What's "normal" in forex?

[tomorton]

I recently posted elsewhere on the site my typical recent experience with a basket of forex pair spreadbets. Maybe this reflects your trading experience too?

Profit increases smartly to a given level and I'm feeling cheery. None of the individual charts show obvious exit points or reasons to be concerned so I continue to hold the lot. Then, within 3-5 sessions, most of that profit has evaporated and never comes back. The signals are effectively dead and I get out with one third of the potential gain.

So much for conventional wisdom of running your winners.

Possible solutions -

1. get in with better timing, stronger signals – limited potential here as I'm already doing this: I'm making money but just want more per trade

2. close each trade earlier to protect profit – close winning trades after x days: but how many? seems a bit random

3. empty the basket when it touches £x

4. empty the basket when unrealised profit suddenly jumps, ahead of market’s reaction to this

From what I've seen of my own forex trading, options 3 or 4 would get me the best deal.


But I’d love to understand if there's a more global view of the market that might highlight when market conditions "change", whatever that means, and thought I'd share this.

Starting with the trend-followers' premise of buying above the 200EMA and shorting below, as prices above have been rising and prices below have been falling. Looked at 35 pairs over the past 23 weeks to identify to what extent the market behaves “normally”, i.e. in how many weeks did price close higher if the week opened with price above the 200EMA and close lower if it opened with price below the 200EMA. The answer is 49.94%. That’s as near as damnit a coin toss.

Doesn’t sound like a very effective filter for eliminating weak signals. But do these 35 pairs behave “normally” every single week? Actually, no they don’t. The % of “normal” results varies from 17% to 74% between different weeks. My recent “good” week showed a 62% normal rating: last week, when my gains were much depleted, scored only 22%.

My assumption now is that this coming week and possibly the one after, there should be significantly higher “normality” ratings.

I’ve read many writers warning of trying to trade in the habitual way when market conditions change but I don’t hear them being specific about what that means or how to quantify it. I’m hoping to get closer to this and welcome anyone else’s alternative ideas.

 

[Dr. Toad]

I don't really consider myself qualified to answer this, but you may also consider:

5. Pulling half of your trade when you reach £x and continuing on as usual with the remainder. I have not done this yet but is probably the first thing I will consider if I feel I am leaving to much on the table in the future.

6. Using a tighter stop when you approach and reach beyond £x. This is something I have been implementing lately with some success. Generally when I approach and surpass £x and I see any sign of the trade reversing (even if it is not as strong of a signal as I normally look for), I start looking for a good place for an aggressive limit order out rather than a defensive stop out.

 

[fxstrategist]

There are so many variables involved and ways your trade may go that there is no clear pathway to follow. Stick to your rules, twick them, dont may huge changes, go little by little, be patient.

 

[NVP]

Hey tom

Good question.....let me have a think and come back a little more when in uk again

In summary a few points though

If you are trading correlated pairs (ie same currency in both).....then you will generally see similar patterns on the move happening on both pairs....it will take some significant divergence between the second and third currency involved to break this pattern as the move progresses and usually retracement phase will be at similar time

It is damn hard to get a strong and consistent extended move in forex baskets....this is because of the rapidly changing correlation patterns seen in forex.......sure the gpb may go long for a while but you need to be on the right pairs to capitalise on it initially before she retraces etc etc ....

I've talked about simple basket strategies over the years on my thread ......generally it's always pretty easy to spot the most divergent and aggressive currencies to buy or sell......the art is to blend in the secondary currencies that complement the move..........usd is the standard backstop most people use.......in my experience you have the aggressive volatile currencies like gpb and yen for example and then more smoother currencies like euro , usd, even aud can be smooth.....but again this is far too generalised so don't quote me .....

From my recollection looking at fx baskets most of the exponents "harvest" the portfolio as they go and then adding to the pot based on new signals .....so,in effect the portfolio is always live with some trades dependent on market ......see below

You might take a peek at some of the basket boys over at FF who have corresponded at length on this over the years .....

Dreamliner , FERRUFX, steve hopgood ......hanover .......the list is endless.......there was a classic thread called basket trading 101 I seem to recall.....I've corresponded with a few re basket forex trading over he years but in truth I find the gang at FF a little snooty and not open to any critique or ideas on forex basket trading ,,,,a shame as a few threads had some errors you coud drive a truck through when using forex strengthmeters......badly !

N

 

[tomorton]

I don't really consider myself qualified to answer this, but you may also consider:

5. Pulling half of your trade when you reach £x and continuing on as usual with the remainder. I have not done this yet but is probably the first thing I will consider if I feel I am leaving to much on the table in the future.

