Topic: TA is pseudoscience - page 3. (Read 5363 times)

Trendlines or regression-fit lines don't predict, they indicate in a simplified way what the current price progression is. If people are using them to predict - that is flawed. Trendlines are useful for staying in a trend when you have a position on, using it as a ruler to see where you're at.

As for chart patterns, some can be useful - especially realizing when you're in an exponential/parabolic move that can break down suddenly after a rise - but others are just hugely subjective and don't produce any actionable trading decisions.

To condemn it all is a mistake. Even statistical evidence suggests the distribution of price data isn't a random walk - it doesn't fit the normal gaussian distribution (bell curve) very well. The mean is usually higher and there are more excursions to the 'tails' that can't be explained. (Leptokurtotic versus the standard "bell" curve.)

If you're familiar with the Hurst exponent, a measure of past changes influencing future changes, you'd know that price series display a persistent reading above 0.5 (randomness), usually in the range of 0.7 to 0.8 - what this all means is that price does have an effect on what prices will be tomorrow. This is why certain techniques can work, and do so reliably.

Of course, of the ones that truly work - you won't see them sold as trading systems are hawked online, because their authors are using them to make money, not to swindle newbie traders.

And I can't tell what you are trying to say but my strategy of holding for the longest time has worked fairly well so far. I entered the market in the winter of 2012.

Basically, they are playing roulette.

this is fair. the buy-and-hold-forever crowd has got a lot going for them. they trade unknown, possibly 100% risk for unknown, possibly many-orders-of-magnitude-large-returns.

i can't tell if im being sarcastic 

Ok. But if you believe that everything below 1000 $/BTC is a bargain then it is more important to enter than to enter at the perfect time.

It's more difficult in any market that is manipulated (by Gox or other forces). IMO Gox does everything to keep it up (cools off the sell-off) and there is no easy way to short, so this certainly makes TA difficult. I believe it still works.

The same patterns present themselves over and over....for a reason. I do not think Bitcoin is magically immune from regular market forces and crowd psychology. And with the extreme lack of solid fundamentals other than overblown press releases, I contend that TA is often the only thing we have.

I agree for normal markets but bitcoin has such a great potential upside that I do not dare not to be a part of it.

The fact that this is even being debated says a lot more about the credibility of you all than it could ever say about the veracity of technical analysis.

I'm not a professional trader but am involved in the capital markets. Many use a blend of fundamental and TA (e.g., fundamentals to choose the asset and technicals to enter the trade). Some choose one or the other.

Algorithmic trading is a $trillion industry. Fully systematic programs account for $billions upon $billions. Discretionary fundamental programs account for $billion upon $billions as well. In some asset classes (e.g., managed futures), fully systematic programs are preferred by institutional investors because it prevents style drift.

I think this is the best analogy for TA...  Weather forecasting.

The difference is weather forecasting is actually science. Weather forecasting isn't always right but there are provable weather patterns and theories and on average it is much better than randomly guessing sunny/cloudy/rain/snow. This has not yet shown to be true for TA in a scientific way.

I definitely agree that it is pseudo science but it can be a bit entertaining

My personal analysis is a fundamental analysis and not TA. I only bought a little during the recent crash/correction. I did not sell anything.

You can read my analysis here:
How to Value a Currency

I might aswell chime in. Obviously I'm a fan of technical analysis. I've never called it a science, I don't think many people would.

Its price forecasting, in a way a bit like weather forecasting. The price chart is a picture of the market, without knowing anything else you can see a lot just from that line. Patterns crop up in charts just like they do all over nature, you can look at these patterns and combine it with your feelings on the overall market trend, your gut instinct, wave your hands oversome tea-leaves and make a forecast.

It is freaky how often the price ends up bouncing more or less exactly where you thought it would though.  

If a lot of people apply astrology in their choice of lottery tickets, some of them will be rich as well. It proves nothing about the viability of their methods.

A list of believers in technical analysis.....and the fortunes amassed due to their use of technical analysis.

James Simons aka the “Quant King”  – $11.7 Billion
Ray Dalio – $12.5 Billion
Steven Cohen – $9.3 Billion
Paul Tudor Jones II – $3.6 Billion

I could continue...

TA is about psychology, more than anything. It's trying to read the emotions of a mass group of investors based on their actions. Which means you'd have to be a psychologist to be good at TA.

No. If everyone knew that, then they would all try to front-run the pattern and buy before the pattern completes, thus never making the pattern appear in the first place.

Markets are funny like that. If something is true, and everyone knows it's true, then it stops being true. If everyone knows the price of bitcoins usually go down on the weekends, then people will sell before the weekend, thus making the weekend drop stop happening.

True. But if you know of a pattern like that, it would be in your best interest not to make it common knowledge. So those who try to spread their patterns are the ones who are most likely wrong.

i already mentioned this in this very thread.

this just means that known regularities are unstable over long periods of time. yet, if there is sufficient informational asymmetry (what if only a single market participant even recognizes the pattern?) he might profitably anticipate it many times before his own correcting behavior dissolves the regularity.


i'm sure this has an effect, but it is certainly not the sole effect. triangles don't form 'because people expect them to'. there are rules that the price follows. if the EMH were true, the price function would look like a random walk, and we all know that it deviates from this model plenty, because we all use these deviations to inform our trades...

again, if you really believe in the EMH, go play satoshidice. you're wasting your time at 50/50 odds in this 'efficient market'

TA is like "board sports". there's a whole bunch of different kinds. from all of the discussions we've had on this point since i started posting my own analysis i've boiled down the "TA-Hypothesis" to an assertion that there do exist patterns, but they are not immediately obvious. as such various price/volume transforms, visual cues such as trendlines, and even fractal analysis such as Elliot Waves could all possibly be used to 'resolve' these less-than-obvious patterns.

so yes, i'm supporting your point haha. i just dislike the label TA because it lumps together a whole number of different analytic techniques which are, at the core of them, surprisingly simple, and makes it sound like some religion people subscribe to.

Look at the stock market as compared to the federal reserve setting interest rates. When the federal reserve lowers rates, the stock market jumps. When the federal reserve raises rates, the stock market drops. The theory is that with lower rates, more money will be pumped into the economy and the big companies will be getting more money.

So when the Fed board releases a rate increase or decrease, immediately people in the stock market react as quickly as possible one way or another.


Now imagine if the people of the stock market consisted of a bunch of people with extra money to play on the market. Kids taking out college loands to invest, families taking money out of their credit cards to invest, people with a few extra bucks here and there. Maybe one or two people who understand the way things always work.

Now, Bernanke comes out and says "we are raising interest rates". Imagine if those people who have all of their money invested in the stock market have no clue what that means and do not really care one way or another. They just got another paycheck that they can invest so they put more money into the stock market and the stock market goes up.

How could this be. Most experts would expect it to go down since that is the pattern. But the people buying do not care about the pattern, they do not follow the charts, they just buy or sell based on their own ideas of what the price should be.

Technical Analysis is a self fulfilling prophecy. Everyone knows that something goes up or down because of that analysis so they trade accordingly, which then re-affirms that analysis.

Bitcoin investors do not care. They see it going up or down based on a multitude of factors and buy or sell on their own whims. That is why TA does not work very well with Bitcoin.

There is a very big problem with TA. Even if you assume that a TA-pattern that offers profitable predictability exists. As soon as people realizes this pattern it would stop existing.

Markets are anti-inductive