Trading Fibonacci numbers

[tech/a]

"Drink till it's gone"

Post number 2
From another quality poster
With nothing to add to the
Community

 

[tech/a]

Some software will draw swing charts for you. I have found that no individual indicator is a reliable stand alone and swing charts are no different.

Here is a small selection of links from a quick google search of "swing charts":

http://www.investopedia.com/articles/technical/04/080404.asp

http://stockcharts.com/school/doku.php?id=chart_school:trading_strategies:swing_charting

http://www.stock-trading-infocentre.com/swing-charts.html

Sails
They can't be used as a trading tool.
For the Same reason you can't use fib function in a trading systems test.
The last swing can't be confirmed until it has corrected the % chosen.

Let's take a 10 % swing
You see a pivot and take the trade stop a few ticks above the
Pivot.
It drops 8 % swings back and adds 15 %
the original pivot disappears and just appears
As a high in a continuing trend.

This happens often-----before a swing pivot can be set in cement
By completing it's % swing.

Same happens with Kagi,Renko point and figure etc.

 

[CanOz]

Sails
They can't be used as a trading tool.
For the Same reason you can't use fib function in a trading systems test.
The last swing can't be confirmed until it has corrected the % chosen.

Let's take a 10 % swing
You see a pivot and take the trade stop a few ticks above the
Pivot.
It drops 8 % swings back and adds 15 %
the original pivot disappears and just appears
As a high in a continuing trend.

This happens often-----before a swing pivot can be set in cement
By completing it's % swing.

Same happens with Kagi,Renko point and figure etc.

I've been playing around with the 6E (Euro) on market replay, a Zigzag indicator with Fibb. I'll record a video and try to post it tomorrow. Then y'all can plainly see how useful this concept is or isn't...

 

[sails]

Sails
They can't be used as a trading tool.
For the Same reason you can't use fib function in a trading systems test.
The last swing can't be confirmed until it has corrected the % chosen.

Let's take a 10 % swing
You see a pivot and take the trade stop a few ticks above the
Pivot.
It drops 8 % swings back and adds 15 %
the original pivot disappears and just appears
As a high in a continuing trend.

This happens often-----before a swing pivot can be set in cement
By completing it's % swing.

Same happens with Kagi,Renko point and figure etc.

Hi Tech, I agree they can't be tested in a trading systems test. I use them but probably not in the conventional way. They are only a very minor part of analysis and definitely not used in isolation.

I'm not recommending them - I simply provided some links I found with a 10 second google search in answer to a very broad question...

 

[tech/a]

Hi Tech, I agree they can't be tested in a trading systems test. I use them but probably not in the conventional way. They are only a very minor part of analysis and definitely not used in isolation.

I'm not recommending them - I simply provided some links I found with a 10 second google search in answer to a very broad question...

Fair enough.

 

[beachlife]

Sails
They can't be used as a trading tool.
For the Same reason you can't use fib function in a trading systems test.
The last swing can't be confirmed until it has corrected the % chosen.

.

Not the type of swing charts I use. I use Gann's way of drawing them, and they are based on highs and lows, not percentages, can be back tested and can be traded on their own profitably. I use MAV and ATR with them as a filter, but I know some very profitable traders that just use swing charts and ranges. They are a great way to remove the noise from the bar chart.

 

[Trembling Hand]

Not the type of swing charts I use. I use Gann's way of drawing them, and they are based on highs and lows, not percentages, can be back tested and can be traded on their own profitably. I use MAV and ATR with them as a filter, but I know some very profitable traders that just use swing charts and ranges. They are a great way to remove the noise from the bar chart.

Can you give us some examples, either ones that come up in backtests or live discretionary?

 

[tech/a]

Can you give us some examples, either ones that come up in backtests or live discretionary?

+ 1

When is a high a high.
When is it set in concrete.
The day it happens
A week after it happens?

You have EXACTLY the same problem.

 

[Caveroute]

+ 1

When is a high a high.
When is it set in concrete.
The day it happens
A week after it happens?

You have EXACTLY the same problem.

You can set up your analysis so for a high to be a valid high it needs to be the highest high n bars before and n bars after it appeared – so n bars after it appeared and its still the highest high over the past 2n bars, it’s a valid pivot that will never disappear - for the given 'n' level analysis.