6. Using a tighter stop when you approach and reach beyond £x. This is something I have been implementing lately with some success. Generally when I approach and surpass £x and I see any sign of the trade reversing (even if it is not as strong of a signal as I normally look for), I start looking for a good place for an aggressive limit order out rather than a defensive stop out.

Cheers for this.
5 sounds like a sound approach and deserves the mention. I'm trying to aim higher, to get the biggest slice of pie. So I won't be doing this - it would be like having a smaller pie if you see what I mean.

6 I do like and if I can't do anything more radical, I would be happy to come back to this. Should reasonably protect 70-80% or better of gains I guess.

 

[tomorton]

Hi NVP - nice answer, thanks.

Harvesting baskets is surely the professional approach and I was hoping to mimic it but failing. It seems the major pairs are so well correlated either positively or negatively, a push in one is replicated in another, so everything goes with trend or against trend, even if the trends are weak or opposite. I'd like to be able to do this but I think I'm going to have to go for the cruder approach for the time being - keep the basket or dump the basket: no finesse and mocks TA but making money is the aim and justifies all.

One thing I obviously don't do is e.g. go long EUR/GBP and short GBP/EUR. I also avoid being both long and short USD or GBP - these charts don't tend to give me opposing signals anyway though it sometimes happens with the EUR.

 

[tar]

The word normal doesnt make sense here .

 

[NVP]

Hi NVP - nice answer, thanks.

Harvesting baskets is surely the professional approach and I was hoping to mimic it but failing. It seems the major pairs are so well correlated either positively or negatively, a push in one is replicated in another, so everything goes with trend or against trend, even if the trends are weak or opposite. I'd like to be able to do this but I think I'm going to have to go for the cruder approach for the time being - keep the basket or dump the basket: no finesse and mocks TA but making money is the aim and justifies all.

One thing I obviously don't do is e.g. go long EUR/GBP and short GBP/EUR. I also avoid being both long and short USD or GBP - these charts don't tend to give me opposing signals anyway though it sometimes happens with the EUR.

If you were doing anything I would advocate a tribal approach based on more global correlation patterns ....

Aussies Vs. Europeans Vs usd / yen


1 of these 3 teams will always be displaying more strength or weakness than the others .....so look to trade it against most opposing currencies when opps ,arise

Asian session is the classic for Aussie plays .......and I am always hoping for,the Europeans to dominate the moves in London session

The other trouble with dynamic basket trading of course is using / not using stops.....but that's another story in itself

N

 

[scott_miller]

I recently posted elsewhere on the site my typical recent experience with a basket of forex pair spreadbets. Maybe this reflects your trading experience too?

Profit increases smartly to a given level and I'm feeling cheery. None of the individual charts show obvious exit points or reasons to be concerned so I continue to hold the lot. Then, within 3-5 sessions, most of that profit has evaporated and never comes back. The signals are effectively dead and I get out with one third of the potential gain.

So much for conventional wisdom of running your winners.

Possible solutions -

1. get in with better timing, stronger signals – limited potential here as I'm already doing this: I'm making money but just want more per trade

2. close each trade earlier to protect profit – close winning trades after x days: but how many? seems a bit random

3. empty the basket when it touches £x

4. empty the basket when unrealised profit suddenly jumps, ahead of market’s reaction to this

From what I've seen of my own forex trading, options 3 or 4 would get me the best deal.


But I’d love to understand if there's a more global view of the market that might highlight when market conditions "change", whatever that means, and thought I'd share this.

Starting with the trend-followers' premise of buying above the 200EMA and shorting below, as prices above have been rising and prices below have been falling. Looked at 35 pairs over the past 23 weeks to identify to what extent the market behaves “normally”, i.e. in how many weeks did price close higher if the week opened with price above the 200EMA and close lower if it opened with price below the 200EMA. The answer is 49.94%. That’s as near as damnit a coin toss.

Doesn’t sound like a very effective filter for eliminating weak signals. But do these 35 pairs behave “normally” every single week? Actually, no they don’t. The % of “normal” results varies from 17% to 74% between different weeks. My recent “good” week showed a 62% normal rating: last week, when my gains were much depleted, scored only 22%.

My assumption now is that this coming week and possibly the one after, there should be significantly higher “normality” ratings.

I’ve read many writers warning of trying to trade in the habitual way when market conditions change but I don’t hear them being specific about what that means or how to quantify it. I’m hoping to get closer to this and welcome anyone else’s alternative ideas.

