Chart Patterns

[beamstas]

Hate to disillusion you, Tech, but there is NO statistical significance in the patterns you are showing.

I'd suggest that you are putting the case that the tails are suggesting that lower prices are being "rejected" - i.e. that buyers are willing to soak up all that the sellers have to offer and then some, but if you actually code up these patterns and test them - there is no statistical significance to be found there.

50% go up, 50% go down and there's no particular correlation with fat tail outsized returns.

Been there, tested that.

Of course, by no means does the above preclude trading such patterns profitably, but it ain't the pattern that makes the profit.

Michael. Im going to comment on this, you most likely know this anyway but it might add something to the thread

When you see those patterns, you said it youself, 50% go up, 50% go down. So what do we learn? Don't buy yet.

You wait until the odds are in your favour. To do this you only buy the 50% of trades that go up.

How do you do this?

Wait until the price breaks out of the consolodation. That's a sure way to get the 50% of the prices going up.

Nothing is 100%. The only way you are going to know what the price is going to do tomorrow is to go and take a poll of every person in the world who is going to place a trade tomorrow. This is not practical so we do the best we can. Instead of polling tomorrows traders we poll todays traders using resistance/support, having a look at volume and trying to come to some sort of conclusion about what is happening to the price of a stock and why

Like i said earlier, all you can do is attempt to stack the odds in your favour.

It is in my opinion that this can be done by using technical analysis.

If the market were purely random then why are some traders better than others? How can they seem to make so much profit while you are still putt putting away? Because they understand why the price does what it does, NOT how to overlay and read an indicator off a chart

Cheers
brad

 

[johnnyg]

Michael.

Sorry to dis illusion you with the illusion that its the tails I'm interested in.
I stated its not conventional and its not.

The pattern tells a story,which is a player in the setup NOT the entry. The entry is triggered by "The Unconventional"
Which is way off beam with conventional thinking.

Thats as far as I go with this as its MINE MINE!!!
Guess away.

Something to do with low volume squat bars?

 

[weird]

Hate to disillusion you, Tech, but there is NO statistical significance in the patterns you are showing.

I'd suggest that you are putting the case that the tails are suggesting that lower prices are being "rejected" - i.e. that buyers are willing to soak up all that the sellers have to offer and then some, but if you actually code up these patterns and test them - there is no statistical significance to be found there.

50% go up, 50% go down and there's no particular correlation with fat tail outsized returns.

Been there, tested that.

Of course, by no means does the above preclude trading such patterns profitably, but it ain't the pattern that makes the profit.

Never understood the 50% rule of being right or wrong ... if no exit has been defined ... where is the logic if no exit has been defined (forget brokerage).

Been there, tested that ... I don't think anyone can exhaust all possibilities. We live in our own small worlds, me included.

Perhaps we can skew the win rate to 90% if using wide stops ? Wait, until a reality margin call or until we call it quits.

I have more to add, but this thread is going off track. I believe entry is important, but also exits and money management is just as important as MD has pointed out, but also disagree concerning the importance of entries.

Pattern trading is a valid entry method, and some patterns are perhaps statistically better than others ... even with a 33% win ratio but most importantly includes MM ... having the stamina to trade them is a different matter.

 

[MRC & Co]

Hi MRC,
Like to hear more of your perception , especially during those periods you alluded too ..

For example, relative position of the pattern. Your better off buying a double bottom that hits a previous swing high or level, than a double bottom in the middle of nowhere.

Behaviour of your market, you don't trade a breakout on the SPI (you either trade the false break or wait for the pullback), on certain FOREX pairs or other trending markets, you can hit the top or bottom of the move. You do it on the SPI, you better have wide stops, conviction, and be right (poor entry) or you will be crucified.

Also agree with much of your post Weird, a lot of smart, practical, trading fellas around here, good for the brain!

 

[alwaysLearning]

BINGO.


A chart of whatever you wish to display be that Bar or Candlestick is a symbol of crowd behaviour for that time frame.

It represents that snippet in time.
Patterns, particularly recurring patterns are a group of singular crowd behavioural events which indicate the likely reaction by the crowd to that group of events. Some see this as an edge,whether it is or not will depend on BINGO.

For those of you still bent on patterns and wish to perhaps take advantage of Michaels wisdom these thoughts of mine may help you in your application.
See Chart below.

