#winning

Your conclusion is that TA is no different than gambling---it offers no edge. Your analogy illustrates this. You assume your conclusion.

The fact is that some of the best traders in the world use TA somewhat. You are not one of the best traders in the world (you're not even one of the best traders on twoplustwo). So what you say doesn't matter, no matter how many words you use to say it. This might be an appeal to authority, but at least it's an appeal to an actual authority instead of assumed authority about finance on a goddamn poker forum from someone who isn't actually concerned with what works or not, just concerned about winning internet arguments.




I would seriously use astrology before taking your advice. It's just as likely to be right.

That's the benefit of talking in absolutes. You have a 50% chance of being right and don't need to do any deep study at all to make a claim. There are people that know better than you though.

on one hand, TA almost assuredly works.... but what is TA? and what aspects of it work? i think it's a small part of it.

someone very knowledgable on this forum suggested that AQR's studies of momentum investing are TA. i would argue not, but if they are then i don't think there's any doubt TA works.

another very popular TA type indicator is something like RS(2). don't have definition in front of me but it correlates highly to market down 2 days in a row = BUY, some definition of market up 2 days in row, SELL that position. add some simple volatility filters and stop loss rules and away you go. and arguably it has worked really well.... but is this really TA? and does it really work?

to me, TA is ultimately support/resistance and i do think this has merit, but not significant merit ...... at an extreme, knowing that the world's largest equity mutual fund company is in the process of buying 10% of Tesla = "can't miss" (relative to market, you could short market against it). and that is technical analysis (albeit strong form).... weak form of TA is looking at support/resistance on charts

here's a question,

do people think that a simple concept like support/resistance on a chart has merit?

if TA has no merit, then why the need to even look at a stock chart for a liquid stock? you could even argue why look at a non-liquid stock chart, but i think everyone would agree that has merit.

Does anyone on here seriously enter a trade or build a position simply on a fundemental thesis without regard to price/momentum/relative performance? The definition of TA is being butchered in this thread...

don't firms like DE Shaw (DA Shaw?) and Renaissance rely heavily on what you might call TA. in both cases, Astrophysics PhD's working on TA, but TA nonetheless.

Renaissance is arguably the most successful HF in history with results that i'd say are impossible to understand (and that's in hindsight)

well said

No. This is painful for everyone reading who's not ******ed. Do try and follow. There is no conclusion in my answer to yours. You are reading a conclusion in because your'e a ****ing ******. This is literally what happened.

You make pointless, worthless point:
I use exactly the same logic in a reductio ad absurdum to show you that it's a pointless, worthless post:
This is me saying "your point is ridiculous bro and so are you". I am making absolutely no claims about TA here. I am making no conclusion other than the utter stupidity of your post.

This goes right over your head and hilarity ensues. Even after I explain to you exactly what is going on, you double down on the stupid, as if you lack the ability to step back. Amazing stuff.
What percentage? Have you surveyed them? Did you miss the point about highly competent people using things which are 0 EV or -EV and swearing by them? It happens in most fields.

We could have an interesting discussion here but you're at tard level, bro. You cannot keep up with the conversation. I anticipated and answered your next post before you made it. This went right over your head - as I literally predicted it would:
Anyway, good times.
Quite a few would disagree with that last statement. I'm top 5 in BFI easily by real time, posted short term trades.

So no one should speak on TA except the most successful traders in the world? Why are you speaking?

The quants who work for firms like Renaissance doing pure, advanced technical analysis at the cutting edge (for tiny returns leveraging large amounts of capital repeatedly that retail can't capture due to fees and spread) laugh their asses off at retail technical analysis. They are way way way way smarter than you and way smarter than me.

Do they get to speak? Do you get to speak given that you disagree with them and are substantially less intelligent and knowledgeable than them?
So you recognize the fallacy but decide to go ahead with it anyway? Good job bro.

What about the authority of people who have studied TA extensively in the published literature, who find that there is no meaningful short term TA signal, and that support and resistance are bull**** with an upper bound on their effectiveness that's less than retail fees and spread?