The more time you spend, the more you'll find out that - as someone already mentioned - there is no normal situation. Nothing is to be taken for granted because you expect it to happen. Expectations are what ruins traders. Adjusting them and confining them to facts and trends is what gives an edge.

About points 1,2 and 3 - I wouldn't count on those. They might work for what you're doing right now, but could limit you in other currencies or timeframes.

Keep gathering the stats, read the news for better contextualization of what's happening and you'll get to something resembling what "normal" is.

 

[tomorton]

Hi tar and scott_miller - Thanks for your input. At first I was going to fully agree, there's no "normal" in forex, in that the market cannot be 100% predictable.

However, in my very narrow context, normal means a chart which respects the 200EMA - that is, if price is above the 200EMA this Friday night, it will end the next week higher, and if its below the 200EMA this Friday night, it will end the next week lower.

A (very simplistic) rule of TA is of price is above the 200EMA, ignore sell signals as it will probably go higher, if its below, ignore buy signals as it will probably go lower.

So, week ending Friday 25/09, 22 of the 35 forex pairs on my spreadsheet ended the week "normally", following the 200EMA rule above. But the next week, ending Friday 02/10, only 8 ended "normally". Because I'm a trend follower, that meant my unrealised profit was well down. But why did performance drop from 22 to 8? What was in the news? Or is this just cycling between extremes?

 

[scott_miller]

I'd go for cycling between extremes of some sort, as the short term can cough them up. News can coincide with that, as well as pent-up pressure in one currency or another.

Btw you saying that you expect something in a week to happen based on a fixed point in time before that has some risk implied. The markets move according to their "internal" time which may vary from currency to currency. Otherwise all big movements would happen simultaneously for all currency pairs.

One way to go about this is add another indicator to see why the 200EMA isn't working. But I'd advise against overcrowding the chart with more than two or three indicators.

 

[tomorton]

Thanks scott, good things to think about here.

Cycling - At the moment, I'm clocking the "normality" ratings of my forex universe and would definitely be tempted towards closing all positions if this goes higher than 60% and certainly not adding new positions at this level. I'm even thinking if there's a need to drill into this to express it as a daily rating rather than weekly.

Markets moving to internal time - Actually, I'm seeing all currencies as being interconnected, not in terms of direction, rising or falling, but in terms of predictability (or normality: degree to which they follow the 200EMA rule).

Extra indicator - I do use a few other EMAs but no off-chart indicators at all. The 200 is the main average I refer to, because it is for so many people and reflects long-term trends. I'm not convinced it isn't appropriate, rather that the implications of the way it works aren't well demonstrated by the TA authors I've read.

 

[tomorton]

One additional point re "conventional wisdom" for traders.

A common rule stands against having many trades open at once: as few as one is enough, watch its chart like a hawk to know when to get out. But if individual chart TA doesn't give the best signals to help manage an open trade, as I find, but a basket does, maybe an even larger basket gives an even clearer picture? As I had to say above, do what the losing crowd does and you'll probably be a loser.

 

[Jason101]

Hi Tom,
Markets behaving 'normally' for me is whatever fits my set up criteria. If currencies are not fitting my criteria then I am not trading them. I have not traded currencies for a while now. Instead I am shorting problem stocks.
I am continually scanning currencies, commodities and bonds, waiting for conditions to coincide with my setup. But in the meantime I have reduced, size and am trading stocks that behave how I need them to, to conform to my 'normal'. It doesn't matter to me how long this takes.
My recommendation to you would be to broaden your market and be more pattern selective and accept you will not always be able to trend follow all markets all of the time.

Edit: ps why isn't this posted in our newly created Swing/position forum.

 

[tomorton]

Hi Tom,
Markets behaving 'normally' for me is whatever fits my set up criteria. If currencies are not fitting my criteria then I am not trading them. I have not traded currencies for a while now. Instead I am shorting problem stocks.
I am continually scanning currencies, commodities and bonds, waiting for conditions to coincide with my setup. But in the meantime I have reduced, size and am trading stocks that behave how I need them to, to conform to my 'normal'. It doesn't matter to me how long this takes.
My recommendation to you would be to broaden your market and be more pattern selective and accept you will not always be able to trend follow all markets all of the time.

Edit: ps why isn't this posted in our newly created Swing/position forum.

Well that's totally sensible Jason, if markets don't print your set-up, there's no green light to enter. My problem is, I am selective and go in on a proper trend signal, but at some point, all my trends turn back on themselves. Its pretty much simultaneous and not news-specific far as I can tell. I accept trends must "stutter", but they don't soon recover (some never do) and I'm certain I'd be financially better off cashing the lot in and getting in on the next round of signals.

