The Problem with Technical Analysis

[wisefoolx]

The problem with TA is simple, it's in its calculation, which is hardly ever objective and almost always subjective. What do I mean by this ? Very simple really, we as traders are trying to forecast where future prices might be. A future price is a number in the future which is itself based on numbers. Therefore why should we not use the claculation of number to analyse prices ? There really seems no good reason why this should not be the case.

Let's take the old example of drawing trend lines. The phrase itself should make you gringe, tell you that something is untoward, it means exactly what it says using your hand to draw a trendline. Bearing this in mind, we should know that no two people are going to draw trendlines the same. Yes I am well aware of the so called rules of drawing where you join higher lows to even higher lows to get an uptrend and lower highs with lower highs to get a downtrend. But even then, where is the objectivity in this, what math was used to get this result ? None, so just because some guru says it is so then all it means is this is the way he draws lines , it is no more valid than the way how you or I or anybody draws lines.

Solution: Use linear regression , that will produce an objectively drawn straight line through all the data points calculated by maths, available freely on many broker software

NB. This is still very very basic there are even better ways to produce trendlines.

Now let's come to another bugbear of mine RSI. This is supposed to measure changes in the momentum of prices in hope that you can forecast where they will go in the future.

Momentum is a concept well establish in physics and is calculated by p = m * v, where is p = momentum , m = mass and v = velocity.

So why is it calculated by this concoction:

RSI = 100 - (100 / 1+ RS)

Where RS = Average of x days\' up closes / Average of x days\' down closes


This has no foundation in any concept in hard science and has been plucked out of the air or even other places by Welles Wilder. It is thoroughly unobjective and totally subjective. What true value does it have ?

Solution : Use equations based on proven hard science concepts.


NOTE: THIS IS NOT ABOUT IF TA WORKS OR NOT; THIS IS ABOUT HOW MOST TA IS CALCULATED, END OF.

Also nothing to stop people from using both

 

[seekingTruth16]

What if price movements aren't based on any objective laws, but are instead based on subjective judgment of key market participants. Subjective laws. Do your complaints about the need for objectivity still hold?

Anyway, good post. Mostly agree.

 

[EnlightenedJoe]

Solution : Use equations based on proven hard science concepts.

Good idea. But do you dare to place a trade using that ? If not, then it's useless. Things like that can easily be tested through a call out.

 

[wisefoolx]

NO.

I knew this would happen, the article says NOTHING about what works and what doesn't. It does NOT care, that is an individual choice ( what type of TA you use ).

However it demands this : A future price is a number in the future which is determined by numbers in the past. Therefore why should we not use the claculation of numbers to analyse prices ?

 

[wisefoolx]

Good idea. But do you dare to place a trade using that ? If not, then it's useless. Things like that can easily be tested through a call out.

Good idea ? LOL.

What do you know about TA, you're a MM trader , you stick your finger in the air and blow with the wind.

Get out of here, stop chasing me around.

 

[wisefoolx]

What if price movements aren't based on any objective laws, but are instead based on subjective judgment of key market participants. Subjective laws. Do your complaints about the need for objectivity still hold?

Anyway, good post. Mostly agree.

Your answer is in #4

 

[EnlightenedJoe]

Good idea ? LOL.

What do you know about TA, you're a MM trader , you stick your finger in the air and blow with the wind.

Get out of here, stop chasing me around.

I don't need to know about TA. But, I do know people using TA or your pseudo-TA won't dare to call out trades using it.

 

[barjon]

NO.

I knew this would happen, the article says NOTHING about what works and what doesn't. It does NOT care, that is an individual choice ( what type of TA you use ).

However it demands this : A future price is a number in the future which is determined by numbers in the past. Therefore why should we not use the claculation of numbers to analyse prices ?

That would make sense if future price is indeed determined by price in the past. Your argument is based on that assumption which I think is doubtful. You could say equally that future pattern of prices is determined by past patterns - which most ta is based on. Just substituting numbers for patterns makes no difference.

 

[EnlightenedJoe]

That would make sense if future price is indeed determined by price in the past. Your argument is based on that assumption which I think is doubtful. You could say equally that future pattern of prices is determined by past patterns - which most ta is based on. Just substituting numbers for patterns makes no difference.

Your doubt doesn't seem too genuine given your ongoing interest in the tall candles. Isn't that also based on past patterns ?

The failure in the OP's thinking along with the TA, the big candles crowd is they believe they can extract money out of someone's wallet by using a mechanical formula. This cannot be done except for the formula of pointing a gun at the wallet owner's head. Otherwise the wallet owner will do everything in his power to preserve his capital, include violating all the big candle patterns you care to throw at him. Wouldn't you do the same if you were in his shoes ?

The OP went from trading £5-£30 a point in the past to now not daring to call a trade worth no more than a penny just proves my point. There is no value in these numbers and patterns. It's a dead end, a cul-de-sac.

 

[tomorton]

Hi wisefoolx - You've followed a line in your opening post concerning TA that many critics of some TA approaches or even TA as a whole often use.

You've suggested that your approach is better, which is fine and might well be correct for all I know. But in order to demonstrate this you've selected aspects of TA that are inherently weak, the implication being that all TA-based traders equally respect these weak approaches.

If you had only set out to criticise trend-lines and RSI I'd be right behind you. Or if you were showing better types of trend-line TA or a better way to use them and RSI.

But as it is I think you just might be dragged into the usual circular arguments. Are you going to say more about your own approach?

 

[barjon]

Your doubt doesn't seem too genuine given your ongoing interest in the tall candles. Isn't that also based on past patterns?

