Zeusjoes Market-Neutral Hedge Strategy

This is a big problem!
If you give for StartingTimeCorr an other date as our "2008.07.01 00:00", you can see more bars, BUT the range values changes!
The algorithm from the original "OverLay Chart" indicator is very nice, but is functioning only on the current window.

I just try to find an other algorithm with a moving time-window, but the solution is not easy...
Here my first attempt:
.

Try this modification, which paints the bars back about a month now...:

(I used the "old" v2, because I don't have the mq4 for v3, but the modification is very easy.....)

If we use v3... .then it would calculate the threshold based on 500 bars, which would be GREAT!!!

//-- Calculate Visible Bars
_BarsCount = WindowBarsPerChart() + 500;
int _FirstBar = firstBar; // WindowFirstVisibleBar();
int _LastBar = _FirstBar - _BarsCount + 500;
if ( _LastBar < 0 ) {
_LastBar = 0;
_BarsCount = _FirstBar + 500;
}

Looks like we still have to input the start date, but it really is "now" then works back... can we just leave it be the current time, no input??

Here the Version 3.1 for you.

The problem is that for trading decision today, you calculate the range from today. After x days you have for today an other range as today (is clear...?).

The value what you see in the oscillator 5 days ago (e.g. 100), was 5 days ago maybe 80 or maybe 110!

The chart and the oscillator works with the position of the close between the HiLo bands. For short periods (50-60 bars, e.g. 2-3 days) you can use the range values for trading, but for larger periods (>100) the range values are unrealistic.

Can you please post (or PN) the 8 entry times in June? I will check if with the version 4 is possible to find (historical) the same entrys. If not, I must find an other solution.

I had good results on Thursday, which was a NFP day (released early due to holiday.) It may have been luck, I don't know. But I did trade by the rules and am up some on my account.

So far today (Friday) the market is flat and I don't expect much more movement with the US markets on holiday. I'll hold all open positions over the long weekend and see what happens on Monday unless the spreads reach 0 before the market closes today.

I'm also playing with the MM and the stops on this since I don't have a clue how far the spreads can go against you. It's going to take some time to learn. Who knows, I may blow out my account because I'm doing this all wrong. I'll let you know. That's why I'm using a demo.

I plan to just watch and see how these spreads work out before I enter any more trades on other pairs so I can keep from getting the results all mixed up together.

I'll keep you posted.

I was wondering if it is more useful to set the mirroring on the indicators to "true" or "false" or does it make any difference?

Thanks.

Thanks! SwingMan... I know what you mean about the threshold determination over time... But it is great to be able to paint back a month, see where the extremes were, and set a threshold with our good ol' brains!!

One question though... Looks like we will have to daily update the StartTimeCorr entry.. If I set this at say, 2008/6/5, the indicater will start at about 6/10 and paint BACK... set at 6/30, the indicator stops at 7/3,. So at a 7/1 setting, I figure it will work until about 7/6 and then not paint any current bars... meaning we would have to reset the date then??

Is there a way to have that value pick up the current date/time, thereby always updating itself? And we don't need that input? Would that mess up the calc? Doesn't appear to.. If I set it at 7/4 or 7/1, the indicator looks the same, and accurate...

Very nice work!!

And Pistol Dave.... sort of depends on you, and how you like to "see" the relationship between the two pairs. Having the setting allows us to setup the indicator (either one) to show us what we like to see for either positively correlated pairs (GU/EU) or negatively correlated pairs (EU/EChf)...

Hello Swingman,

I am impressed with your work. Do you think a fully automatic ea could be developed from the indicators?

Jeff

sorry if this a stupid question.

but while the line made by mrkam (EU) was close of GU since the ende of the day the oscilator was showing the correlation as super oversold.

i suppose the oscilator was wrong, is it?

thanks in advance.

To better use our brains, is possible to give two dates (begin/end), and to draw the prices for this period. Then can you the "what if" back analyse.

My idea to "updating itself" the begin of the calculation period is to give as parameter "2-3 days back". With this input we have ranges that not changes every hour.

Yes

I think there are some minor calculation errors, but you can test the new indicators. I'm going
(The old v3 indicator has also a small error in the calculation of the threshold).
.

A few small changes in the indicators.

- If the main chart is GBPUSD or EURUSD, the appropriate SubSymbols EURUSD and USDCHF and the Mirroring parameter are set automatically.
For other SubSymbols you must enter this symbols.
.

Wanted to give everyone a report on the results I had from last week's trading this strategy.

I am going to try to trade the system as well as I can, by the rules that I understand so far. I'll do it as long as there is some hope that it is profitable.