What I'm hoping for is a way to time the early exit and looking here as it doesn't seem to come from the individual TA charts.

I posted here as this I suspect has a forex market characteristic answer?

 

[Jason101]

Hmm, my personal opinion on a forex market characteristic answer is that it is only possible retrospectively. Perhaps you could reduce fx exposure to correlated trades?
It's my belief that no matter what the pair, fx (and most other markets!) there seems to be some correlation of some sort to some degree. My personal exits came in the fx when my trades failed to achieve a higher high (but equalled the last high) and then the retrace made a lower low.
Also I just think that is the way it is. The cost of doing business
Sometimes you will give up more paper profits than you want, in exchange for other times when you will receive greater profits when the trade continues.
I just don't think (as you eluded to) there is any way to tell if a retrace is the end of a trend. And sometimes I can end up adding to a position just at the wrong time, when the retrace is in fact the end of the trend! Very painful.

 

[tomorton]

Hmm, my personal opinion on a forex market characteristic answer is that it is only possible retrospectively. Perhaps you could reduce fx exposure to correlated trades?
It's my belief that no matter what the pair, fx (and most other markets!) there seems to be some correlation of some sort to some degree. My personal exits came in the fx when my trades failed to achieve a higher high (but equalled the last high) and then the retrace made a lower low.
Also I just think that is the way it is. The cost of doing business
Sometimes you will give up more paper profits than you want, in exchange for other times when you will receive greater profits when the trade continues.
I just don't think (as you eluded to) there is any way to tell if a retrace is the end of a trend. And sometimes I can end up adding to a position just at the wrong time, when the retrace is in fact the end of the trend! Very painful.

I am trying to identify a changed market characteristic that tells me to get out - I don't need to do it predictively, real time will do. Intuitively and what all traders say and what all trading writers write, is that there are "changes" in market charcteristics or behaviour that will eventually suddenly throw your plans out, whether for entry or trade management. At that time, you should get out and re-set. They don't say what they're talking about, even retrospectively: I'm trying to work out at least one candidate.

I agree, all fx prices move simultaneously with respect to each other. That's exactly the issue. So there is some relationship between e.g. EUR/JPY and GBP/CAD. I don't necessarily have to learn what it is, as long as I can see if its proceeding "normally" or are we in a storm situation.

Tootally agree on trends. Just as a retrace may not be the end of a trend, I also never see retraces coming - if I look at the charts. But I'm eaten up with the notion there is more going on than the sum of a lot of TA and if I can get a scent of it I could keep a lot more profit per basket of trades.

I hope this isn't tedious, thanks for your thoughts. I'm excited that I'm getting caution but not being certified loony.

 

[tomorton]

The "normality" score fell further, to 14% on the weeklies, so I'm hoping this will prompt a reversion to the mean and the pairs of interest will from here resume following their primary trends for a few weeks at least.

 

[Forexmospherian]

Hi Tom

There are ways to catch interim highs and lows in intraday trading FX and one of the major clues is time

Every day I use my so called "time rules" as another clue to assist my other TA reading

I will never be 100% accurate to the pip on a interim high or low for a day - and will only know within 30 mins of it generally happening - but even 70% accuracy with tight stops makes great returns

Time intervals are so important in intraday FX trading but very few gurus etc want to share what they probably know about . Its become more important over this last 10 years with more automated systems ( HFT ) etc

This last week as been brilliant for intraday FX trading some great moves both ways on my major pairs.

With regards to swing or positional FX trading - not really my area. I have seen in the past some correlations on turns after 8 and 12 and 13 trading days as well as always the start of new months when the daily / weekly / monthly charts all paint a similar picture.

I require tight stops with multi trading and larger RR's off winners - so although I do look at the weekly monthly and even occasionally a quarter chart view - I still then revert back to my tick or 1 min charts for all new trade entries.

Hope that helps - and also agree with your comment in another thread - the FX markets are not totally random - they are readable - but its takes years and years of experience and chart reading to fully understand the whole "picture"

Good Trading to you

Regards


F

 

[tomorton]

Cheers F. Encouraging words from you especially re "cycles". This goes beyond the obvious trends/pull-backs and accumulation/distribution on individual price charts. I do think there's something more than each chart can tell us (we're over-focusing on the wrong charts). Hoping to acquire sufficient evidence that such pictures will become quite self-evident.