X, Im pretty sure I am with you on this one. I can't draw worth a damn, but I find myself trading best with a naked chart and some patience. Point is, it is more of an artisan approach compared to pure models. Now where my visual approach ends is the stats of w/l and p/l. I believe those come first to any serious trader. Being right or wrong is what newbs want, pros get $ in the account and walk away. Simple as that.
Absolutely. Many patterns of the reality are easy to handle with use of our brain, but literally impossible yet to replicate using algorithms. Car driving comes to mind among first examples: we all do it easily, but is there yet a robot which drives the car as well as a human does? Does it mean this given human ability is something weird only because it requires too fine information processing vs. what is available with modern technologies? I don't think so. Our society often gets way too technocratic, we say if something can't be "objectively" tested by computers it doesn't exist... Yet we survive mostly thanks to abilities which are very subjective, most of our instinctive reactions we don't even consciously control, yet they are the basis of our behavior and this subjective practice is very successful looking at how humans as a specie dominate the world.
Oh man, the TA apologists are out in full force. Yes, these defenses of TA always fall back on anecdotal evidence-- the I DO IT, so it must be true claims are humorous. The very premise of your post is in error, as is the premise of TA which is the same thing. Human behavior does not drive price, money majority drives price. The majority of money can be controlled by one person or by 1000's at any one time--- A single fund manager gets a bad burrito at Chipotle-- he hates being sick to the stomach, so he decides to PUNISH Chipotle by dumping his shares. He is a large share holder and the price drops 20 points due to his selling. All the longs can't support the stock, it collapses. One person's move not "human behavior patterns" moved the stock-- so unless you believe repeatable human behavior patterns can predict individual actions-- TA is nonsensical. Money Mass not sentiment moves price. Secondly, the argument from the TA brigade swings from claims of objectivity to subjective chart reading. sometimes in the same post. You do this for example, "it takes a human to drive a car" I can infer you are referencing "it takes a human to read a chart"--- Although I have seen ZERO evidence-- possiblily some humans can succesfully and profitably read charts--- this is not my issue-- one can't argue with an art form. I don't believe it/you-- but that doesnt mean its impossible subjectively. I take it, you are admitting in the above post that TA can not be tested and is a subjective art form? We completely agree, if this is the case??
Good morning Surf. 1) It's your example of fund manager being sick for the stomach and "punishing" (wow, amazingly unprofessional for Wall Street pro!) the stock is anecdotical, not the TA. I don't think you will argue that most of the time markets are driven by routine buying/selling mostly of institutional nature. That's their everyday job, their traders do it literally every day and they do have recurring patterns, it's simply impossible to tweak one's game every day to make it completely different if they care of it at all which I guess they don't. 2) I tell you that I trade by reading charts. My live trades are publicly available. I manage OPM and those people see the same trades in their account. It's not the matter of your belief it's the matter of you accepting this evidence or not. 3) I agree with you that many forms of TA are hard to mechanically test for an outside person who simply doesn't know all nuances of the approach. In this sense yes, it's an art form and not the exact science. But I wonder why do you say "artform" as it's somewhat inferior to purely mechanical approach. What's wrong with "artform" if most things we do in life are actually an artform and not the science. Don't you think that you are too technocratic in relation to trading and probably restrain your own progress, which would be possible if you accepted that artform can be as if not even more reliable than an algoritm. P. S. You know, my father is a professor of theoretical physics, my mother was professor of literature, I grew up in the family of continuous discussion of art-form vs. exact science and learnt to grasp the best of both worlds. It really is possible and rock n rolls when you do that.
TA can be tested and Ol Yella has taken it upon himself to conduct a technical price analysis of a single setup based on price action: The opening range breakout (ORB): From here, you could test the same constrained setup further by adding to the profit target and finding out if the result can be boosted; you could also decrease the stop loss and find out the level at which doing that decreases the result. Also, for intraday scalping using a 5-min chart, I like to analyze at least 3 months of data. Then the testing can continue: What if we eliminate the open/close (candle color) constraint and instead bracket both sides of the opening bar? What if we filter out bars that are narrower in range than X and/or wider in range than Y? What if we exclude long trades that have you buying into a falling 20EMA or selling into a rising 20EMA? And so on, until you compile stats that demonstrate an acceptable enough net profit to consider adding this setup (and the trade management rules) to your arsenal, or demonstrate a slim enough profit or negative result that excludes the setup from your plan. This is what I did for a variety of setup ideas I had based on observation of price patterns leading into price swings of N or better. It took a long time.
Quote from Ol' Yella: RESULTS UPDATE: <b>Of 30 trading days (Sep 12 to Oct 23) there were 16 on which a trade triggered. Using the above rules we have: 4 winners of .24 = .96 ($960) 5 breakeven = 0 7 losses of .12 = -.84 (-$840) assuming no slippage that gives .12 profit = $120 per contract for commissions I assume $5 RT * 16 trades = $80 So $40 profit per contract theoretical result. In any event it is at least profitable on paper. On days 31 and 32 a trade triggered each day, giving a .24 winner and a .12 loser. So that would have boosted the stats.</b> Thank you for the testing, Ol yella. Perhaps running these tests over the last 2 years rather than 30 days would yield differing results? Thanks! surf
Not much cause I just recently started, but that can be an indicator of how "real" things are nevertheless. If reality is objective then chart reading TA patterns do exist and it is possible to make consistent profits using them. Because your position is simply "I don't believe" while my position is objective proof of my words.