Last line should read Price action AND volume---within the pattern

Thanks tech that does help Man, lots of good things to read this week on this forum!

Gives me an idea of how might be able to better categorise some patterns i've been looking at.

 

[johenmo]

This thread has a wide range of views for a learner & just reinforces there is no one way of trading.

Humans are creatures of habit. So this has to impact on trading (likely subconsciously for most) & this is where I see patterns occurring & being of use. But only to reduce risk by trying to tip the probability being 'right" of a move one way or another in. Anyone who thinks they predict the future is crazy. Too many influences.

Yet again, the thing that came out was quitting losses quickly (MM).

I look at a pattern type (say a symm Triangle) and sometimes it goes up & sometimes down. Isn't the way it moves a response to the "stimulus" surrounding the stock? (good/bad news, sector performance/perception, etc)?

 

[tech/a]

Michael is a systems trader and I to trade systems although not currently.

What your not thinking about in the 50/50 scenario is there will be times when you have 75% wins and times when you have 25% losses.

If your wins are 2R and your losses 1R or less.
BINGO.

As systems traders and developers we are constantly looking at Reward to Risk.
As a discretionary trader we should also remember the above and be striving for the same.

T/H what were you going to add?

 

[Trembling Hand]

T/H what were you going to add?

I think patterns in charts are used to confirm our biases and enable us to act.

Its a big subject and haven't much time at the moment but I think its perceived that TA is used like this,

1. Find a pattern.
2. deduct from that pattern is bullish
3. Act on that pattern

My thinking is more along the lines of implicit knowledge.

1. Through repeat exposure you develop implicit knowledge of how your fellow traders act.
2a. receive info from price action & "other sources".
2b. become bullish
3. find a pattern that fits such views
4. trade it.

The rally from 3200 is a classic example. I made note to you that wasn't it funny that the guys that actually trade the markets have been long from then. They chose to weigh over sold more than lower lows, lower highs. Why? TA theorist where looking for shorts. TA practitioners were LONG. I believe they consciously or subconsciously decided they have seen this game before, to much negativity, over sold Blah blah, and found the pattern that enabled them to act to get long while ignoring the same valid patterns telling them we are going lower,

Experience, knowledge whatever enabled them to pick the "right" pattern.

 

[MS+Tradesim]

What do people mean by 'patterns'?

I think the majority interpret it in reference to traditional TA ideas like triangles, H&S, wedges, etc.

Yet, a pattern is any event or process that repeats over a given period. For example, it could be a pattern for a stock to close higher than its lows 90% of the time. Or it could be a pattern that a certain index or sector in Australia will do the opposite of last night's DJIA. Or maybe a sell-off is a re-occurring event every time XYZ stock comes close to a round number. Or a new high on low volume in ABC is normally followed by a large retracement....etc.

IMO, it's a basic necessity to objectively determine the edge a particular pattern has. Don't just use it because everyone else does or because widely read TA books tout it. Screening charts with the eyes is not, IMO, a great way to deploy patterns. Our biases will cause us to skip over the false positives and accentuate the times it clearly works, no matter how few. This is one reason why testing (backwards and forwards) is important. It allows you to quantify the edge (or lack thereof) of a particular idea.

Despite the fact that I've posted some traditional TA pattern ideas in those threads, I don't actually trade those patterns per se. But they can be used as guides to what other people are thinking. I'm looking in the chart for confirmation of my own idea and knowing how a lot of people use traditional patterns allows me to find my own entries and exits. There is nuance here. Some people believe a pattern/predicts causes an outcome. This is a common misconception from beginners and FA-only guys. A pattern can't cause/predict anything. It is a summarised representation of crowd behaviour on a chart which is a representation of a market, not the actual market itself. I think the majority of experienced traders understand this and use traditional patterns to see what the herd is thinking.

For me, the valuable patterns are the ones I quantify myself and are not found in books or websites. They are discovered.

 

[MichaelD]

When you see those patterns, you said it youself, 50% go up, 50% go down. So what do we learn? Don't buy yet.

You wait until the odds are in your favour. To do this you only buy the 50% of trades that go up.

How do you do this?

Wait until the price breaks out of the consolodation. That's a sure way to get the 50% of the prices going up.

Nope, that's a sure way of buying something at a HIGHER price that subsequently has a 50% chance of going up or down.