 

[tech/a]

You can set up your analysis so for a high to be a valid high it needs to be the highest high n bars before and n bars after it appeared – so n bars after it appeared and its still the highest high over the past 2n bars, it’s a valid pivot that will never disappear - for the given 'n' level analysis.

Yes you can
But many will disappear after n bars.
Trade it----
Code it and test it.
The " n " becomes the old % swing.
Which " n " do you choose to confirm the high and take the trade.
Too early too many stops
Too late trade is over.

 

[Caveroute]

Yes you can
But many will disappear after n bars.
Trade it----
Code it and test it.
The " n " becomes the old % swing.
Which " n " do you choose to confirm the high and take the trade.
Too early too many stops
Too late trade is over.

None will disappear after n bars, the rear view n bars after a valid peak has been identified never changes.

I have coded it, and tested it - and yes I learnt the hard way until I adopted this approach.

n is an arbitrary number of your choosing. If you make it 1 you can build a pivot map of the entire chart which opens up a raft of pattern exploration options.

Not that any of them seem to work particularly well, programmatically anyway, but that's another story.

 

[tech/a]

None will disappear after n bars, the rear view n bars after a valid peak has been identified never changes.

I have coded it, and tested it - and yes I learnt the hard way until I adopted this approach.

n is an arbitrary number of your choosing. If you make it 1 you can build a pivot map of the entire chart which opens up a raft of pattern exploration options.

Not that any of them seem to work particularly well, programmatically anyway, but that's another story.

I understand exactly what your saying.
Played with this about 15 yrs ago.
The bar that is identified a the pivot will remain that within n
But the chart will remain dynamic.
You can alter n's
After n bars and stay pivotal for a different n

Found exactly what you did.
Particularly in isolation.

 

[beachlife]

The simple answer for the swing chart questions is that there is a method for drawing them, so they are programable, hence can be included in software and therefore can be back tested and used in scans, and once a bottom or top is drawn it is set in stone and it's range can be used as a projection or retracement tool in what ever time frame you want.

It's a bit like comparing a trend line to a moving average. A trend line is subjective. Different people will draw different lines on the same chart. A MA is a formula and once drawn does not change and is the same no matter who presses the MA button. The swing chart is a formula or more correctly a set of if then statements. It is what it is and once drawn does not change.

Anyway back to topic, xjo & spi hit the 38% mark today, so now I look for two things, a move up where I wait for a swing top confirmation below 23% to take another trade down, or a move down where 50% is the next possible target. For now tighten stops on the shorts, and patiently wait for confirmation.

 

[beachlife]

Here is how Fib help save me from a 12% of account loss on QBE.

On 30th Oct my scans flagged QBE as a possible long. First step was to draw big picture Fib retracements.


The lines are to show the ranges - not trend lines. I saw that 50% of the small range coincided with 23% of the big range. Next was to draw the entry stop and 2:1 target to see how it fitted.



The target was well beyond the 50% level so I passed.

Had I taken the trade my trailing stop would have been below the 38% line which was just below the swing bottoms so I would have been stopped on open this morning for a 12% loss.


A good example of how being cautious about 50% can keep you out of dud trades.

Once it came back below 50% I have been waiting for a break below the swing bottoms, but unlike my CCL trade I didnt put orders in because the MMA was up, so I have been waiting for a confirmed break of the swing bottoms and the 38% level, ie where my stop would have been. Waiting for confirmation means I just missed a 14% profit - that's life.

There will be a lot of very unhappy investors and options writers today. But I am sure there will be a small number of aggressive traders that went short on the 29th once the first lower swing top confirmed at 50%.

 

[tech/a]

Again the example is shown in hind site.
If your so sure this works as perfectly as hind site
Put up 5 trades.AS IT HAPPENS WITH COMMENTARY.

I'll put up the usual $500 to Joe if you get 3 out of five
To play out as your forward commentary.
Be specific.
2 or less I would hope you'd support your local ASF.

 

[Caveroute]

Tech is right, yes the are programmable but retrospectively.

Look at the rightmost bar as the action unfolds in the example provided earlier [at the bottom of the page] , or better still the final bar - the final top has been anticipated and cannot be confirmed until the following the day is over, and even then maybe not.

http://www.investopedia.com/articles.../04/080404.asp

Therefore you cannot make trading decisions based upon it and you cannot back test it.

