increase.

Did you have a look at my Simple Strategy?

I'm just working on Range from 00.00GMT, Stop loss and Take Profit to see which one works best and then i'll start testing it from tonight
Kwackers,
I looked at your strategy. Basically it's the same i'm using. The diference are the parameters used.

This are mine:
- Each day i put two STOP ORDERS based on 00:00 GMT OPEN trade of GBPUSD pair.

- Enter 58 pips above (buy) or below (sell) OPEN GBPUSD.
- Limit 15 pips from enter.
- Stop 27 pips from enter.

I backtested this parameters from 29/05/2006 to 29/12/2006. And this are de results:
1% double loss (sell and buy).
2% double profit (sell and buy).
62% profit (sell or buy).
15% loss (sell or buy).
19% didn't trade.

In 155 trading days the systems returns 834 pips of profit. It represents 27 pips per week.

It is not so profitable in pips, buy with a good combination of money management rules, it gives 4756% in the period analized. Starting with an hipotetical balance of USD 250 on may 29, at december 29 you would get 7317 USD.

In addition, the sistem is very simple, and you don't need to be in front the charts to much time.

Thats are the reasons why i decided to start using this method in my real account: simple and no time demand.

The backtest was made without considering fundamental information. But analizing the worst weeks i realized that there isn't a real trend. What i mean it was very tighty. The same that happened today and yesterday with de GBPUSD pair.

This kind of days it is better don't trade.

I think tomorrow, with USD PPI report, will be a nice day for this method.

If you want, i can explain the reasons of the parameters used.
I also backtested this parameters:
ENTER LIMIT STOP
58 22 25
58 27 27
56 60 38
56 40 30
56 50 100
56 110 100

I hope this help all of us.

I beg you pardon by my english. It's not may natural language.

Regards,
  • #745
  • Jan 16, 2007 12:06pm Jan 16, 2007 12:06pm
Quoting traderone
Here's something to try. Just place your straddles at the end of Monday and Wednesday and Thursday. (That way we would trad only Tues, Thurs, and Frid). I have some stats that show those are the historically most volatile days of the week. Some of my backtesting shows that it doesn't always work that way, but it might be a method to increase probability for success.

Hey Kwackers! I'm down over 1500 bucks today. Both lots on the buy went bad and now I've triggered the short and gone the wrong direction there too! After two months I'm now about break even. A month ago I was well ahead in the game. Sigh.
Hi Traderone,
I agree with you that tues, thurs and frid are the best days for this strategy.

Today i lose 27 pips by 3rd. consecutive day.

I can see that if you place the orders at 9:00 GMT, very near London market open, instead do that at 00:00 gmt you will avoid enter in a tighty market o an extreme volatile market.

I'm using this strategy on my real account, and this give me a profit on 05/01/07, and didn't trade on 09/01/07.

If you watch a 5' minutes chart of that days, you'll see a loss in both if orders were scheduled at 00:00 gmt and before 9:00 gmt.

Regards.
  • #746
  • Jan 16, 2007 2:34pm Jan 16, 2007 2:34pm
RAMA,

I think you might not have noticed that Kwackers' entries are for reversal trades from what you are doing. Here is a snip from his doc....

Example

1) 1.8800

2) Buy 1 lot = 1.8750 Sell 1 Lot = 1.8850

3) Stop Loss :

Buy = 1.8720 Sell = 1.8880


Buy Take Profit 1 = 1.8765 Sell Take Profit 1 = 1.8835

If you will notice, if the price goes DOWN 50 pips then he BUYS! and vice versa. I'll have to look at this very closely!
Don Life is expensive, but includes a free trip around the sun.
  • #747
  • Jan 16, 2007 2:44pm Jan 16, 2007 2:44pm
Why do you think his nickname is Kwackers?


Quoting Rama
Kwackers,
I looked at your strategy. Basically it's the same i'm using. The diference are the parameters used.

This are mine:
- Each day i put two STOP ORDERS based on 00:00 GMT OPEN trade of GBPUSD pair.

- Enter 58 pips above (buy) or below (sell) OPEN GBPUSD.
- Limit 15 pips from enter.
- Stop 27 pips from enter.

I backtested this parameters from 29/05/2006 to 29/12/2006. And this are de results:
1% double loss (sell and buy).
2% double profit (sell and buy).
62% profit (sell or buy).
15% loss (sell or buy).
19% didn't trade.

In 155 trading days the systems returns 834 pips of profit. It represents 27 pips per week.

It is not so profitable in pips, buy with a good combination of money management rules, it gives 4756% in the period analized. Starting with an hipotetical balance of USD 250 on may 29, at december 29 you would get 7317 USD.

In addition, the sistem is very simple, and you don't need to be in front the charts to much time.

Thats are the reasons why i decided to start using this method in my real account: simple and no time demand.