I don't know how many times I have to repeat that I'm not at all interested in tall candles nor have I ever traded them. It was all dredged up by fl from an old discussion thread I started about their merits or otherwise.

So far as TA is concerned I use it as a tool just to point up places where it's reasonable on past experience to take a trade or to end one. Certainly not as any form of predictor.

 

[EnlightenedJoe]

I don't know how many times I have to repeat that I'm not at all interested in tall candles nor have I ever traded them. It was all dredged up by fl from an old discussion thread I started about their merits or otherwise.

So far as TA is concerned I use it as a tool just to point up places where it's reasonable on past experience to take a trade or to end one. Certainly not as any form of predictor.

My impression wasn't from what fl did. Only a couple of weeks ago, you were talking to someone who offered to test those candles for you. Obviously you are still holding out hope for the big candles. Otherwise I don't see why you would waste someone's time with something you know doesn't work. Maybe you were talking about different candles. Nevertheless it was about patterns that has no basis other than the past.

 

[wallstreetwarrior87]

NO.

I knew this would happen, the article says NOTHING about what works and what doesn't. It does NOT care, that is an individual choice ( what type of TA you use ).

However it demands this : A future price is a number in the future which is determined by numbers in the past. Therefore why should we not use the claculation of numbers to analyse prices ?

Because every market is different. Some markets are used to hedge, and some are more the domain of directionally based decisions.

So TA in "any form" can not be used with any consistency, whether it "works" or not. As the underlying calculations are based on different meanings although its a "physical" number. Its more to do with what the number represents.

Numbers in the future are determined by a lot more than numbers in the past, more to do with time and opportunity, but thats another thread

 

[barjon]

My impression wasn't from what fl did. Only a couple of weeks ago, you were talking to someone who offered to test those candles for you. Obviously you are still holding out hope for the big candles. Otherwise I don't see why you would waste someone's time with something you know doesn't work. Maybe you were talking about different candles. Nevertheless it was about patterns that has no basis other than the past.

We were talking about retracements, not big candles. As usual you seek to twist things to suit your own agenda .

 

[wallstreetwarrior87]

We were talking about retracements, not big candles. As usual you seek to twist things to suit your own agenda .

Very wise, as my next point was going to be - what if we take price off a chart? How will that affect TA? We would then have to move onto the nuts and bolts of the market -

We are left with time and movement (baring a set tick chart of course).

Retracement = movement, not numerical, but physical.

So ultimately, what would the real benefit be to using calculations based on an independent number in the past?

 

[EnlightenedJoe]

We were talking about retracements, not big candles. As usual you seek to twist things to suit your own agenda .

How was I to know ? I wasn't into TA. One candle two candles, they look all the same to me.

Let's put the big candles aside, this retracement candles work then ?

 

[dbphoenix]

Very wise, as my next point was going to be - what if we take price off a chart? How will that affect TA? We would then have to move onto the nuts and bolts of the market -

We are left with time and movement (baring a set tick chart of course).

Retracement = movement, not numerical, but physical.

So ultimately, what would the real benefit be to using calculations based on an independent number in the past?

If you were to remove price from the chart, you'd be getting back to traditional technical analysis, i.e., the analysis of price movement, specifically the trader behavior that moves price. Several decades ago, a "trading for dummies" attitude took over, and indicators became the sine qua non for beginning traders in particular (and those who sold products and courses to beginners). After all, any idiot can buy or sell when the red line crosses the green line. And now that people are at last beginning to realize that it's all nonsense, technical analysis gets a bad rap. But if they were to examine the origins of technical analysis, as distinct from fundamental analysis, they'd find a wealth of useful information.

 

[wallstreetwarrior87]

If you were to remove price from the chart, you'd be getting back to traditional technical analysis, i.e., the analysis of price movement, specifically the trader behavior that moves price. Several decades ago, a "trading for dummies" attitude took over, and indicators became the sine qua non for beginning traders in particular (and those who sold products and courses to beginners). After all, any idiot can buy or sell when the red line crosses the green line. And now that people are at last beginning to realize that it's all nonsense, technical analysis gets a bad rap. But if they were to examine the origins of technical analysis, as distinct from fundamental analysis, they'd find a wealth of useful information.

Exactly!

All of this seems to get missed. But, if we delve further, we get so much useful information as you say. Then we really can analyse the market from a "TA" POV, and it really starts getting interesting:

 

[wisefoolx]

That would make sense if future price is indeed determined by price in the past. Your argument is based on that assumption which I think is doubtful. You could say equally that future pattern of prices is determined by past patterns - which most ta is based on. Just substituting numbers for patterns makes no difference.

Yes, I have rephrased that part, hopefully that will make the point better. Sorry but that was written at 2 in the morning.

 

[wisefoolx]

Hi wisefoolx - You've followed a line in your opening post concerning TA that many critics of some TA approaches or even TA as a whole often use.

You've suggested that your approach is better, which is fine and might well be correct for all I know. But in order to demonstrate this you've selected aspects of TA that are inherently weak, the implication being that all TA-based traders equally respect these weak approaches.

If you had only set out to criticise trend-lines and RSI I'd be right behind you. Or if you were showing better types of trend-line TA or a better way to use them and RSI.

But as it is I think you just might be dragged into the usual circular arguments. Are you going to say more about your own approach?

I don't think I am agreeing with other critics because I am not saying it is bunk which most other critics are..., only stating that the most of it is SUBJECTIVELY calculated.

Indeed if you read correctly, I did add that one could use both tA and maths together.

And I have stated what my approach , please re-read.