I'm quoting these guys. They make strong claims with strong evidence. The strongest claim you have made is that "some of the best in the world us TA somewhat". I've already explained why that's a ridiculously weak claim. Some of the best in the world have used astrology and various other totally worthless methods some of the time. See the problem with your claim? Hell, the entire expert-level options pricing market makers horribly mispriced tail risk for 20+ years - yet they still made money despite widely doing and believing in a model that was -EV. See the problem with your argument, you moron? Of course you don't.

It is very easy to prove that TA works and has a nice edge if it did in fact have that. Many researchers have attempted to do just that - it would make them famous. No one has managed to do it. That lack of evidence alone says quite a lot.

i think MrBaseball will agree but we'll see...

i think trading most futures traders in the pits put most of their analytical weight on supply and demand, aka technical analysis. whereas an org. like PIMCO spends alot of time/effort in analyzing world economies, Fed statements etc. they too look at supply/demand

i think TA is analogous to efficient markets debate,

strong form of EM: no way to beat the market, even inside info, 100% not true as you'd make fortune if you knew large cap stock earnings reports before they were released.

weak form of EM: easy to significantly beat the market with very simple analysis... assuredly not true. of course people say "i loaded up on Amazon and trounced the market". i'm not sure how to judge that based on weak form EM.

semi-strong: somewhere in-between weak and strong and how the world really works. alot of time/money to get a moderate advantage over market (and if you are big, you will have all kinds of other impediments that your skill has to overcome)

i think this is very accurate for TA being short term i.e. support/resistance......... 12 month momentum is different. is that TA?

one question for everyone:

i have a very successful hedge fund based almost entirely on 12-month momentum of large stocks. i.e. right now i'd be very long Amazon.....

is this TA or not?

I have no opinion on whether or not it works, but don't think it's been adequately addressed why TA couldn't be done by a computer based on an algorithm? (Therefore making it a non-advantage for humans)

Your posts are layer cakes of fallacious reasoning.

The implied conclusion of this analogy

is that TA is no more efficacious a trading strategy than roulette. I mean that is your claim isn't it? My insinuation that you are only here to win internet arguments is illustrated by the fact that you sometimes even argue with yourself lol.



Assertion fallacy: Proof by assertion, sometimes informally referred to as proof by repeated assertion, is an informal fallacy in which a proposition is repeatedly restated regardless of contradiction or evidence to the contrary.

Inserting words into a statement isn't a reductio, ******.

Here's a good example of why not:

1) I know a guy who made $15 million dollars playing poker because poker is a skill game.

2) I know a guy who made $15 million dollars playing roulette because roulette is a skill game.

The difference between these two statements is that poker is a skill game. You have done nothing to show TA is non-efficacious but make assertions and call people dumb. Nor did I claim the guy I know who made 45k last week was evidence that TA was efficacious. I simply stated I knew a guy who made a significant win using TA, and that a good deal of highly successful traders use TA in a limited way to increase their edge. You're simply a moron who can't keep his mouth shut.



Neither did I.

Even if that was true, it's the same ****ing claim I didn't make that you're railing against.


I'm not speaking on TA. I have no opinion in the matter. I just wanted to bait you into flying into one of your hubristic rages and showing yourself the fool again, and it worked.

[QUOTE=JKC;54306536]I have no opinion on whether or not it works, but don't think it's been adequately addressed why TA couldn't be done by a computer based on an algorithm? (Therefore making it a non-advantage for humans)[/QUOTE

excellent point

that is being done right now and been done for many years.

i guess the question revolves around AI and how much the machine can just learn on its own.

but someone owns that machine. and someone has to give it some concept of what its supposed to be doing.

we've already seen this i think...... funds used to do really well just trading earnings surprises after the fact i.e. post drift. but the everyone starting doing it and now the post drift happens right when earnings are announced (and it's become a real game to figure out what is an actual positive or negative surprise)

some top hedge fund practitioners have a concept that there's X amount of Alpha in the world, say $100B. and more and more people/computers compete for it every year... so it's more people/computers and realistically probably slightly less alpha over time.

of course there's a debate as to what is alpha....... loading up on FANG stocks the last 5 years, is that really alpha? not sure.