The backtest was made without considering fundamental information. But analizing the worst weeks i realized that there isn't a real trend. What i mean it was very tighty. The same that happened today and yesterday with de GBPUSD pair.

This kind of days it is better don't trade.

I think tomorrow, with USD PPI report, will be a nice day for this method.

If you want, i can explain the reasons of the parameters used.
I also backtested this parameters:
ENTER LIMIT STOP
58 22 25
58 27 27
56 60 38
56 40 30
56 50 100
56 110 100

I hope this help all of us.

I beg you pardon by my english. It's not may natural language.

Regards,
Trade what you see - Not what you think
  • #748
  • Jan 16, 2007 3:04pm Jan 16, 2007 3:04pm
Quoting traderone
RAMA,

I think you might not have noticed that Kwackers' entries are for reversal trades from what you are doing. Here is a snip from his doc....

Example

1)1.8800

2)Buy 1 lot = 1.8750 Sell 1 Lot = 1.8850

3)Stop Loss :

Buy = 1.8720 Sell = 1.8880


Buy Take Profit 1 = 1.8765 Sell Take Profit 1 = 1.8835

If you will notice, if the price goes DOWN 50 pips then he BUYS! and vice versa. I'll have to look at this very closely!
Yes i notice it's a reversal system.
I answered at its own thread.
  • #749
  • Jan 16, 2007 3:09pm Jan 16, 2007 3:09pm
Quoting summitfx
Why do you think his nickname is Kwackers?
Isn't it?

Perhaps i confused a bank name, or a software with a nickname because english is not my natural languge?

I beg you pardon, whatever your nickname is.
  • #750
  • Jan 16, 2007 3:15pm Jan 16, 2007 3:15pm
Me - well errr....kwackers i guess?

Anyway if anybody would care to comment, please do so on my thread which is 'Another Range strategy GBP/USD', trying not to HIJACK micardos thread, as like i said, he probably has a better theory than mine.
  • #751
  • Edited 3:34pm Jan 16, 2007 3:19pm | Edited 3:34pm
Just joking around rama obama sama- I think you missed the jest of the joke-I mean you cats were talking about averaging 27 pips per week and I thought that was a bit far-out so I said why do you think they call him kwackers which to me is similiar to krackers which is sorta nutty like woody allen if you catch my drift. Oh well, I'm feeling groovy so I'm going to split this joint.
cheers~
Trade what you see - Not what you think
  • #752
  • Jan 16, 2007 11:32pm Jan 16, 2007 11:32pm
Kwackers, 27 pips p/wk is a marginal profit. I hope that figure is net otherwise spread will blow it out of the water. You will need to achieve a win/loss of around 80% for a trade with a 15TP and 30SL to be profitable. Hope you are running a TSL on the 80TP cause it is a lot of profit to give away if price reverts after entry. Best of luck.

Traderone, We actually need directional/trending movement rather than volatility. The whipsaws and generally non-directional movement challenge the profitability of this system.

My latest thoughts are: the most profitable days are when the market shows strong direction. The least profitable days are when the market is volatile and whipsaws around taking out both entries and SL's. On a non-trending day I have a good chance of the 14TP being triggered, however it is very unlikely that the 50 or 80TP will meet its target so my thought about refining the strategy is to apply the larger TP's (eg 50TP & 80TP) on trending days and do not apply them on non-trending days.

This bodes the question what is a good indicator to determine what is a trending v's non-trending day. Then would the indicator be sophisticated enough to switch on the TP's depending upon the strength of the trend. For instance: non-trending=14TP, slight trend=14 & 30TP, trending=14, 30 & 50TP, strong trend=14, 30, 50 & 80TP.

I haven't explored this further at this stage. If anyone can suggest an indicator it would be appreciated.

Good to hear from you Micardo.
  • #753
  • Edited 10:37pm Jan 24, 2007 9:04am | Edited 10:37pm
Hi Mikhail.

I've been thinking of some ways of developing the DRS EA further. I haven't written MQ4 before, so any help would be appreciated, but if not, would you mind posting the MQ4 version of your DRS EA so I can continue this project?

My current thoughts are to:

Stage 1
Optimise it for time and for price targets. A larger first TP, for instance, 20pips may work better than the 14TP because of the difference in the profit/loss ratio.

Stage2
Add in Money Mgt such as in Phoenix

Stage3
Add code to switch the Entries and TP's on or off depending upon the likelihood of them being sucessfully achieved. The method I have thought of is to look back to the period between the start and end time settings of the previous two days, take the difference between the open and the high of that period over the last two days and average it. If the TP is >= the look back then allow the position to open else exit. Ditto for open and low. These two events should be independent of each other. So for instance, if we are in a strong uptrend all the long positions will be allowed and the short positions will not be allowed. I haven't backtested this theory however I think it should increase the number of positions in a trending market and reduce the number of positions in a nontrending market which should hopefully increase the profitability of the system.

Stage4
Allow the money mgt % of free margin (allowed per trade) to be incrementally increased (by a % of free margin) upon consecutive wins up to a ceiling. Each trade should be treated individually. For example, if the Long 30TP consecutively wins the % of free margin per trade gets incremented, however if the Long 50TP doesn't win it doesn't get incremented. This should allow larger positions to be taken when the market is trending or at least when you are consecutively winning for whatever reason. This should increase the profitability of the system. I think this may be in Phoenix as well.

I think after completing these stages of development the EA would be really quite powerful.

Thankyou in advance.
  • #754
  • Jan 30, 2007 8:08pm Jan 30, 2007 8:08pm
At the risk of this thread becoming a monologue. I've been backtesting and optimising the time and price inputs. The methodology I've used is to optimise over 2006 data and test against 2005 data. My view on multiple TP’s is that each plan (as in Entry, TP & SL) should hold it’s own weight individually and can be combined to smooth the equity curve and profit multiple times from one event. So I tested each plan independently.

The good news is that after optimisation I produced some fantastic results; TP1 50% ROC, TP2 100% ROC. The bad news is when I tested it against 2005 data it produced a negative result so the strategy for 2006 did not hold true for 2005. This got me thinking - does that mean the methodology of optimising over 1 year doesn't produce a robust result? Does that mean I should ditch that method and try perhaps a rolling optimisation eg: optimise for dec apply to jan, optimise for jan apply to feb etc. Maybe that will help.

Then I thought maybe optimising for time and price should be the last stage of development rather than the first. The stages of development mentioned in my previous post will all effect the way this strategy works and the profitability of different inputs. So I think the stages of development need to be completed first.

I've noticed the EA is producing OCO orders when the strategy is not meant to enter an OCO order.

I've also noticed the trailing stop the EA is using is a trailing profit stop. That is, price needs to move into profit by the quantum of the trailing stop for it to engage. When I conducted my manual backtesting I found that a trailing stop loss which turns into a trailing profit stop was significantly more profitable than a trailing profit stop by itself. That is, a trailing stop which engages immediately when the bid-TSL>SL. Example code below:

If TrailingStop > 0 then // if trailing stops are used
{
If OrderValue(cnt,VAL_STOPLOSS)<(Bid-Point*TrailingStop) then
{
ModifyOrder(OrderValue(cnt,VAL_TICKET),OrderValue(cnt,VAL_OPENPRICE),
Bid-Point*TrailingStop,OrderValue(cnt,VAL_TAKEPROFIT),Red);
Exit
};
};

After making these changes and completing these stages of development I think we would have a powerfully profitable EA.

Mikhail, if your out there, help!!!
  • #755
  • Jan 30, 2007 9:17pm Jan 30, 2007 9:17pm
Hello,

Im sorry I cannot post my code in public because as you know people sell codes like this in ebay. also some people already bought the code from me which would not be fair to them if I post it in the public forum.

As for the trailing stop part, you can change that in the EA I posted here. go to the exten vars change "trailingstoptype": 1=After Profit Trailing Stop; 2=Trailing Stop even before profit; 3 Trailing Stop 50% of the profit;

Hope you understand and hope that helps,

Regards,
Mikhail
  • #756
  • Jan 30, 2007 9:55pm Jan 30, 2007 9:55pm
hello again Trav ,

It is very good to see that you are trying to develop the system for it to become robust but I cant help but comment on a few things if you don't mind.

The EA I posted here is not a trading system it is just an expert adviser that makes trading the DRS just a little bit easier by automating the task for you. But if you truly want to make a robust trading system out of this strategy, that would be profitable in the long run I suggest you stop looking at fixed levels of entry and exit, It just will not work. The market is dynamic so your system must also be able to change as the market changes.

To achieve this in the DRS you have to look at ATR or percentages. An example would be the "SBS EA". its logic is similar to this system but the levels of entry is not fixed. It bases its entry level by calculatuing 70% of yesterday daily range, then thats where it puts its entry levels. and closes all orders at end of the day no tp. Very simple but proves to be profitable.

Also if you are able to make a system profitable both in 2005 and 2006, then you would have found your grail, its because 2005 is mostly ranging while 2006 is mostly trending.

Just a few suggestions but I encourage you to continue to look at the market and it will reveal itself to you.

Regards,
Mikhail

Quoting Trav
At the risk of this thread becoming a monologue. I've been backtesting and optimising the time and price inputs. The methodology I've used is to optimise over 2006 data and test against 2005 data. My view on multiple TP’s is that each plan (as in Entry, TP & SL) should hold it’s own weight individually and can be combined to smooth the equity curve and profit multiple times from one event. So I tested each plan independently.

