Forex strategies - page 6

 
khorosh:

Sart If the day is predominantly flat, like today on the Eu, then you can trade in the direction of the inside of the channel without waiting for the width of the daily channel to reach its average width. If you look at today's eu (M5) you may see a lot of entries during the whole day. But not with the help of limit orders, but with the help of market

orders. If the price immediately bounces from the border inside the channel we perform an operation. If the price breaks through the channel, we wait for it to reverse inwards and open a position again.

I wonder how you can tell if the day is going to be flat before it ends... You can tell the week or even the month to be flat... That's the true grail:)

 
khorosh:

Sart If the day is predominantly flat, like today on the Eu, then you can trade in the direction of the inside of the channel without waiting for the width of the daily channel to reach its average width. If you look at today's eu (M5) you may see a lot of entries during the whole day. But not with the help of limit orders, but with the help of market orders.

orders. If price immediately bounces from the border into the channel, we execute the trade. If price breaks through the channel, we wait for it to turn inwards and open a position again.

One can talk differently about the daily average range...But look again, today the Canadian reached its 60-day average range exactly at the last "minute", so to speak - exactly 108 p., and then it froze in anticipation ....

In other words, the average daily range may serve as a very good reference point when considering any issue, it may be referred to.

 
Lord_Shadows:
We should alsoopen a position:

Sart If the day is predominantly flat, like today on the Eu, then you can trade in the direction of the inside of the channel without waiting for the width of the daily channel to reach its average width. If you look at today's eu (M5) you may see a lot of entries during the whole day. But not with the help of Limit orders, but with the help of market ones.

orders. If price immediately bounces from the border into the channel, we execute the trade. If the price breaks the channel - we wait for the reversal inside and open a position again.

I wonder how you can tell if the day is flat before it ends... You can tell if the week or even the month is flat too... And this is the true grail:)

We define visually at the current moment. During a flat day the daily range borders are spread approximately equally and almost alternately broken by the price. On a trendy day, one of the range boundaries moves mainly.

Sart:
khorosh:

Sart If the day is flat, like on the Eu, then you may trade in the direction of the inside of the channel without waiting for the width of the daily channel to reach its middle width. If you look at today's eu (M5) you may see a lot of entries during the whole day. But not with the help of limit orders, but with the help of market orders.

Place orders. If price immediately bounces from the border into the channel, we execute a trade. If price breaks through the channel, we wait for the price to reverse inwards and open a position again.

One can talk differently about the daily average range...But look again, today the Canadian reached its 60-day average range exactly at the last "minute", so to speak - exactly 108 p., and then it froze in anticipation ....

In other words, the average daily range can serve as a very good reference point when considering any issue, it can be appealed to.


I agree. This is an important characteristic. The only pity is that if it is reached at the very end of the day, then by your method the day has been wasted. At best, if it is reached by 2pm, which is not always the case, you can trade for less than half the day. The efficiency of using the trading time is less than 50%. Although for those who work during the day the method is quite suitable.

 
khorosh:
Lord_Shadows:
We should alsoopen a position:

Sart If the day is predominantly flat, like today on the Eu, then you can trade in the direction of the inside of the channel without waiting for the width of the daily channel to reach its average width. If you look at today's eu (M5) you may see a lot of entries during the whole day. But not with the help of Limit orders, but with the help of market ones.

Orders. If the price immediately bounces from the border inside the channel, we make a trade. If the price breaks through the channel, we wait for the price to reverse inwards, and again open a position.

I wonder how you can tell if the day is flat before it ends... You can tell if the week or even the month is flat too... And this is the true grail:)

We define visually at the current moment. During a flat day the daily range borders are spread approximately equally and almost alternately broken by the price. On a trendy day one of the range boundaries moves mainly.

Sart:
khorosh:

Sart If the day is flat, like on the Eu, then you may trade in the direction of the inside of the channel without waiting for the average width of the daily channel. If you look at today's eu (M5) you may see a lot of entries during the whole day. But not with the help of limit orders, but with the help of market orders.

Place orders. If price immediately bounces from the border into the channel, we execute a trade. If price breaks through the channel, we wait for the price to reverse inwards and open a position again.

One can talk differently about the daily average range... But look, again, today the Canadian reached its 60-day average range at the last "minute", so to speak - exactly 108 p., and then it froze in anticipation....

In other words, the average daily range can serve as a very good reference point when considering any issue, it can be appealed to.


I agree. This is an important characteristic. The only pity is that if it is reached at the very end of the day, then by your method the day has been wasted. At best, if it is reached by 2pm, which is not always the case, you can trade for less than half the day. The efficiency of using the trading time is less than 50%. Although for those who work during the day the method is quite suitable.

So, the pivot points in forex:

1. Average daily range

2.
 
Sart :

So, the fulcrums in forex:

1. Average daily range

2.

2. I think the general trend on large timeframes is 4 hours minimum... (4H+D is ideal).

As they say, the trend is our friend...

 
goldtrader:

Gentlemen, are you looking at the date of the post you are replying to?

Sart

02.06.2008 12:59

Pay attention - today the pound has reached exactly the average range for the previous 60 days - 176 p.

i.e. "today" was yesterday, or rather now the day before ;)

Yes, my bad. I did not look at the date of the post.

