Discussion of article "Working with currency baskets in the Forex market"

 

New article Working with currency baskets in the Forex market has been published:

The article describes how currency pairs can be divided into groups (baskets), as well as how to obtain data about their status (for example, overbought and oversold) using certain indicators and how to apply this data in trading.

How should we decide on performing a trading operation with a currency basket? It would be quite difficult to do so based solely on currency pairs behavior due to the large size of the baskets. We need additional technical data on basket currency status parameters. Let's focus our attention on the RSI indicator graph for the three currency pairs from the GBP basket, namely GBPUSD, GBPJPY and GBPCHF. D1 graph is displayed in the image below:


As you can see, these three graphs have much in common: many extreme values are located similarly, while up and down intervals also have much in common. This similarity can be explained by the same basket currency — the British pound. Its strengthening and weakening lead to such up and down intervals.

Author: Andrei Novichkov

 

Excellent article Andrei. This theory I practice since a very long time and is very profitable. Jet I do not trade baskets but the strongest currency against the weakest. This will reduce our number of trades because we do not need open 7 trades to buy or sell one currency basket, instead we have only one trade to execute which is paired of the strongest and weakest base and quote currency.

If you dont mind to link, it is here

 
Bernhard Schweigert :

Excellent article Andrei. This theory I practice since a very long time and is very profitable. Jet I do not trade baskets but the strongest currency against the weakest. This will reduce our number of trades because we do not need open 7 trades to buy or sell one currency basket, instead we have only one trade to execute which is paired of the strongest and weakest base and quote currency.

If you dont mind to link, it is here

Hello, Bernhard

Thank you for your review), I agree with your strategy on your link, very interesting.

I'm writing an article about patterns, there will be about it - "strongest currency against the weakest" )


 

Hi Andrei,

Excellent article. Some people have tried Basket trading based on index of hi/low bid price/2, some have based on MTF Moving Averages. You have shown RSI and Stochastics use. 

I really find them useful only for momentum based short-term frame trading.

Why and when momentum accelerates? If you can somehow tie up basket trading with patterns, then it will be useful.

Can you please make testing indicators/EA for Breakout of basket hi and low of certain number of bars (zig zags) to identify momentum pattern?

I think basket trading only works when market makers need to guide the market on the eve of news events or when the currency gets tired and reverses direction.

So it is important to have direction bias first and then keep an eye for momentum to time the trade. 

 
Peejay Lal :

Hi Andrei,

Excellent article. Some people have tried Basket trading based on index of hi/low bid price/2, some have based on MTF Moving Averages. You have shown RSI and Stochastics use. 

I really find them useful only for momentum based short-term frame trading.

Why and when momentum accelerates? If you can somehow tie up basket trading with patterns, then it will be useful.

Can you please make testing indicators/EA for Breakout of basket hi and low of certain number of bars (zig zags) to identify momentum pattern?

I think basket trading only works when market makers need to guide the market on the eve of news events or when the currency gets tired and reverses direction.

So it is important to have direction bias first and then keep an eye for momentum to time the trade. 

I think that baskets work best for trend reversal. When one currency pair in too enhanced, and the second is too attenuated.
 

The indicators doesn't work in the Metatrader Strategy Tester ( data displayed are different  of the real data).

Is it a limit of MT4 or should be possible to code them working in the strategy tester ?

 
AndreaTrade1:

The indicators doesn't work in the Metatrader Strategy Tester ( data displayed are different  of the real data).

Is it a limit of MT4 or should be possible to code them working in the strategy tester ?

Multi pair indicators usually never work in strategy tester because it needs too many other chart feeds.

Strategy tester does not provide all needed data.

 
Bernhard Schweigert :

Multi pair indicators usually never work in strategy tester because it needs too many other chart feeds.

Strategy tester does not provide all needed data.

....And we need a very good historical data. For all chart feeds.
 

Don't know if this is too late, being like 5 years ago lol, but your indicator has a major flaw. It's assuming that ALL of the data that you've got for all of your currency pairs is exact. Meaning the time in AUDUSD 10 bars ago is the exact same time as the time from AUDNZD 10 bars ago. Unless you go to each one and make sure all of the data is downloaded, that would cause issues when loading the indicator. Also, there can be problems/gaps in history, which would also cause (compounding) issues. If just one is off, then it will be off the rest of the candles going back, and will get even more off as others aren't exact.

You need to use iBarShift with the current time of the bar that you're working on to get the actual shift of the bar in different currency pairs. basically if "i" is your working bar, then..

 barShift = iBarShift("AUDNZD",_Period,Time[i]);

Then check to see if it's returned an error, and if not..

otherOpen = iOpen("AUDNZD",_Period,barShift);

 That will guarantee you're getting the correct value instead of blindly walking backwards and building something that probably isn't totally accurate.

 
KEatonFX :

Don't know if this is too late, being like 5 years ago lol, but your indicator has a major flaw. It's assuming that ALL of the data that you've got for all of your currency pairs is exact. Meaning the time in AUDUSD 10 bars ago is the exact same time as the time from AUDNZD 10 bars ago. Unless you go to each one and make sure all of the data is downloaded, that would cause issues when loading the indicator. Also, there can be problems/gaps in history, which would also cause (compounding) issues. If just one is off, then it will be off the rest of the candles going back, and will get even more off as others aren't exact...

Of course, I completely agree with you) But this article did not imply the development of indicators. This is just a theoretical basis. The indicators accompanying this article are for educational purposes and are for illustration purposes only. They cannot be used for real work.

Indicators that can be used on real accounts are much more complicated. This is a completely different task.

 
KEatonFX:

Don't know if this is too late, being like 5 years ago lol, but your indicator has a major flaw. It's assuming that ALL of the data that you've got for all of your currency pairs is exact. Meaning the time in AUDUSD 10 bars ago is the exact same time as the time from AUDNZD 10 bars ago. Unless you go to each one and make sure all of the data is downloaded, that would cause issues when loading the indicator. Also, there can be problems/gaps in history, which would also cause (compounding) issues. If just one is off, then it will be off the rest of the candles going back, and will get even more off as others aren't exact.

You need to use iBarShift with the current time of the bar that you're working on to get the actual shift of the bar in different currency pairs. basically if "i" is your working bar, then..

 barShift = iBarShift("AUDNZD",_Period,Time[i]);

Then check to see if it's returned an error, and if not..

otherOpen = iOpen("AUDNZD",_Period,barShift);

 That will guarantee you're getting the correct value instead of blindly walking backwards and building something that probably isn't totally accurate.

error in data feed must always be checked! Do not trust broker data. Its only something like 95-99%.

Reason: