Zeusjoes Market-Neutral Hedge Strategy

Just a word of caution....please do some research and import some data into Excel using MagicTraders Excel spreadsheet and backtest. I've done 4hrs and 1hr for 2006 and 2007 (with my own formula) and I've seen spreads get to 1000+ pips. It's just like anything else you need to get a game plan together cause this has the potential to wipe you out if a persons not careful.

Trade Well....
Brent

Yes.... very interesting math to address the same issue....

One question:

In your spreadsheet, you set 7/25/08 at 20:00 as the "ZeroPoint", that is, where each currency is the same, as your basis. Then you track the up and down from there, etc...

How do you pick that point in time to set zero? Is it arbitrary, or do you have a method for determining When to "reset" it to zero...

I have the same concerns. It's tough to know where quilibrium is. That is what I like about SwingMan's NeutralPriceOsc_v1 indicator; it is dynamically recalculating equilibrium and showing the standard deviation from it.

Magictrader's approach is less scientific, but fundamentally can work, but IMHO is difficult to determine where the trade triggers should be. Set for too low or too high a spread and you may not get very many trades. With some trial and error you can probably come up with some decent triggers, but they may need to change with time and fundamental changes.

-

It was just first history data from broker, but I think to place "0" maybe at market opening in sunday and reset it every week.

There are two possibilities with different indicators:
1. Repaints and more probability that pairs will cross in future, also time frames to choose.

2. Set the start data, then no repaints, and I don't know with what probability pairs will come back, but I think that setting starting date e.g at the begining of the month, and reset it every months it is like larger TF from previous indicators, and setting e.g. one week starting date is like short TF. Maybe something is in it.

So spreads for TF reset every week should be smaller.

Yes the starting point is the key. It all reality your starting point could be when the spread is already 200 pips apart. Not sure what you mean by SwingMan's indicator dynamically reclculating equilibrium. He has the same delima. He has to pick a starting point from something.

Personally I go with the peaks/valleys. So on my charts i go back to where the two currency's last peaked or had a V-shape at the same 4hr candle and I use that close of that candle as my starting point.

Trade Well....
Brent

SwingMan's indicator, as I understand it, does "mean" calculations from the current bar going backwards, so in simple terms it is calculating the most recent avg price difference with every new bar.

Apparently he is using a logarithmic method to weight the avg in some manner. He posted an excerpt from Ehrman's book on "Normalizing Pairs Divergence", which is a bit over my head at the moment.... but I like the results I'm seeing.

-

Serios repaint here
I notice that the only diferent thing is the first/last bar date. Whats the problem?

For your review and comment, I am attaching the .ex4 of the current SynchBar Indicator. Thanks all who have shared in the programming expense of this, all costs are covered, and those who "helped" received the mq4 file,... (and THANKS for the help on that....) That will be that.

Also shown is an example screen shot.

THIS IS A WORK IN PROGRESS!!!!

SynchBar needs the NeutralHedgeOscillatorv3.4 on the chart in order to paint it's bars accordingly. If you don't want to paint the bars according to the NHO calc, you don't need it on the chart. I have it on the example, then minimized by dragging the borders. (Ii have noticed it takes a few minutes for some reason for Synchbars to get matched up with NHO...

The Indicator works like this:
1. You set a "# of Bars" parameter, which is simply then number of bars the indicator will show. I have it default to 240, but it can be anything you want. If you use NHOv3.4, your #of Bars HAS to equal Period_range in order for the colors to line up.. (240 bars at 15m is about 2.5 trading days back).

2. Each bar is simply the mathematical difference between Curr1 and Curr2, with NO equalling factor applied. JUST THE DIFFERENCE, raw data. Short bars, the prices are closer, long bars, the prices are further apart.

3. The Indicator looks at the number of bars shown, finds the High and the Low, and calculates the mid-point by (Hi+Lo)/2, and plots it as the Gold line. This is simply inteded to be a reference "ZeroLine", halfway between the period Hi/Lo you select.

4. The red lines are manually set "Entry" Lines. Again, you can make these be whatever you want. They are set at X pips from the ZeroLine.

NOTE: Both 3 and 4 are dynamic, meaning they may change. This is NOT repainting. What they do is change dynamically according to the market. If you add a new bar that sets a new hi/lo, then you get a new ZeroLine calc'd, and new EntryLevels set. If you drop off a high or low at the first (oldest) bar, then you also will get a re-calc, and new Hi/Lo/Entry lines.

(There is a trading logic to this, but you can noodle that out!!)

This is similar to SwingMans' difference Log indicator, just a bit more basic, more "raw". The indicator itself is very basic, showing strictly price action, and the ebb and flow of the two prices as they move closer and further away from each other. It is this flow that we are endeavoring to trade.

The purpose of tieing to the NHO indicator is to try to get a determination of trade direction. (and there may be better ways to do this!!)