And on June 2, the system really worked brilliantly - a clear rejection of the 60-day chanel. The only thing is that the specified range was reached too early(10:00 terminal time/12:00 Moscow time). According to my observations regarding the history of quotes, it was too risky to buy on the 2nd quarter. risky. Reaching the border of the channel so early sometimes (at a stretch, I would even say usually) provokes a strong impulse towards the border of the channel, which price touched last (the lower border, in our case). Nevertheless, sometimes everything works out according to the system, which was the case on 02/06.

 
So, this is what we have agreed upon:

Conditions for Open Buy/Sell:
- opening time: from 1-21 to 2-22 hours (time is chosen during optimization) - it is possible to divide it into three sessions and choose parameters for each one separately, but then the troubles with the range calculation will increase, so it is questionable....
- amplitude of the daily movement is equal (almost equal - +% is selected during optimization) to the N-day mean daily range
- price is near (+% is selected) its minimal/maximal daily value
- exit by TP and SL in percent of the N-day mean daily range
- all orders are market ones
- re-enter (if needed) by the pending orders (or trace, not by eye) when price returns (+-%) to its daily maximum/minimum.....


Is it correct? Or not?
Then it is not clear, where yesterday's range should be applied :(

And in general, there is some confusion about the terms: I understand "range" as amplitude without reference to concrete prices, and "channel" - ..... well a channel is a channel....
 

все ордера рыночные

I assume that just about everyone is a deferrer. At 00:00 of the next day, we already have all the information we need, because we know everything about the past period (day). Therefore, there is no problem to specify: here (+/-) we scale from here, and here - to settle. Why wait for the market?

re-entry

Strong IMHO - the system is clearly one-order. Either we have an order (I mean a pending one - see above, but will it work?) and it was "hooked" by the market, or the market did not reach us. That's it. But, again, this is IMHO.

Then it is not clear where yesterday's range fits in.

What do you mean? First of all - what range of the instrument we're measuring?

I mean the "range" - I understand it as amplitude without referring to specific prices, but the "channel" - ..... well, a channel is a channel....

Again, IMHO:

DailyHigh=1.2300, DailyLow=1.2100. Then we say: the range of the day was 2100 to 2300. And the channel width was 200 pips. Here, it is approximately like this.

In general, all the entry/exit conditions (from the previous post) are correct for the system. I do not like yet too many parameters to be optimized - it may easily slip into chicken-fitting. I think (I assume) that 1-2 parameters can still be calculated algorithmically based on current market conditions instead of being an extern variable.

 
SamMan:

all orders are market orders

I assume that just about everyone is a pending order.


I'm already talking from the point of view of creating an expert: it doesn't work like that - what kind of pending orders? We are talking about a certain range (+-%), through which it is unknown whether it has already passed or not.... and if it has only touched..... however, then Limit is definitely a limit :)..... but we still have to wait..... until we have decided on the ranges and then it's up to us to decide

re-enter

Strong IMHO - the system is clearly one-order. I.e. on a given day either we have an order (I mean a pending one - see above, but will it work?) and it has been "hooked" by the market or the market has "failed". That's it. But, again, this is IMHO.

Then it is not clear where yesterday's range fits in.

What do you mean? First of all - what range of the instrument we're measuring?

There, in the beginning of the post, we were talking about the range of the previous day.... That's why I'm asking, because I don't understand how it may be related to this case. I would understand if the direction is considered on higher TFs, but it's not clear...

What about one-order - it may be calculated in the opening variants. Very often, when the jag is formed, the first order is loss-making, and the second one is winning......

On the whole, all the entry/exit conditions in the system are correct. I do not like yet too many parameters to be optimized - one can easily get down to chicken-fitting. I think(suppose) that 1-2 parameters can be calculated algorithmically based on current market conditions instead of being an extern variable.

I don't like it myself :) .... i just think you can use the most important ones first and then all the others.... Optimization will become more complicated, of course, but if the results are acceptable, it will be worth it.... but if there are thoughts on "algorithmic", then open to perception :)

In general, standard ATR can be adapted to this case, improve it a bit and that's all :).... I'll give it a try.

 

пока с диапазонами не определимся и тогда уже решать

All the widths of all the channels (let's call it that) for the coming day are known to us at 00:00 of that very day. What else can we NOT decide on? If our system = waiting for a bounce from the 60-Day Channel - we have everything in hand by 00:00. At the same second we are ready to place Buy/Sell stop orders(since at 00:00 the price is in the conditional middle of the channel).

there, at the beginning of the post, we were talking about the previous day's range.... That's why I asked, because I don't understand how it is connected with this case.

Ah, yes! The point is that you can build a channel on its borders within (last) 60, 55, 30, 25... days. Or you can do it in 1 (yesterday). No one knows which channel is the right one. So we'll take any or the average of them all. We're not seriously thinking of taking the absolute low and settling in the opposite direction, are we? So we take it like this - a channel of 1...60 days. :) They, in fact, will not make a cosmic difference.

A more tricky approach - to take the channel for each day, based on some market feature (volatility, for example). But this is where the nuances come in.

But if you have some "algorithmic" thoughts, then you are open to it
Similarly, but for now it is nothing more than a wish to myself and the future TS. To begin with, it is quite possible to create a TS with parameters, and then "algorithmize" them (parameters in the sense).
Reason: