Chart Patterns - tosh?

Good points, I wouldn't argue with any of that - as Richard quite rightly suggests (in my opinion) trading is something we do in a system that has a huge number of variables - striking the right balance, allotting a 'good enough to work' (rather than 'correct weighting') balance to all the inputs is more than a case of identifying triangles etc.
Level 2 - I must say that whilst I find it hard enough work interpreting the price and volume info on the chart, I am making some progress... it's not so much a case of not 'getting' it, it's a case of reliably spotting items, I generally have 4 charts on view for intraday and two of them are showing markets whilst I have text fields showing $Trick and $Trin, so I'm attempting to ensure I'm in step with the markets... BUT Level 2 I find very difficult.

Looking for a decent guide to it I bought and read Velez/Capra (it was something like the 80th 'hit' on Global's website using 'Level 2' as a search <g>). They talk about a number of ways to interpret the display, various shenanigans etc from a 2000 viewpoint, other things I've read would suggest some of that 2000 info is now outdated. Is it me, or is Level 2 a bit of a game where the rules change frequently?

Ah me.... I know it's not meant to be easy, but progress is glacial at times!

Dave
 
Marvellous, well put sir ! And in addition, do not overlook the message that Richard imparts here which is the requirement to excercise judgement over a dynamic, unfolding scenario that requires multitasking skiils of the highest order with very quick response.

Some people have difficulty in rotating attention from one concept to another, for example from a chart and what it illustrates, to a series of changing numbers, all of these in different colours in different rythms.

At the other extreme there are people who purport to derive benefit from watching many screens at the same time. There is a point at which the observer has to decide what constitutes information overload and what does not. The refining of this concept takes a long time, and is the consequence of repeated exposure, which leads to experience.

As all of this cannot be hurried, a wrong turning along the road can take a long time to be recognised and indeed even longer to be corrected and adjusted, which is also the opposite to what could normally be expected.
 
Splitlink said:
I think that you underestimate yourself- the toss of a coin, surely, will give a better expectancy than 90:10.

I tend to consider triangles, especially the ones plotted against a nice double bottom like the one in the chart Rossered offers , as more of a confirmer than a buy signal. If I had been watching that one and interested in trading it, I think that I would have entered a few days earlier than the triangle's completion with a stop under the low. TA being a statistical game, don't you think that DB's have the same level of success as triangles? If you did trade the triangle where would you place your stop, below the low? So isn't it better to get in earlier trading the DB?

Split

I wouldn't know where to begin with the charts that rossored posted- No volume, no historical context. To my way of thinking chart patterns like rising triangles, ross hooks, double bottoms, etc. are are symptoms of the buying and selling pressures that caused them. There is nothing to say that pressures that formed them will continue - events happens, players change positions, I can only observe - does the flow continue? Is the same group still in control?

Recently I am trying to break the habit of predicting what will happen. Now when I enter a position - it's because I think I see that something has just changed between the balance of the buyers and sellers. It's a place to enter with a really tight stop. Then I can only watch. If I am correct and the balance has changed - then I stay with that position until I see something changing again. If I'm wrong, I have a reason to get out very near my entry - and not dither around wondering what to do.

All this is very new to me and I do not have any results to report yet. I'm way up this week, but it could just be luck because I have only begun to define and test my criteria for recognizing that the balance has changed.

So your questions about if being earlier is better: I suspect the cause of the pattern happens before the pattern exists or 'confirms'. But I don't have any facts yet.
JO
 
Hi again d998

ducati998 said:
Rogue,
It is a simplistic view. However, just because of its simplicity it would be a mistake to dismiss it without at least some thought.

Some thought.
Simplicity in itself is a very desirable quality, indeed it is often when people over complicate simple tasks that problems arise. However one has to be equally careful not to oversimplify. If one reduces everything in life to it's simplest conclusion, it either works or it doesn't, then probability for any undertaking would be 50/50. Clearly this is not the case, to refer to my previous post, I have in my lifetime crossed the road on many occasions and have yet to be knocked down, by your simplistic view I would appear to fly in the face of probability. Of course with this example we can all see that it is not that simple. There are a lot more influences to take into consideration than simply will it, or won't it.


ducati998 said:
Because technical analysis uses price levels generated by the market to determine entry and exit points, and assuming discipline of trading plan execution, then your expectancy by definition will come in at 50%, as price staying static, will in a timeframe of the traders choice mandate an exit.

Sorry but this makes no sense to me whatsoever (though of course it may to others). Everyone who enters the stock market uses price levels generated by the market to determine entry. A value investor determines those prices generated by the market to be below the true value of the stock and so enters. A momentum player determines those prices generated by the market to be moving with sufficient momentum to jump on board. Your point? This in itself does not lead to any expectancy by definition, much less 50%
From my perspective there is no X% probability to a trade, why? because calculation of probability is a mathematical process which requires the mathematician to assign values to the variables he factors into the equation. With trading there are so many unquantifiable variables that this in my opinion is impossible. A trader must simply assess the variables as he sees them at a given moment and make a judgement whether or not the trade is higher or lower probability, if it is the latter then he need not waste his time looking for confirmation.




ducati998 said:
This is data mined for set-ups that show a historical tendency of doing the wanted and expected. The fallacy is that the better known a set-up becomes, the lower its probability trends, until the final ironic twist and the contra-pattern becomes the set-up.
The problem thus becomes one of guesstimating the level of probability that remains within the set-up that you wish to utilise. For those into backtesting, if you take one pattern and backtest it in enough historical data you will find variable levels of probability.
What you have correctly identified here is why a high % of wouldbe traders fail. Trading a technical pattern in isolation without considering any other influences will probably have widely varying results and these will tend towards the negative, the more the pattern is recognised and traded. Patterns and historical S/R levels are not predictors of what will happen, but simply predictors of what may happen. A chart pattern, a S/R level, all require confirmation. Confirmation comes in the form of price acceleration accompanied by volume. This injection of volume accompanied by this price acceleration attracts more volume and the dynamic of probability on the trade changes. Those who enter a trade early, either in the vain attempt to catch as much of the move as possible, or driven by the fear of missing out, risk the setup failing and fall into your realms of low probability..
As for the dynamic of widely recognised patterns failing, those who wait for confirmation need pay little concern to this. And those who jump the gun soon move over to your camp discounting patterns leaving them back in the domain of those who know how to play them

ducati998 said:
That however does not alter the fact that expectancy for technical traders is 50%
It is only "fact" from your perspective of how you "as a technical trader" would enter manage and exit a trade.
 
JumpOff said:
I wouldn't know where to begin with the charts that rossored posted- No volume, no historical context. To my way of thinking chart patterns like rising triangles, ross hooks, double bottoms, etc. are are symptoms of the buying and selling pressures that caused them. There is nothing to say that pressures that formed them will continue - events happens, players change positions, I can only observe - does the flow continue? Is the same group still in control?

Recently I am trying to break the habit of predicting what will happen. Now when I enter a position - it's because I think I see that something has just changed between the balance of the buyers and sellers. It's a place to enter with a really tight stop. Then I can only watch. If I am correct and the balance has changed - then I stay with that position until I see something changing again. If I'm wrong, I have a reason to get out very near my entry - and not dither around wondering what to do.

All this is very new to me and I do not have any results to report yet. I'm way up this week, but it could just be luck because I have only begun to define and test my criteria for recognizing that the balance has changed.

So your questions about if being earlier is better: I suspect the cause of the pattern happens before the pattern exists or 'confirms'. But I don't have any facts yet.
JO

I don't draw many trend lines or triangles on charts. Everyone has his own way of trading and I am inclined to look for stops being taken out- it happens all the time. I have spent so much time trying to understand volume, without success, that I have decided that I don't need it. Put that down to sour grapes! I've also had enough of intraday and index trading. Now, older and wiser, I trade shares overnight and longer and that seems to be my niche. Footsie, too, because I like sociable hours.

Split
 
Spot on,
requires the mathematician to assign values to the variables he factors into the equation. With trading there are so many unquantifiable variables that this in my opinion is impossible. A trader must simply assess the variables as he sees them at a given moment and make a judgement whether or not the trade is higher or lower probability,

Confirmation is a double edged sword if overdone, another fine balance. If I believe I have seen signs of a reversal that will move the other way (rather than simply going into consolidation - even that is a landmark in education, trends might end, but that doesn't mean a reversal) then I will look for the price to start heading that way, but not for too long... if it hasn't moved after a short while I assume I was wrong. If it does then I enter. This is seldom a matter of more than 2 bars. Confirmation of support might well be a single Doji or long lower shadow appearing in a downtrend where I have previously noted earlier bottoms. On a 15 minute chart I can see strong intimations of the current bar being a Doji or long shadow perhaps as early as 5 minutes into the bar itself - I'm not saying I'll assume it will end up as the Doji when the bar is complete, but it's not unusual to have identified S/R levels from longer term charts (I pick them on daily charts as often as not, then trade 15 minute charts) and when I see the price starting to trace out a Doji I'm not overly surprised... at the end of the bar I've got my confirmation, in part at least.

Dave
ps - Thanks Albert
 
Socrates

Reply:~ No, your assumption is not correct because you approach the topic with the wrong perception. Charts are not inefficient. Charts display the facts, and only the facts. But in these facts is contained information. The ability to extract the correct conclusion from this information presented is what determines the outcome.

If you wish to argue the semantics that is fine by me, but please at least be accurate. My observation was that OTHERS OPINIONS indicated that charts MAY be an INEFFICIENT TOOL. This was not my assumption.
However, as this topic has been broached, my assertion is that chart reading is not a science, and that its practice cannot be continuously successful.
That chart reading is not a science is clearly demonstrable. If it were a science, its conclusions would be as a rule dependable.

The theoretical basis of chart reading runs somewhat as follows;
The action of the market reflects the activities, and the attitude of those interested in it.
Therefore, by studying the record of market action, we can tell what is going to happen next.

The premise may well be true, but the conclusion does not necessarily follow. You may learn a great deal about the technical position of the stock by studying its chart, and yet you may not learn enough to permit you to operate profitably in the issue. Missing from the chart is vital information and that is of "INTENT"


Reply:~ No, the chart does not have any shortcomings if the information presented in it is correctly structured and clearly displayed. It is the human element that mucks up the excercise. The perceptual bias is in the eye and mind of the observer, and not the other way round.

Which is exactly the point made. You need to read the text .

Reply:~ No, you are mistaken. I have in the past done this in a live market with accurate time stamps to punctuate prediction of events before they occurred, ahead of time. I took the precaution of not using my own charts so as not to be accused of doctoring the record.

People do not like to see displayed what they themselves cannot do for whatever reason.

This attracts unjustified and unwarranted abuse. Therefore I now restrict analyses on chart action to commentary, not otherwise, on the wise advice of Chartman the Moderator. But it will dishearten you to hear, that when an absolute bottom is picked twice, for example in sequence in real time with a time stamp to prove it, in the same instrument in the same session, that again, to my wry amusement and quiet contemplation, also stimulates people to become abusive, which causes us those who can, not to

I was actually referring to this particular thread. You however would seem to be indicating that somewhere on these boards lies your real time analysis. I am perfectly willing to accept that it exists.

As to it disheartening myself, why should that be the case? I am perfectly willing to accept that there will always be exceptions to the rule, however the general rule still holds true, and that is , the vast majority of short-term technical traders find it difficult to be profitable and create wealth in the long term. This should be the purpose of a forum, to stimulate thinking and possible solutions to that end.


JUMP OFF,

I can personally attest to the fact that up to now, my expectancy is closer to 90:10. That is - I enter at the wrong time about 90 percent of the time. Your statement does not take into account the fact that the price goes up, down, (sometimes both in a matter of moments) and can also just poke along sideways for a couple months. And it has no reason to trend sedately in one direction - just because we have decided to open a position and you think there is a 50/50 chance it will go one way or the other .

Expectancy is in relation to will THIS TRADE MAKE ME MONEY.
The answer (expectancy) is about 50%, viz, you will either make money and exit, or you will lose money and exit. Your expectancy, therefore is 50%.

Probability, is related to expectancy, but different in the following way. Probability is a mathematical construct designed to identify the likelihood of a specific action coming to pass, while also assigning a weighting to the conditions at the time.
To use ROGUETRADERS example, the probability of my crossing the road successfully when there is no traffic in sight, should be high. If the conditions change to a very busy 6lane motorway, and I am blindfolded, then my probabilities are lower, but my EXPECTANCY remains at 50%, viz, I shall succeed, or fail. If the choice is to abstain, then this nullifies the operation.