How can you backtest something when your anticipated buy signal disappears on the completion of the next bar.

Try using the market replay option in tools like AB to actually see this happen.

 

[beachlife]

Again the example is shown in hind site.
If your so sure this works as perfectly as hind site
Put up 5 trades.AS IT HAPPENS WITH COMMENTARY.

I'll put up the usual $500 to Joe if you get 3 out of five
To play out as your forward commentary.
Be specific.
2 or less I would hope you'd support your local ASF.

Of course it was an explanation after the event, sadly I cant show broker statements for trades I dont take.

If you think that this doesnt work thats ok, no one is forcing you to change the way you trade. So just leave the thread for those that do believe in it to share ideas, and for those that want to learn. If you want to argue, start a thread on politics or religion.

If you want to follow my trades the ccl was done in real time and is still live. The swing chart has swung down today so stops can be moved to break even. The swing is set in stone even though the day is not over. My spi trade is still live with stops at profit and has been explained above. That's 2.

Heres' my wbc, and amp, both live, both in profit, but of course they are hindsight now so no point talking about the entry and management logic.




Here's my oats, all over now, so yes hindsight, wont talk about that one either.



Here's my soy beans, live and just starting to move, but again entered without your scrutiny, so hindsight.



And here's one that failed aio. But again hindsight so not worth explaining what happened.



The next one in my sights is oil, I have provided enough for you to figure out how to trade it.

 

[peter2]

Beachlife: Your application seems inconsistent to me. The swing should define the trend and the direction of your trade. In the first chart you marked the swing up and noticed the 38% - 50% pullback for a possible buy. You mentioned that you didn't take the trade due to the small R:R based on the swing down (pullback). Which swing defined the trade direction? If you become concerned by the 50% level of every small counter trend swing then you'll never start a trade.

My suggestion for a more consistent approach: Trade the daily swings only when they match the weekly swing. As the weekly swing of QBE was down it would be logical to avoid daily swing buy setup and wait for a daily swing down sell setup.

IMO scrap the fibs. The only retracement level worth looking at is the 50% level and it isn't a fib level.

ps: I did buy the QBE near $15.00 as the ledge (after the swing down) was sitting on old resistance, which was at the 50% R level. Yes I lost 18% of the value of my QBE holdings but this was only 2% of my account.

PPs: Noticed that the QBE weekly swing direction changes depending on how you construct the swings (2-bar or 3 bar Gann swing).

 

[Trembling Hand]

Of course it was an explanation after the event, sadly I cant show broker statements for trades I dont take.

If you think that this doesnt work thats ok, no one is forcing you to change the way you trade. So just leave the thread for those that do believe in it to share ideas, and for those that want to learn. If you want to argue, start a thread on politics or religion.

Geesus dude! You were the one that claimed something could be back tested and traded. Tech And I couldn't see how that could be done, or actually what you were trying to get at so have asked a few times for more explanation. Each time you have got more defensive and more evasive and no doubt at the point now of wigging out. While still not showing how swing points can be back tested. Yet we are just asking for what you are trying to get at out side of a few selective post that are the "perfect" examples but have nothing to do with what we asking.

Here is a tip for nothing. Its a trading forum for people to discuss trading. You want try answering the questions? How do you back test, as you have claimed, something that is only set in hind site?

 

[beachlife]

Peter, my approach is consistant, but is probably different to what you do, so it may seem inconsistant to you. I approach very setup the same, its profitable at a level that I am happy with. I agree, I place the most importance on 50% which is strictly not a fib number, but the others are good for short term profit targets.

TH, I did answer your question. Again, swing chart construction is a formula, so you can code it and test it. Or you can manually back test by stepping through a chart one day at a time. And then I gave real examples of how I trade it, including two that are current and live. And of course the old trades that are real, are dismissed as selective. As for sensible discussion, this from you who posts gems like

LOL so now we have 12.5%, 23.6%, 25%, 33%, 37.5%, 38.2%, 50%, 61.8%, 62.5, 66%, 75% and 100%. That really would work every time

and

You guys are all stark raving mad!

Your mind is clearly closed on trading with Fib, you are not here to learn or discuss, just to argue and demand proof, why you think you are so important that we should jump to your demands is a mystery, so just go away.