>I use TA
If technical analysis worked, wouldn't the working strategy get jumped on by everyone and then stop working as it gets priced into the market almost instantly? And did you know that automated bots based on TA fail en masse EVEN WHEN BACKTESTED?
Other urls found in this thread:
jspauld.com
twitter.com
>Ib tegnigal analyzss workkkDd wudnt duh workknng stratgeehh gtt jamppid on by evrryaan den stopped werrrking
>TA fag is incoherent
What a surprise
I just thought it would be the case if the shitty free TA indicators posted here all the time worked. Is it wrong?
Sort of.
For example, this guy: jspauld.com
made half a million with a machine learning approach and it worked for two years. It takes time for the market to adapt. And it isn't as if one or two guys adopt the same approach it stops working -- there has to be a large shift in the way the majority of the market trades to counteract this and correct the inefficiency.
Interesting. I knew all these TAfags were bullshitting.
it works if you have 140+ iq
t. 300x since july
yes it's all quite obvious. no one ever got rich from trading (hedgefund managers have/do, but it's mostly in fees, and their luck eventually runs out)
There are several ways to interpret data in TA. Does it matter why it works? Wether you think it's a self fulfilling prophecy or not, just be aware of it is all. Clearly a lot of people pay attention to the same patterns, price ranges, fibonaci and all those memes
>And did you know that automated bots based on TA fail en masse EVEN WHEN BACKTESTED?
There are very very very few who used TA to make a shitload of money. But not a single one of them just used TA - they did it with a shitton of other means and not a single one of them ignored fundamental analysis.
>A bunch of free indicators on literally every trading site ever will make me rich
DELUDED
stop posting these traps brah
Not a trap trust me.
pfew i thought i was turning gay or something
>implying there some super secret indicator that make everyone rich
T h e DOTCOM bubbles are call you
its very easy op, you make money when there is an exploitation premium to be had. With TA, that premium is either not there or its been arbitraged away a long time ago. In general, anything available publicly will not work, or will not work for long. Thats the whole nature of any exchange, its self correcting. to various degrees. Currently, based on my own personal experience, good money is in infrastructure via arbitrage (those opportunities are still plenty), datasets and data engineering (for example, on sub 1m timeframes limit book pressure is a very good predictor), pure computing power and data analysis (throw 100mil of uncorrelated sets at an algo and you are bound to find something), novel algo trading approaches (personally my favourite, things like svd, signals decomposition and so on), machine learning.
That last one doesn't work by default because every kid has access to keras now. What works is understanding how and why machine learning could actually be applied. Most people, even most hedge funds end up losing money with ML because they make several assumptions that are established in other ML heavy fields (visual recognition, nlp, and so on) that simply don't apply to finance. Pretty crucial mistakes too of very insidious nature. For example sampling. Sampling prices on a any kind of timescale be it 15min or 1min or whatever and then building features from those only introduces heteroskedastic behaviour into the algo. It also breaks the IID assumption that is required for most ML tasks. Labeling observations is done all wrong in 99% of literature and papers out there. Backtesting and cross validation is also all wrong in most papers and results can be safely dismissed most of the time. I could literally write a book about this. Lets just say ML without a ton of adjustments will cost you a lot of money.
TL;DR if its easily accomplished chances are the premium is already long gone
When I read stuff like this it reminds me how powerful a single human brain is when it sorted through all the noise to buy eth at $0.50 two years ago
After all it has an unbroken chain of organic development going back approximately 4.4 billion years ago
well luck is undebatable of course
Nope.
Recognition of opportunity, evaluation of risk, these are common to human survival
Everyone with a brain inherited genetics commensurate with some level of ability
> Recognition of opportunity
> evaluation of risk
people are notoriously bad at this. Especially evaluation of risk. Whole industries exist to make money off of that. Poor evaluation of risks and various biases established by behavioral science is what broke the EMH assumption in the 80s. But yes, you can certainly luck out and purchase something pre-hype and then it moons. Anyhow, I was talking about trading and op was asking about trading I assume. If you wanna do value investing then yeah its possible to make good bank with little to no math, look at Buffett for example. But thats of little interest.
I'll put it this way, luck helped some genetics survive for short periods of time.
Ability enabled genetics to survive for millenia
Natural selection overcomes short-term luck for long-term viability. In other words, its not luck that you possess genetics that help evaluate risk/reward and opportunity.
I bought btc at 1k without knowing what TA was, what an ICO was. I never heard of Ethereum. I had never looked at a chart. I was using different tools. TA is just another tool like fire or a stick.