Dave,



The reliability of patterns does change over time, and with market conditions, that should not occur if the chance of a correct call is indeed 50/50 except within the normal run of variation (a run of consecutive heads or tails within a chosen number of coin tosses, as an example). There are multi year periods where the reliability of some patterns will hit 80% during bull markets, swap to a bear market and the same pattern will show 50% or less - that doesn't make the pattern or TA a waste of time, it means the sensible TA user is aware that in certain conditions this is not the pattern to trade... although 50% is enough to extract profits with suitable trading.

Again, without assigning absolute numbers to the patterns, this is what I have said. Patterns will change for all sorts of reasons, and the pattern that you have assidiously discovered via backtesting historical data as a 70% probability chooses just this moment to drop into the 40% probability basket. How much money will that cost you? Possibly more than you think, as, your money management on a 70% probability may mandate a position size of 5000 shares, but only a position size of 800 shares on a 40% probability.

Perhaps I should point out, as Socrates is posting here, that 'pattern' is no necessarily meant to imply 'triangles, catapults' and the like - I mean something more like 'pattern of behaviour' - if I see a bar make a new high on substantial volume, then the next bar is struggling to approach the high and volume is notably absent, then that to me is also a 'pattern of behaviour' inasmuch as I have seen it before and regard it as intimation of a change in direction. Dojis, long shadows... the mind, or those who bring these to our attention might provide names for them but it's just a convenient way to ensure the price (and sometimes volume) activity is easier to recognise next time. We - some of us at least - find it easier to 'memorise' price and volume action of significance by giving those price/volume events names

Dave this is interesting, and I for one do not believe that charts tell anywhere near the full story. The point of difference is one of philosophy. As daytraders, you subscribe to "efficient market theory" I do not. Charts only display price traded, not intent. They are therefore lacking very important information. Also, price, for daytraders, and purist technical traders is without context, and thereby devoid of useful information.

Rogue,

Simplicity in itself is a very desirable quality, indeed it is often when people over complicate simple tasks that problems arise. However one has to be equally careful not to oversimplify. If one reduces everything in life to it's simplest conclusion, it either works or it doesn't, then probability for any undertaking would be 50/50. Clearly this is not the case, to refer to my previous post, I have in my lifetime crossed the road on many occasions and have yet to be knocked down, by your simplistic view I would appear to fly in the face of probability. Of course with this example we can all see that it is not that simple. There are a lot more influences to take into consideration than simply will it, or won't it.

I have clarified with Jumpoff

Sorry but this makes no sense to me whatsoever (though of course it may to others). Everyone who enters the stock market uses price levels generated by the market to determine entry. A value investor determines those prices generated by the market to be below the true value of the stock and so enters. A momentum player determines those prices generated by the market to be moving with sufficient momentum to jump on board. Your point? This in itself does not lead to any expectancy by definition, much less 50%

Yes, apologies, I was not terribly clear. Price to technical traders has no real context, ie it denotes no value, or lack of it as it does to a Fundamentals based trader.
Momentum however is momentum, and as such has no relevance to price. Price simply becomes a starting point and finishing point. The two concepts are worlds apart. Expectancy, still remains at 50% for any technically based model.....how can it otherwise?

It is only "fact" from your perspective of how you "as a technical trader" would enter manage and exit a trade

I would no more trade "technically" than fly to the moon.
 
D998 -
I think we disagree fairly fundamentally on the value of TA, which is fine.... it would be a dull world indeed if all were to operate in agreement all the time. As a physicist (however) I feel compelled to argue against your statement:-
That chart reading is not a science is clearly demonstrable. If it were a science, its conclusions would be as a rule dependable.
Your conclusions about science are naive, in reality science is red in tooth and claw, with brutal fighting determining the 'truth' passed on to the rest of humanity. SOME science is demonstrably 'right', but many topics in science are very much open to debate. TA is very much like Physics to my mind, the 'language' of both is mathematics, many patterns and tools rely on finding mathematical relationships between elements, and some areas are hotly contended by their adherents on the one hand and a mob armed with flaming torches on the other. No doubt the similarities between the two accounts for why the UK Institute of Physics has a sub group devoted to the physics of investing.

