Zeusjoes Market-Neutral Hedge Strategy

This is a great topic, one of the best so far on FF.

However, I have read so many topics like this on other forums and at the end of the day, it all comes down to speculation, risk, and reward.

Is a trading system like this really that much better then a very simple moving average system? I am still undecided.....

Better?? you should know "better" than to ask that!!!

Better in what way? Better to who? Profit? Just making the "most pips" is not "better"... what is your trading psyche? Do you like "always in" systems... if so, you just made 400 pips with the $... but you have lost 100p here 100p there for months... Do you like scalping... 10p here, 8p there... no 400p gains... but you always crank in the pips day in day out regardless of the market... How about risk adversness... are you good with losing 200p? How about 40p to gain 10, but happens once out of 20?

Better? Maybe not the question.. Different! now we are talking...

Everything is speculative in forex. What we are endeavoring to do here is fresh. Who knows if it is "better", but regardless, it sure is fun!!

Stick around see what happens. Help us with your insight...

It's a good life!

Vip

I still get the same results.... it does not work...
I have Microsoft Excel 2002 version.... Is it possible for you to save the worksheet in that version?
Probably by now you have updated what you have been doing with the worksheet...
Thanks for sharing.

Bruce

The best way I can answer this is: Look at friday's action when EUR/USD fell much further than GBP/USD. But when GBP rallied a little EUR rallied even less. In some instances GBP rose while EUR stayed still. So it appears to me that EUR was the one out of synch. Therefore, it made sense to buy GBP/USD and sell EUR/USD.
Regarding the lot size I prefer 1 to 1 but if you like 1 to 1.26 I suppose you can try it. Like I said before, I am not looking for the perfect hedge. With a higher ratio, you may well make more profits but since this system in not foolproof, it will also increase your losses. Also you must time your entry very well or else you will be carrying a hedge that never correlates you into profit.

Hi guys. I'm a big fan of this kind of operation. In the stock markets we know it as Long/Short position. Here in Brazil we do for example long on the EMBR3 (Embraer- a flight company) and short on PETR4 (Petrobas - oil and energy), and we got a lot of bucks. So im trying this on forex... My doubts:

1) Which pairs and why to operate then? What's the main link between the pairs Ihave to observe?

2) For the GbyJpy and EurJpy what is a good spread to open a position? I got some profit from spread araound .4110.

3) Whats the behavior of the spread? Does it usually stay in range or can go to "infinity"?

4) does someone have the historical data of the spread?

Well, i made some program on visual basic. It takes the information from the MT4 and saves on a text file (values separated by semicolons).

If you want, try it. The file generated is in the format: "GBPJPY at the start of the period, EURJPY at the start of the period, GBPJPY at the end of the period, the spread in the begining of the period, the maximum spread, the lower spread and the closing spread. Each data line corresponds to 5 minutes.

To use it start MT4, start its DDE server (Tools > Options >DDE Server), turn the program on, click on "Register All". It will create a file and register the data. After u're finished u can stop the program, export data to excel and plot some graphic. I hope it helps. There are some words in portuguese on the text, but they are similar to the english counterparts.

OK, just to keep things going here!! Looks like we just went through the "risk" part of this approach,.... that we get a sustained move that "uncorrelates" our currencies, then re-settles....

I traded this two ways:

1) Standard Synchline & RSI entry, which signals the trade at a Synchline entry level, and determines direction by the RSI on the crosspair one time-frame higher. (>50 buy GU, sell EU, opposite for <50)

This entry has me at -85p right now and did not exceed 100p negative during the whole move... not bad if this is "worse case".

2) Advanced "2nd RSI touch" entry. In this one, I get a trigger from SynchLine, at a given price difference, then again check the crosspair RSI at one time frame higher. But I only enter on a touch of the RSI movement from above the 70 or below the 30. This means that it is the 2nd touch, once to break the RSI, and the 2nd to go back. The thinking is to allow the market to reverse it's pair strengths, and return back towards our profit direction. This one has me at -29p. Not bad at all.

But now, the pair difference seems to be settling in, perhaps for a new "synch". So, I could closeout for a loss, call it good, (this is not a bad option, really, for if this is our worse case, not bad!!), or leave it open to see if the Synch works it's way back (could take a long time...), or leave it open, but continue to trade at the new Synch Level it is establishing now...

Very interesting!!

Where are your stop loss? Or when do you get out the position to cut your losses?

Hello,

check this out:
http://www.fx-hedgefunds.com/modules/tinyd1/

Here is the simplest way to trade with this strategy:
Spread EU-GU in 90-95% looks like EUR/GBP chart, so on EG chart insert Bollinger bands with 200 period average and 2 and 3 standard deviations.
If you look closely on EU, GU, EG, or any other chart you will see that almost always price bounce from 3rd standard deviation on every TF.
How to trade:
watch EG chart, when it's going up just wait for price to be between 2nd and 3rd StDev then you could SELL EU and BUY GU if you will loosing then add position on every 15-20 pips lost from last position.
The same for EG in down trend.

You can try with EG price touching 2nd StDev, then you will have more entry signals, but when trend will be strong then your entry will be too early and drawdown could be big.
You can do backtests because Bollinger bands won't repaint.

Hi Magic... got your PM... I have been using the CrossPair chart to determine direction as well... but using the RSI on the next time frame higher...

But your BB idea looks great as well!!

As a further update, my SynchPair position that I had just prior to the big drop last week just closed at +8p exiting at the SynchLine "ZeroLine" touch!!!