Well. that is pretty much what HFT is. Computers trading based on technical data. Competing with that is pretty much fruitless. The thing is as had been mentioned earlier is that the definition of TA has been butchered pretty badly in this thread. There are infinite ways to apply it using infinite time frames and infinite trading vehicles. What this means is there are several TA niches and approaches that the algos aren't even considering.

Your own goals are amazing man. The above is precisely a reductio ad absurdum of the implied premise used to justify the "because".

God, this is classic. You owned yourself in the Tesla thread and slinked away when people started telling you what a ****** you were, and now you do it again here? BFI's dumbest poster from a rich field.

So you got butthurt with other people calling you a fool in the Tesla thread from your stupid own goals and lack of honesty (tell us again about methane?), and now you're entering other threads I'm in you have no opinion or interest in because you (self-admittedly above) want you my attention?

BFI's dumbest poster without a doubt. Congratulations. I actually like talking with stupid people but you're so dumb and lacking in original thought as to not even offer interesting stupidity. Ignore.

I don't know why the financial industry would put out research showing that the retail investor can beat the market with technical analysis. It is the financial industries business to make money off of clueless retail.

This thread is a trainwreck. You can't seriously discuss the matter until everyone more or less agrees on a definition.

But, ignoring the edge cases, we can probably all agree that trend lines are TA. I have had this conversation before and probably won't have time to fully engage here. But trend lines are real:

An Analogy:
They are simply a relationship between price and time that repeats. Take a simple example of PS4 consoles at a Walmart in some town in America. The store has 100 consoles in stock. And every two weeks they sell out of them and do a just in time inventory update and restock. This goes on for 52 weeks let's say.

This is a trend line. And the same thing happens in the markets. It is obvious and anyone who argues otherwise is _____________ (fill in the blank for yourself).

However, the trend line is a model, and especially in this context, a model and reality are two very different things. The above example assumes more or less a steady state:

Say after a year in the above example Sony releases PS5. This creates a shock to the system and results in a new state. What do you think happens the relationships between supply and demand...price and time, for PS4, in a world that now includes PS5? The equivalent in the markets is a fed rate announcement, or a favorable court decision, new legislation etc.

Complex Systems:
Another reason that TA has at least some validity is that financial markets, and especially modern ones, are level 2 complex systems (an example of a level 1 complex system is global weather). They observe themselves (or we observe our actions, the distinction is academic) and this observation affects their state, a feed back loop. Soros calls this reflexivity.

So, because some actors believe in and act on TA, that in and of itself, makes it some degree, valid.



Anyway, just my two cents. View it as a tool and use it as much or as little as you like. But it should really no longer be debated here that there are successful traders that use solely TA, successful traders that use no TA, and successful traders that use both fundamental analysis and TA.

Also, the ideas of state space and fractals are very real in the markets, trust me there are momentum algos. You are truly fooling yourself if you don't think, for example, that a bunch of bulls place their stops below recent lows...

Momentum algo is actually an interesting case.

Basically momentum was a detected phenomenon. Academics broke into two camps:
1. that's proof of market inefficiency
2. it means the CAPM model is incomplete and momentum is another beta

People in 1 camp started momentum funds and made out pretty damn well... then they shut down shop once they couldn't find easy things to target.

Same thing happened to Fama's 3 factor model proposing that market cap and p/book ratios (small caps and companies with low p/book did better) have their own betas. The idea was the market demanded a premium for investing in small cap and high p/book.

That inefficiency got driven out over time. At this point in academic literature, it's basically settled Fama and French were wrong those were risk factors market would compensate for.

Good starting point and way better than mrbaseball's cowardice.

I completely agree. Trend lines are definitely real and are ubiquitous in nature. For example, random walks exhibit powerful trend lines (as well as support and resistance):



A few more (literally the next three I pressed generate on):

1. Very powerful trendlines (who could deny this??) as well as support and resistance. Notice how the market buys it hard off zero at about 700, an event which sets up a strong momentum run after it bounces of support as often happens (a classic TA signal):



2. Here we see a breakout move shot down hard, followed by the failed breakout initiating a large momentum selloff. Perhaps some fundamental news happened at about step 200 of the random walk, but it doesn't matter - the TA trader doesn't need to know it and can simply read the reversal in the charts and make money. The random walk market is making its sentiment clear through the price action. Also notice how it crashes once it breaks support at the zero line at around day 300 - no doubt a lot of traders set stops here and panic selling started, which was responsible for the fast dive through zero:



3. Extremely strong momentum trends in this one - aTA trader would have made a fortune here following momentum upward for 800+ days. The edge is obvious here and anyone who denies it is an utter fool. I know a guy who played this stock (random walk) for its entire uprun and make amazing money, buying every dip then getting out when momentum finally broke down at around 820 where it broke the previous higher low. Clearly trend and momentum following TA is working here and anyone who denies that is a fool.