The good news is that after optimisation I produced some fantastic results; TP1 50% ROC, TP2 100% ROC. The bad news is when I tested it against 2005 data it produced a negative result so the strategy for 2006 did not hold true for 2005. This got me thinking - does that mean the methodology of optimising over 1 year doesn't produce a robust result? Does that mean I should ditch that method and try perhaps a rolling optimisation eg: optimise for dec apply to jan, optimise for jan apply to feb etc. Maybe that will help.

Then I thought maybe optimising for time and price should be the last stage of development rather than the first. The stages of development mentioned in my previous post will all effect the way this strategy works and the profitability of different inputs. So I think the stages of development need to be completed first.

I've noticed the EA is producing OCO orders when the strategy is not meant to enter an OCO order.

I've also noticed the trailing stop the EA is using is a trailing profit stop. That is, price needs to move into profit by the quantum of the trailing stop for it to engage. When I conducted my manual backtesting I found that a trailing stop loss which turns into a trailing profit stop was significantly more profitable than a trailing profit stop by itself. That is, a trailing stop which engages immediately when the bid-TSL>SL. Example code below:

If TrailingStop > 0 then // if trailing stops are used
{
If OrderValue(cnt,VAL_STOPLOSS)<(Bid-Point*TrailingStop) then
{
ModifyOrder(OrderValue(cnt,VAL_TICKET),OrderValue(cnt,VAL_OPENPRICE),
Bid-Point*TrailingStop,OrderValue(cnt,VAL_TAKEPROFIT),Red);
Exit
};
};

After making these changes and completing these stages of development I think we would have a powerfully profitable EA.

Mikhail, if your out there, help!!!
  • #757
  • Jan 31, 2007 1:51am Jan 31, 2007 1:51am
try this: note the tokyo closing of the 4 majors usdchf eruusd gbpusd and usdjpy and use this close as starting point for your trading. use an excel spreadsheet where u have marketprice - opening of your ccy. if the usd gets positive against all 4 cceis = long usd market and if the usd gets neg vs the 4 ccies = bear usd market
  • #758
  • Jan 31, 2007 2:57am Jan 31, 2007 2:57am
Thanks for your reply. I appreciate and respect that as a programmer you have the ability and right to comand a fee for your Intellectual Property.

As a trader I seek to profit from trading the strategy and hold the belief that the input of many people will create a better strategy than what one person can create. Hence, I would be happy to post the code (like Hendrick has done with Phoenix) and develop the strategy in a forum. Besides, my programming ability is currently very limited so I would need input at some stage.

Thanks for the definition of the trailing stop inputs.

My attempt to manage risk and capitalise on the dynamics of the market is to turn on/off the trades with higher TP's if they are/aren't likely to be hit. I really like your idea of percentages and will definately research SBS EA.

I'm hoping that stage three and four mentioned previously will have the effect of decreasing my positions and size in a non-profitable market and increasing it when it is profitable. I'm not sure whether this is my grail, but I'm quietly confident that this strategy will be profitable. As you said, 2005 is a good testing sample for a directional strategy.

My plan will be to see if I can construct an EA from similar ones such as Amazing EA, Straddle EA, Phoenix and SBS EA. I would appreciate any comments or input you are willing to give.
  • #759
  • Jan 31, 2007 3:14am Jan 31, 2007 3:14am
Hi to everyone still left on this very interesting thread. I spent the last 5 hours sorting through this very long thread. I saw things that worked in the beginning and not later on. I think mikhail hit the nail on the head. The market is not constant so your system cannot be. One thing I learned about this market, is that support and resistance levels are constantly changing but do repeat themselves alot also. Mikhail, you stated "It bases its entry level by calculatuing 70% of yesterday daily range, then thats where it puts its entry levels. and closes all orders at end of the day no tp." If you could please explain further on how you calculate the previous daily range and making sure you close at the end of the day with a profit. I have seen many times that a good profit turns into a bad lose all within one day? I would really appreciate the info. that you can give. I want to try the Kamanda trading style, but like you said, the market is not a constant market. Things change thats why you could spend the rest of your life looking at threads and systems that work now but not the following year. I think you have something with that 70% range. PLEASE HELP. LOL. Thank you and take care.
  • #760
  • Jan 31, 2007 8:34am Jan 31, 2007 8:34am
Hello All,

If you want further information on the strategy I am referring to PM me your email, I will gladly email to you a very good trading book written by Larry Williams. They based the system I was referring to form one of his strategy in that book that he merely discussed it to illustrate his point, and now we know his point is very profitable

Trav I would love for you to read this book I'm sure you would be able to create great systems from the knowledge you would acquire here and I would gladly help you and everyone else in developing a robust system that could conquer the market.

Regards,
Mikhail
1 trader viewing now