This is great news. Clearly, I was on the wrong side of this move... 400p move!! We don't get those too often. At one point, I was -85p, but closed at +8p. So, if we can weather such a large move being on the wrong side, to come out at breakeven, we are doing pretty good. Now, there still is NO guarantee we will come out that way every time, but we know a couple of things:
1) This kind of 400p move, and it's corressponding wide de-correlation (or out of synch), does not happen very often.
2) There is always some sort of re-correlation move after the big move.. maybe not all the way to breakeven (this one was), but at least some sort of move.
3) Looks like a 100p MaxLoss would be a reasonable setting.
4) Even if we took a 100P loss, I have had (3) 30p wins since then, so in less than a week of this move, being on the wrong side, I would have paid back the loss.

All in all, very encouraged...

Today was a good day. +86 still demo pips.

Magictrader,

Could you explain your entry strategy? Noiced they were taken at different times and always the same sell eurusd buy gbpusd.... and then most were closed around the same time..

First I establish that spread between GU and EU is like normal distribution, so entries should be between -2 and +2 standard deviations (95% values in normal distribution are between +2 and -2 Standard deviations). So I've done in Excel standard deviations for GU-EU. When spread between GU-EU goes to +2 or -2 StDev then I look at cross pair EG, if there was up trend then I Sell EU and Buy GU, if downtrend then opposite.
But what I noticed for EU, GU especially, that their spread almost looks like EG chart, so I just add 2,3,4 standard deviations from Bollinger bands 200 period average. But remeber we are trading spread between two pairs, not a cross pair. If you try to study other pairs, then you will see that e.g. spread between EU and AUD/USD looks more different from their cross pair EUR/AUD. So standard deviations for spread between two pairs will not be the same as for the cross pair, so different StDev = different entry points.

But for EU, GU, you can use EG chart because in 90-95% it is the same as spread between them.

ENTRY POINT:

So your first entry point could be when EG is between 2nd and 3rd StDev (I use M5 chart - usually 1 to 2 trade signals per day). Agressive entry could be when price touches 2nd StDev, and conservative when price is at 3rd StDev. To know what to sell and what to buy simply look at EG chart and take opposite trades.

DURING THE TRADE:

There are two ways:
First is to add positions when your first trades are loosing in total 15-20 pips. Then if your last positions again loosing 15-20 pips then add another positions and so on. If you are playing like this you should have maximal drawdown (e.g. 200 pips, when total loss hits -200 pips then close everything).
Second is to open one pair of trades with less maximal drawdown, maybe about 50 pips.

TAKE PROFIT: the best take profit point is when you enter at about -2,5 StDev and price now is at about +2 StDev (some kind of stop and reverse strategy).
Other way is to exit when you reach your TP level e.g. you want to earn 20 pips, when you got it then close all.
And last way of clossing positions is when you add more positions then you could close it on different StDev levels e.g. if your entry was at -2,5 then close some at -1,5, -1, 0 or choose another levels you want.

TIPS:
you can use for calculating StDev linear weighted average, it better follows strong trends and thanks to that entry point won't be to early. Below is the modified BB indicator.

Hope this will help.

Very nice magictrader...congradulations on very smart method.

As I write Eur/Gbp is at about +3 standard deviations on 150 pd 30 min. single regression channel. This is a significant departure from median line. +3 SDs equates to a price of 0.7870 Eur/Gbp, while median line is 0.7797. At this point it must be determined which of the two curriencies is primarilly responsible for the divergence/decorrelation. Again single regression channel is used. I look at Yen crosses. Gbp/Jpy is nearly -3 SDs below the median line on 150 pd. 30 min. channel at 207.215, as I write. +3 SDs equates to a price of 207.031, while median line is 209.422. This is a significant departure from median line. On the other hand Eur/Jpy is remains near to it's median line of 163.248 on 150 pd. 30 min. single regression channel, at 163.16. There is little departure here, nothing significant, probably about -.25 SDs. Weakness in GBP, not strength in Euro, is primarilly responsible for the divergence/decorrelation and the large scale movement in the Eur/Gbp pair. In this case I will buy Gbp/Jpy and sell Eur/Jpy in proportion to their relative contributions to the state of divergence/decorrelation, and in proportion to their valuation to the Yen. Note Well! : when using regression channels it is necessary to keep in mind that they are always in a state of being drawn, and because of this, the distances of historical prices from the median line/SD lines were different at the time these prices were registered on the chart from where they appear at the time the current bar is being drawn. This is because the angle of the channel changes as the accumulation of price values becomes either more negative or more positive. This can be very misleading. Of course, this is not trading advice, just the way I trade this set-up...I hope this helps. Pippo

Today is very interesting day. Without SL technics we could clean out our accounts. Look at this strong trend on EG. Spread now is huge.

hei everyone, any news? hows the result of your research? do you have other trading idea related to this technique?

I've found something safetier. I'm just calculating standard deviations for spread EU-UC. When it hits +2 or -2 I'm waiting for confirmation (usually candle formations on M5 or M15) then I pull the trigger only on EU.
SL=25 pips
TP=60 pips
so it's nice R:R ratio without adding more positions, that's why it's safetier.

Here are the statistics for last 8 signals (I start follow it last week):
5 x +60 pips
3 x -25 pips.

Even with 30% of good signals it's still profitable. I'm counting on 50%-60% of good signals. In one day there are one or two signals.

Hi magictrader,

what do you mine EU-UC ? EURUSD - ? what is the pier UC ?

can you add a TPL ? or scrin pic' ?

tanks

benseg,

Please try to type whole words next time.
To supplement magictrader's post, and hopefully answer your question:

EU means EURUSD
UC means USDCHF

By taking the difference, or ratio of prices between two well co-integrated pairs, you establish a much more stationary time series. That would be the EU-UC time series that magictrader is referring to.

Of course, you must compensate for the fact that USDCHF's correlation with EURUSD is near -1, which means that when you calculate the price differential or ratio, you had better invert the USDCHF price to CHFUSD beforehand.

Hope it helps,
-Shuli