So yeah, the evidence is in. Clearly, random walks have trendlines and support and resistance anyone who argues otherwise is _____________ (fill in the blank for yourself). Anyone denying that you can generate +EV using trendlines and support and resistance on random works is a total fool.

Hi TA experts,

Saw this classic and very strong technical pattern on a new stock I'm looking at, ticker RNDW. Bounced off support no less than eight times around the even line before breaking through and rapidly dumping after support broke (obviously a bunch of traders and investors had set stops at support - how else do you explain the price action?). Then it entered a very strong and clearly defined down channel. Investors are exiting in droves.

This has to be a sell, right?

Come on man. Is that really the best you've got, or are you holding back?

1. First of all it is trivially obvious that locally, just because you observe trendlines in a random walk does not invalidate them in the markets (or any other domain...).

2. While mathematically two points determine a line that is only useful when you are drawing it. You trade the ensuing (3rd, 4th, 5th touches) moves. I will post a real world example is I come across one. But I really doubt I will bother to look for one. Maybe someone else has one handy?

Ha you should make YouTube videos and sell your picks through a subscription site.

Of course, but it's not interesting to you that 4/4 random walk graphs I generate show notable support, resistance, trendlines, momentum, etc, on various timeframes? What method do you use to separate noise from signals? We already know that technical analysts on this forum are Megatards because they don't even correct for the most basic and prevalent of noise, market noise, which drives most individual stock inflections. Why should I trust that people this stupid and incompetent at TA aren't simply being fooled by the fake signals that are as common in random walk graphs are as they are in stock graphs?

All manner of people get fooled by randomness and the confirmation bias resulting from high placebo/noise effects with weak to non-existent signal. From scientists (50% of published peer reviewed studies can't be reproduced and are thus false) to doctors (net success rate of antidepressants is upper bounded at around 5% and is probably 0 yet the widespread narrative is that they work/are needed; success rate of all chemotherapies are upper bounded at 1% - 52% survive without treatment and 53% survive with it - and less than zero when taking out non-Hodgkiins lymphoma and childhood cancer which have high cure rates) to hedge fund guys into astrology and other ideas.

The semi-smartest people in the world are fooled by randomness all the time. Why wouldn't some loser cucks who can't find alpha and get drawn into TA be similarly fooled by randomness, and see TA working in their successes? That's exactly what's happening here. TA is one of the greatest cuck-traps of all time, given the huge volume and variety of data to snoop and the fact that the long run is quite a long time and that people are terrible at intuitively understanding noise, causality and randomness, and are horribly confirmation biased.
I agree - the last graph I posted of RNDW showed no less than 8 bounces off support before the breaking of support put it into a long downchannel trend. These are clearly tradable patterns and you're leaving money on the table ignoring them...right?

Also note that these are random walks of pure a 50/50 decision. Add in random volume as well as random direction and you'll have beautiful TA patterns made of pure unexploitable noise.
Thank you. I understand the cuck mind well.

Strangely though, selling subscriptions via Youtube videos is what most of the examples given of claimed successful pure TA traders end up doing. Curious, no?

The key question is not whether a certain pattern of prices or volume can predict future good bets. Whether it does or not is not a logic question. It is a human behavior question. And there is no reason to think the answer must be "no".

But if I'm not mistaken TA advocates claim something more than the above. They seem to claim that even if everyone knows about the historical pattern and its past ability to predict, that somehow that ability remains even if all the traders take it into account. That's similar to believing that there can be a perpetual motion machine. (Notice that what I wrote is not refuted by those outfits who use patterns to buy and sell within the space of a few seconds. Those patterns, even if they work, could be known yet still work, because very few people can take advantage of them.)