I will try to report after the end of each trading week. Keep in mind that I have a job where I may not be able to enter/exit every trade at the best possible times. So how well I do may not be as good as a full time trader could do with this. I'll do my best to take only trades that go by the best rules we can come up with.

I hope everyone keeps sharing their results, charts, ideas, so we can see what works best. A good EA would be nice too.

I started out trading on the 1h timeframe and I'll keep using that unless I see it's not going to work for me and I may try others. I'm setting stops at 130 pips on each pair and using mini-lots on each pair (0.1 lot).

Beginning balance: Opened Demo with $1000 on July 03 2008

Since I opened the account late in the week I'm going to use this amount as the Beginning week balance for the week June 30 - July 4.

Beginning week balance: $1000
Ending week balance: $1508

I realize that this is an exceptional result and it was only due to the NFP release. And I probably was using way too much leverage. But over a month or two of testing the results should even out to a more normal variation. Or I may blow out, who knows.

Time will tell.

Only one idea for the analysis of entrys:

Take the trades if

1. Setup
HedgeRange > Threshold

2. Confirmation
Parabolic condition
or
LWMA Close(5) cross the LWMA Open(8) condition

If one look at the historical HedgeRatios (osc_v4), I believe that we can avoid large drawdowns.
.

Which pairs?

Hi SwingMan,

Thanks for doing the research on the indicators. Looks good. I'm still using the osc_v2 but I'm going to try osc_v4 and see how it works. I think I tried osc_v3 and had some problems with it.

The pairs I had success with are:
GBP/JPY - EUR/JPY
EUR/USD - USD/CHF
GBP/USD - EUR/USD

I'm posting this screenshot of that days trading so you can see the entry times, etc.

The last version for the NeutralHedge indicators:

- The threshold is calculated as middle of the HedgeRanges + Factor * StdDev. The indicators takes as default value Factor_Threshold=0.0, and you can test other values.

- If you enter the Last_OverlayTime (e.g. 2008.07.02 20:00), you have in the Overlay indicator and in the Oscillator 3.2 the historical analysis at this time.
(In the Oscillator v4.2 you have automatically the HedgeRange and the Threshold values as Histogram and Points).
- The Overlay indicator draw also the ranged prices of the SubSymbol from the last overlay time to the current time (dot line).

- If you enter the Trade_EntryTime (e.g. 2008.07.02 20:00), and not forget(!) to enter the OrderType (Buy=0, Sell=1), you have the current result of the trade. Also you have the maximum Win between!

- The default value for the calculation of the HedgeRanges is set at 72 (3 days x 24 hours = 72 bars).

- For four pairs in the main chart, the indicators take automatically the SubSymbols and set the Mirroring parameter:
GBPUSD - EURUSD, true
GBPJPY - EURJPY, true
EURUSD - USDCHF, false
USDJPY - AUDUSD, false

Remark: is possible that the indicators repaints, but now at the weekend can I not find this.

As example, a study for the USDJPY / AUDUSD pairs in the H4 time frame:
.

SwingMan, I can see you have been hard at work here. It looks great. The indicators are going to be very helpful.

It looks like you have been able to determine a high probability entry level with the NeutralHedge indicators. The psar should also help to tell when the price begins to swing back toward correlation/non-correlation so that would be a high probability entry point. I don't know yet if we would be giving up too much profit by using a lagging indicator like a MA for entry though. We will see.

All that makes sense to me so far. (But it will still take some time in real trading to see if it works out.)

I'm excited that you have made this much progress. And I hope I am understanding all this properly.

One other important aspect to consider is the exit of the trades. For my exits, I used the original idea that was suggested in one of the first posts where someone said they had been exiting when the spread (I think you are calling it the HedgeRange) goes back to zero. That's where I closed my positions and in this particular case it caused me to lose over 100 pips potential profit from the trades. By looking at the chart it's easy to see that I should have stayed in a lot longer.

Obviously, where you exit will determine whether this system proves profitable or not over the long run. Do you think closing out at zero is the best way, or maybe there is something else we could look for to indicate that would let us know that the run is about over for that particular trade?

The stops will definitely make a difference too. I was originally thinking around 100-150 pips. That sounds like too much, but because this method is so much different than anything else I have used I'm going to give it plenty of room to move until we see what is going to happen. We still have to see what the risk/reward is going to be.

I'm open to hear any ideas you might have. BTW, I hope we can get some input from the other guys that were reading this thread earlier since I know some of them are some very sharp traders.

Dave

Just a stupid question How can attach the indicator and where about? I can not make it works thanks