Dave
 
Ah ! Yes, now I understand your posture. You are correct in saying that percieved inefficiency is a matter of others opinions. Also you say something very interesting here, and it is that chart reading cannot be, in deference to Dave JB above, in laymans' terms, a precise science. This is also correct, because if it were the case, as a precise science again in laymans' terms all of whose elements could be quantified accurately, then all the indicators and signals created would work reliably, without fail. The tragedy is that very many people can be persuaded that indicators can be made to work, in this way, which is nonsense.

In fact chart reading is so far removed from the conventional concept of a scientific approach that it travels to the other extreme in which an appropriate description has to be found and the closest I can think of is that it is an art form in its own right, but the filter that precedes it has to be logical deduction and reasoning......and intuitive guesswork. <TIC>. No doubt this extra twist is what makes it perversly difficult for some people, but not difficult for others, rather like cryptic crosswords.....and this baffles people even more.

The ultimate risk is that for an unskilled reader there is a strong propensity to look at a chart and decide he sees what he wants to see in it, and not what there is.
 
Last edited:
Technical analysis is at the fuzzy permeable border between science and art, so is as much definitive and objective as interpretive and subjective.
Richard
 
There you have it, I could not have said it better myself, thank you Richard, perfect.
 
ducati998 said:
Expectancy is in relation to will THIS TRADE MAKE ME MONEY.
The answer (expectancy) is about 50%, viz, you will either make money and exit, or you will lose money and exit. Your expectancy, therefore is 50%.

Probability, is related to expectancy, but different in the following way. Probability is a mathematical construct designed to identify the likelihood of a specific action coming to pass, while also assigning a weighting to the conditions at the time.
To use ROGUETRADERS example, the probability of my crossing the road successfully when there is no traffic in sight, should be high. If the conditions change to a very busy 6lane motorway, and I am blindfolded, then my probabilities are lower, but my EXPECTANCY remains at 50%, viz, I shall succeed, or fail. If the choice is to abstain, then this nullifies the operation.
An unusual perspective, since two entirely different scenarios leave you with exactly the same expectancy. For me expectancy falls into one of two catagories, either it is perception or opinion based, as in the road is empty and I can see for miles therefore I expect I will cross safely; or I am crossing a six lane motorway, blindfold, therefore I do not expect to get to the other side.
In neither situation would my expectancy be 50/50. In truth again I could not put an X% value on it since I have no mathmatical basis for which to make the calculation, I simply have a high or low expectancy in these outcomes.
The other form of expectancy that I recognise, is a mathmatical based expectancy, the kind that an actuary would carry out in calculating life expectancy on insurance related matters. However I doubt that this is the expectacy to which you refer.

Since we are not looking at a specific trade or any statistical data related to one then we are not referring to the latter. However since you are very specific in your 50% expectancy we cannot be referring to a perceptional opinion based expectancy, since that expectancy would rise with the perceived probability of the trade.
 
Okay.... my tongue WAS wedged inside my cheek when I posted, I'm not really going to shop you all to the Institute of Physics Fuzzy Logic Goon Squad <g>
Fascinating as I'm sure it would be I will avoid writing at length on how much of Science is in reality far from the words carved on stone that many believe to be the case.... in radioactive decay we know that an unstable element will produce a decay event precisely, but we haven't a clue which atom will do it... just that one of them will. It's not wildly different with TA - given a set of circumstances we have a rough estimate of where we stand as the latest bar completes, only rough because the overal market, interplay of sectors, current economic climate and hundreds of other factors ensure no one moment will ever match another. From that estimate of where we stand, and a backlog of experience from which we try to identify similar moments in the past, we predict where we expect to be standing in the future.

Contrary to TA not working, I'd say it's remarkable that it works so well - if climate modelling were anywhere near as accurate we'd be able to trust the weather forecast.

Dave
 
Rogue,

Since we are not looking at a specific trade or any statistical data related to one then we are not referring to the latter. However since you are very specific in your 50% expectancy we cannot be referring to a perceptional opinion based expectancy, since that expectancy would rise with the perceived probability of the trade.