For example: (and this is just one idea of how to trade this, maybe....)
1. Find a # of bars you like. Set SynchBars and NHO to match.
2. Set the Hi/Low entry lines at what YOU want for entries. You can enter either at the high or low, for a high(er) probability movement to the Gold, centerline.
3. Red bars mean SELL Curr1, BUY Curr2, Green mean the opposite.
4. Take the trade at a touch from the inside of either entry line, Buy/Sell per #3.
5. Exit wherever you want. Possibilities would include: fixed # of net pips profit, or at the touch of the ZeroLine (gold) or, all the way to a touch of the opposite entry line.
6. Set a MaxLoss, of course....

REV: PS: you need to have BOTH charts open on the time frame you are dealing with in order for the Indicator to get it's data for BOTH currencies....

Any other ideas??

SwingMan told us his NeutralPriceOscillator v3.4 repaints. I don't really even use that one.

The only area of importance, again in my humble opinion, is the current/last bar on the NeutralPriceOscillator_v1 and whether or not it is surpassing my upper/lower thresholds. You trade in realtime and you're looking for a deviation that may last only minutes, so your trigger has to be in the last bar.
-

Agree, same experience here with NeutralPriceOscillator_v1 has precise entry again today making pips in hedging pairs. This is the one to shine the way running it for me. Very nice work by SwingMan. This forum is super.

For now on USD/CHF and USD/JPY spread is over 100 pips with starting time of first candle in this week. So maybe repainting is good for us, because everything is changing when market developes.
But with repainting sometimes there is no gain on pair cross, and we need to add some positions. With starting time, we will have our spread gain when it will come back to "0", but when it come back, and how big spread could be? Nobody knows.

Agreed, and we have all discussed this in the past as well.... past bars on the NHO don't really mean anything (in realtime trading), so it doesn't really matter if they repaint or not. The current bar is what matters. The past bars would only be valuable if they did not repaint for past "backchecking", but I for one put a lot more credence in forward testing...

As far as SynchBar goes, the same thing applies, although the past bars DO help us in giving us a visual (and quantative) view of how the prices have been moving with relation to each other.... which possibly can give us a better "feel" for entries and exits.

The ONLY calc SynchBar uses from NHO is the color of the bar. Color, that is it. This "might" be helpful in determining which currency is stronger compared to the other, which was the purpose of the colors. We can then determine Buy/Sell or Sell/Buy in our entry.

In my testing, the colors seem to work pretty darn well! As does using the current bar for entry... it is just that determining an actual entry level via NHO is tough because the view of history is skewed... hence the idea for SynchBar to give an accurate, non-changing, view.

Great approaches and ideas!!

If I may ask, after which rules are set the red and green bar colours?

Hi Swingman! Sure.... per your NeutralHedgeOscillator v3.4... SynchBar "gets" the colors from your indicator! and I don't know the calc it goes through...

The bars don't match up exactly, but I have found that in realtime, at the extremes of the swings (hi & lo areas), the SynchBar colors match the v3.4 colors...

OK, if you say so ...
Try also the two indicators from Ehrmans book. They should also be very useful. Unfortunately, I have no time to test it, I must develop other indicators ...

There is also one interesting thing in selecting price (time) strategy, that if we enter e.g at delta 30 pips, then if delta become e.g 90 pips, then our loss is exactly 60 pips. At the beginning we know our gain (entry delta) but we don't know our loss. We can know possible average high delta in forward testing because bectest every week could be difficult.

The best thing is, that delta is exactly the ammount of pips to gain. It looks like in this kind of trading risk to reward ratio is mirroring to standard what everybody learning. But it looks that we've got very big probability to success in each trade.

Indicators, Indicators, Indicators... what is it with people and Indicators?!?!

I do not even know how half of these indicators work... you guys are really complicating this system more than it needs to be.

MagicTrader is using a spreadie with what he 'quotes' with the price deviation, I think that is more than enough... I did a trade with a chart overlay, I requested Swingman to do a chart overlay with a date/time parameter, but not heard anything.

G/U and E/U are not a highly correlated pair, but I made money today, by the simple chart overlay and that is it... price deviation and chart overlay... nothing more.

This was done on a 5 min TF - scalped around 10 pips.

Regards,
Vip

guess121


Indicators are only the visualisation of spread sheets, that's all.
I think the first version(s) of the overlay indicator has a date as begin of the calculation.

Hello,

look at this: no news, but so big movement on EUR/GBP. It's theory but maybe some big boys closing their hedge positions?

Ok, here it is in reality what we are trading if we set some start time:

The thing is, that somehow delta "0" point is still moving. It's not like this if we start counting delta from 20:00 monday with value of EUR/GBP at e.g. 0,7850, then delta go to 60 pips, and "0" will not have to be at 0,7850. Question is, is it possible to calculate where "0" point could be. And now if we look at EUR/GBP we've got sideways trend, but since 2007 was strong up trend. How then this system works?