Be pragmatic. If a tool isn't working/doesn't produce results, throw it away and make a new one. For me, I used other signs to predict future events. Stories, aka the "news". Within those stories, traces, like animal tracks in the snow.
Big scores are like game herds on frozen arctic tundra. TA is like footprints telling you where they came from, where they might be going. "indicators" like a piece of fur, a pile of dung, tell you a story. Machine learning is the translation of data into a story a computer understands. Intuitively, you might realize you know the answer, and thoughts like that "gut feeling" shouldn't be discounted so quickly.
> anything available publicly will not work, or will not work for long
Simply put : normies ruin everything
Sold 80% of my stack when niggers started tweeting tron and crypto raps.
Feelsgoodman
so in other words, you got lucky? cool
hope you used that money for some good
Ughhh I'm depressed now. Wanted to learn trading Fx mostly buy reading your essay makes me wonder.
Can one survive as a ''basic'' trader?
>people are notoriously bad at this. Especially evaluation of risk.
Its not that they are bad at it. People weigh bad consequences more intensely, on average. It is instinct, but also a consequence of learning. You learn that avoiding bad outcomes is probably safer than taking a risk. So, you find reasons to decline an opportunity or mitigate to prevent an anticipated loss or harm.
Instincts can be overcome, for some, harder than others, depending on experience and genetics. There are two paths you can take. One, overcome one set of genetics with another. Replace fear with control by experiencing and testing risk. Two, accept bias toward safety by changing a strategy.
Instead of hunting the mastadon on foot with spears, figure out where it is going to be at some point and roll a boulder off a cliff onto its head with little risk to oneself.
In the end, it could be the better strategy to be safe. Humans survived by having some bias toward risk. Take advantage in trading by making safer trades, for example, not riskier ones.
> its not that they are bad at it.
> People weigh bad consequences more intensely, on average
so we agree that people are bad at evaluating risks?
FX is dominated by bots at this point. But look man, I have some buddies killing that shit with only minor algorithmic assistance (trading still done manually). Experience is important too.
>so in other words, you got lucky?
No. I took a risk when I saw an opportunity. I had the means too. How? I had saved a large sum of money for the express purpose of investing when the opportunity presented itself. Was it luck I saw the opportunity? I read every single article of business section every single day.
My struggle was, like everybody else, the risk part. Being inexperienced investing my own money, my struggle was strategy and emotions.
Luck had nothing to do with it. I really think luck is a marginal phenomenon.
Its something that people use to characterize a story about the past. Since effects have causes, luck is a shortcut to explaining something properly. Luck is never the right answer. Things have discrete causes.
People say, "he got cancer, it was bad luck". But in fact it arose from some discrete factor that you happen to be ignorant of.
What your talking about is called the efficient market hypothesis, and yes TA doesn't work if it holds.
see those are the biases I'm referring to above. And thats why EMH was fucking laughable at best, even pre Kahneman et al. Its not that I'm assuming luck, although when evaluating stochastic phenomena thats a legit assumption to make, its that you are retroactively justifying your decisions.
>so we agree that people are bad at evaluating risks?
I am someone who has been cautious in life, and taken a modest amount of risk. I play with risk in various domains, for fun, like motorcycling. I also study genetics.
From my experience and observations I wouldn't say people are bad at evaluating risks. For example, people seem to be good at avoiding possible bad outcomes. There is a bias toward dangerous situations. Its not necessarily maladaptive. I don't think most people play the market because of this reason. They see alot of downside.
I think people are limited in what they can know and that uncertainty biases them. The other problem is believing they know something when they don't. The market behaves irrationally because it is comprised of people constantly re-evaluating risk and reward, with a individualistic, conservative bias.
To be honest, I am not that much different than anyone else as far as evaluating risk. I still wouldn't say humans are bad at it, just biased. They can overcome that through learning and conscious processes, and it has to be continually and consistently tested. The minute you lose your risk discipline on a motorcycle you can get killed, or lose your trading discipline in the market, you can get rekt. Self control is key.
thats not what EMH states. EMH simply refers to prices alone not carrying any premium information. What op refers to is simple strategy arbitrage effect of self correcting markets. Fama proposed perfect EMH at all times, did it so well that 99% of academia fell for it for over a decade until behavioral science gradually poked a million holes into that assumption. Its still widely subscribed to (the imperfect version), hence why anything divergent from that assumption is called "anomaly" (like momentum as a predictor, which should not be possible under an EMH regime). Its mostly bunk though regardless, prices do carry information, TA is just the wrong tool for that.
Yes, the stochastic nature of the price is like the ultimate challenge to a brain calibrated to the real world. Its a continuous effect without a clear cause that confounds prediction and certainty.