Expectancy is for the methodology of technical analysis as a broad generalisation. I accept that there will always be talented individuals for whom this generalised 50% will not hold true, as their "probabilities' are so much higher than average that it would seem to make a lie of "expectancy". That it does not, however has no practical significance, just theoretical interest.
The more important point however is for the average trader, still perfecting his craft, the fact is that on each individual trade he places the outcome will always be either a loss, or a profit, (excluding the small # of breakevens). To add clarity to the difference between expectancy and probability an example would be in order.
In bankruptcy investing the expectancy is higher than in technical analysis because the outcome is determined in large part by the laws pertaining to bankruptcy.
The probabilities will vary and be enhanced by astute analysis.

Therefore,(for the short-term trader) assuming this outcome ratio is acceptable, he must now look to identify high probability setups, which is where there is ample discussion.
Of greater interest and relevance is the consistency and durability of the higher probabilities. This is so, as if the higher probabilities are both consistent and durable, the position size can be increased with safety of principal ensured. This will increase profitability dramatically.

The foregoing of course relates to risk. Reward calculations are the flip side to the previous.
Viz, how reliable (probability) are your profit projections (calculations).

Cheers d998
 
Last edited:
ducati998 said:
The theoretical basis of chart reading runs somewhat as follows;
The action of the market reflects the activities, and the attitude of those interested in it.
Therefore, by studying the record of market action, we can tell what is going to happen next.
I believe you and I have a differing view about the theoretical basis of chart reading. I look forward to more of your posts.
JO
 
This thread has taken on a highly philosophical turn in its latter stages. I'm not sure whether that was what more practical traders are looking for, although it is good to take on board the opinions of some that we have no more than a 50:50 chance of making money I believe that that there are better odds for traders using TA than that. Theories on man made lines drawn across charts are always going to be dangerous when zig-zagging bars are involved and anyone trying to cross Finchley Road in a straight line, with no dodging involved, has more than a 50:50 chance of getting slaughtered. However, it is possible to cross that road by looking for gaps in the traffic. Certainly, I would advise against taking fixed views against triangles and the like. The libraries are full of books on trading theory and practice and the trading world is full of burnt flesh by those who have read those books.

Suffice to say that, by using the brains that God gave us, it is possible to work out a better than 50:50 chance of coming out alive.

Split
 
Split,

although it is good to take on board the opinions of some that we have no more than a 50:50 chance of making money I believe that that there are better odds for traders using TA than that. Theories on man made lines drawn across charts are always going to be dangerous when zig-zagging bars are involved and anyone trying to cross Finchley Road in a straight

I am obviously not being as clear as I should. Your "ODDS" or more correctly your probabilities of making money are not definitively 50%, just that with a random entry your win/loss ratio will be approximately 50%

Now, by utilising a methodology, viz chart patterns, or technical indicators, level2, you may find that the methodology improves the expectancy from 50%. Nothing new here. What should be considered however is the consistency and durability of your methodologies probability.

If you were to compare a triangle breakout with a Level2 buy signal and you had 100 examples of each for the last 10yrs, or 1000 examples in all through a variety of market conditions, which one gives you the most consistent and durable results?

Why is it consistent and durable? Is it unknown, limited access, would the results change if the reasons for the results changed? If and when you know the answers to these and other questions, then you need to ask, how often can I find tradeable examples, will the frequency of opportunity mandate modifications in my position size and stoploss levels to make it profitable?

Within the daytrading and short-term trading horizon traders are competing with each other, and the larger $value traders can dominate due to size in short time frames, pushing price in their direction certainly enough for scalping purposes. So you may be right, and still lose money......a galling outcome.

Therefore money management serves an additional purpose, which is to safeguard your position, not just your capital, when the probabilities are being distorted by scalping moves designed to break patterns without the commensurate size to safeguard them.
To ride out adverse moves requires some true understanding of your methodology, without which you will become a statistic that is often quoted regarding daytraders.

Cheers d998
 
Of course chart patterns aren't tosh - whether the price moves up, moves down or stays still there's always some aspect of chart TA that says it was going to do just that :LOL:

If I start to cross the road my expectancy that I'll reach the other side is 100% (I'm not bothered if that's mathematically correct or not) . The probability of achieving the crossing depends on my timing. If I walk out in front of a ten ton truck (tested resistance) there's a very low probability that I'll make it.

good trading

jon
 
Top