If the market were purely random then why are some traders better than others? How can they seem to make so much profit while you are still putt putting away? Because they understand why the price does what it does, NOT how to overlay and read an indicator off a chart.

A fine question indeed.

Perhaps we can skew the win rate to 90% if using wide stops ? Wait, until a reality margin call or until we call it quits.

That's easy - set a profit target of 1 tick and use a wide stop loss or no stop loss. 90% win rate. Popular with black box spruikers. Unfortunately pretty likely to be fatal to trading capital.

1. Through repeat exposure you develop implicit knowledge of how your fellow traders act.

Despite the fact that I've posted some traditional TA pattern ideas in those threads, I don't actually trade those patterns per se. But they can be used as guides to what other people are thinking. I'm looking in the chart for confirmation of my own idea and knowing how a lot of people use traditional patterns allows me to find my own entries and exits.

And there you have the answer. Charts are random. People are not.

 

[tech/a]

M/S&T

When I first started to type I used to look at every key infact I'd search for it.

It would take me a day to type a page.
Now I do it without thinking and can get pretty fast.

Same with reading a chart.
But you must always return to BINGO

 

[motorway]

markets are random

markets move like coin toss 50/50
from any point in time move up or down

in any time period

Looking at wrong things

Market not random
markets do not move 50/50 ( in terms of size of moves )


The STEP SIZES are different

That is why random first definition
non random under second definition

Don't trade time
trade moves

Time ticks

moves trend


statistics back up

markets trend based on moves
not proportion of up or down days
this is close to 50/50

But who cares about that ?

motorway

 

[Trembling Hand]

Wait until the price breaks out of the consolodation. That's a sure way to get the 50% of the prices going up.

I'm not so sure about that. especially when you take into count that you have to have a stop somewhere and still a positive R:R.

you got any stats?

 

[kam75]

G'day

Chart Patterns..can they work or are they useless??

cheers,

ceasar73.

For some traders yes, for other no. You need to learn how to recognize real patterns from the rest of the crap thats there and learn how to trade them. Its an art. Takes years. And its hard work. You'll eat a lot of **** along the way too. But if you learn to manage your risk, it can be profitable. I've made hundreds of thousands of dollars over the years trading chart patterns.
regards

 

[ceasar73]

For some traders yes, for other no. You need to learn how to recognize real patterns from the rest of the crap thats there and learn how to trade them. Its an art. Takes years. And its hard work. You'll eat a lot of **** along the way too. But if you learn to manage your risk, it can be profitable. I've made hundreds of thousands of dollars over the years trading chart patterns.
regards

Kam thanks.
Can you recomend a book on chart patterns?
cheers,
ceasar73

 

[Mr J]

What do people mean by 'patterns'?

My thoughts as well. Everyone trades some sort of pattern, as we're all looking for entries that produce an expected result (in terms of probabilities, not certainty).

 

[kam75]

Kam thanks.
Can you recomend a book on chart patterns?
cheers,
ceasar73

'How Charts Can Help You in the Stock Market' - William Jiler
'Exploding the Myths' - Frank Watkins

If you like to study, you may get yourself 'Technical Analysis of Stock Trends' - Robert Edwards,John Magee ...the bible of technical analysis.

I have a lot on my blog on patterns too.

 

[ceasar73]

'How Charts Can Help You in the Stock Market' - William Jiler
'Exploding the Myths' - Frank Watkins

If you like to study, you may get yourself 'Technical Analysis of Stock Trends' - Robert Edwards,John Magee ...the bible of technical analysis.

I have a lot on my blog on patterns too.

Kam - checked out your blog...nice one dude!

 

[Wysiwyg]

My thoughts as well. Everyone trades some sort of pattern, as we're all looking for entries that produce an expected result (in terms of probabilities, not certainty).

Worth bumping.

 

[tech/a]

The question is what is it that your looking for in a chart pattern.

What does a chart pattern tell you.
What constitutes a pattern
We all know standard patterns
But what about patterns hardly ever if ever
Talked about.

Can patterns be traded alone!
How long are they valid
When are they considered to have failed.
Are time frames important.

There are all sorts of chart patterns
Point and figure
Steidlmayer ( Market Profile )
Candle stick
Renko
Time and Volume Patterns
And many more

Wysiwyg is right a great topic
One which AI first started.