"Branding Quantitative Analysis as "Technical Analysis" will probably bring in some violent reactions from quants. But I just want to point out the similarities that they share. In fact, it can be seen that Quantitative Analysis is a higher form of Technical Analysis....." http://rmquant.blogspot.com/2007/02/quantitative-analysis-highly-technical.html If we want to create a new "higher power" grouping that separates itself from "TA", I would be all for it.
My group and I have been trading for 19 years using our own technical analysis. Back when I started it was much simpler... But I rely on many of the same indicators I used then as I do now... Why technical analysis works... is because so many traders and instituions use it. Think of all the hedge funds that run algorithms.. all the prop shops with thousands of traders using indicators... how much money is being traded all day long from people looking at their charts online.... Do you think for a minute that people can just watch a ticker or quote and make wise investment decisions? where or when do you enter? did the stock breakout? Is the stock being forced lower becasue traders keep selling/shorting each time it hits the moving average? These are just things you cannot do without watching the charts and basic chart indicators.... I do agree though that many of the indicators are garbage. Many of them are similar indicators with a different name... And in my 19 year career I have not met a person who consistantly makes money with "auto bot" trading software... its just impossible...physically to leave the software on all day/night and make money (futures/forex open 24 hrs)... A trader must watch volume, news, and other macro facctors that affect stocks/securities. Oh well... just replying to your post that Techncial analysis is worthless... because I feed my children with my analysis....and take nice vacations with my wife... etc. etc.. =) J.J
Maybe you can further enlighten us as to what exactly is EMH.It has always puzzled me. For example,at spx 666 when there was panic,get me out at any price orders going through,..was that effecient? How about now,this manipulated rally, exactly how effecient is it now? How do several 100,000 traders all know what the EMH price should be right now.. and if they do why would they buy or sell it? (This is not a personal attack,btw) You see,i actually agree with a lot of things you say,and dislike the liars and scammers too. It's just that i find your trust in the 'experts' kinda worrying. As an 'expert' i can almost make anything a 'truth'- THE SKY IS NOT BLUE.. human eye sees it as blue but then,cats and dogs are colour blind,and if they could talk (and they were 'experts') they would disagree. So TA/FA.... Here's why most experts should be routinely ignored.They have a very loooong track record of being completely wrong about most things An economist today,said in an interview,that he agreed that the UK was going through the worst recession in recorded history, but then said that,since the average bear market lasts 21 months,he expected signs of recovery early next year. To me,that is a statement of staggering stupidity. If this is the worst recession ever,why is he expecting a normal bear market? These guys can never say they honestly dont know.Because,let's face it,it is unknowable. To express an opinion on something that is unknowable comes with the terratory for these guys,and is how you get funding.Meanwhile,in the real world,it is an act of stupidity-but i guess i am guilty as the next guy of having an opinion on most things,but then i'm not held up to be an 'expert'. So excuse me if i don't take too much notice of any study made by these guys regarding TA- a bunch of guys who have never traded one thin dime in a market. As for FA why would i trust a liar who hides his losses off a balance sheet unless i'm very good at reading between the lines? I can trade without a chart.. The real issue here is, WHO REALLY MAKES THE REAL MONEY in these markets? who do you visualise these people to be? What would you expect their computor screens to be showing? What books would be on their shelf Let's face it,almost nobody here has access to inside information.You wanna make money in the hardest game in town where almost everybody fails !? and what are you bringing to the table,, canned indicators with your own special blend of herbs and spices. Most of you guys have got to get real. Not suggesting there is a secret to be discovered-canned indicators can work if you have very good money management.. i guess, like i said the future is unknowable. All i'm saying is you need to think outside the box and ask a different set of questions beyond the conventional,to get a different answer that works for you.Trust me,in this regard,you will find 'experts' to be 100% useless on your journey-be they acknowleged or self appointed ones here. If you find something really good you won't discuss it.Maybe hint at the sort of thing it involves.If you do you will get laughed at,ridiculed and there will be real hostility sometimes.Either one of 2 things.. your system is crap or you are really onto something-if you're making money only you will know.That's when you know for certain that most people are wrong most of the time in this game..but some of em can still make money anyhow..go figure
Black-Scholes doesn't do ANY of that. Neither the "near perfectly" part nor the "without any..." part. The latter in particular uses all that other information, it just buries it in proxy inputs to the pricing models.
Go through it then, because it's expected in the derivatives section of the CFA Curriculum to be able to calculate the value of an option using inputs you must calculate based on a given dataset.
Two inputs to options pricing models - price of the underlying and volatility - are functions of balance sheets, earnings growth, etc etc etc. Volatility is also a function of the market's consensus confidence in that FA. They aren't the only factors, of course, but they are there. The fact that pricing models generally hide those behind proxy numbers doesn't change the basic fact that they are still dependent on them. Hope that helps.
This is perhaps one of the alltime lame beliefs of TA aficionados. It "works" ONLY if it yields a consistent positive expectancy. TA related stuff (indicators, etc.) send way too many traders into the red. For a long time, almost all sailors believed that sea monsters roamed the open oceans, one reason they refused to go out in the open ocean, but stay in sight of land. Columbus had to keep 2 sets of progress records, it was so ingrained in their belief systems. Much of the rest of your post was too naive to take your claims credibly.. You think hedge funds and investment banks are running ADX and MACD to bank profits? Seriously?
I approached TA very analytically and built some very sophisticated software that created and backtested candidate strategies built from traditional TA primitives like MA's, oscillators but it also had the sophistication to create new indicators from scratch out or price and volume. I tested on normalised real market data across 10 currency pairs over 5 years worth of data with independent training and validation data sets. The software backtested over 300 candidates per second for 9 months continuously which was approximately 7 billion combinations tried in total. This was all text-book stuff, every thing was correct, checked and double-checked - the only thing that I was bringing that was new was the technical expertise to build a system sophisticated enough to do this in a reasonable time-frame. The end result was a fairly long list (around 200) functions that reliably made on money across the currency pairs on the test and training data. Combining these 200 rules using boosting gave me a killer strategy with an excellent sortino ratio and very low draw downs on every backtest that I did. Annual returns of 60% with drawdowns of < 3% were typical. I then traded this on a demo account for 3 months and it made good returns and the drawdowns were reasonable - not as good as backtests but still perfectly tradable. I then took the plunge and went live with it where it bounced around for about 3 months and then I pulled the plug because it was clearly losing money gradually. Looking back, I merely searched for survivorship bias across all the data sets and found some random rules that were lucky random survivors on that fixed data. They had no future trading merit at all. So learn from my lesson, those of you that are successfully trading solely with TA are probably just a lucky survivor up to now but your luck will change. Don't forgot tens of thousands of intelligent people try their hand and trading but only a handful last more than a few months - those that remain are the lucky ones up until that point.
Would you mind to explain in more detail exactly what kind of strategies you tested? Did you differentiate between bear markets and bull markets? Trending markets and range bound markets? Mean reverting markets or momentum driven markets? Super liquid markets and thin markets? Did you simply buy a breakout or did you fade breakouts? Did you add any filters for your entries? There are many people who do not even use indicators or charts for that matter. TA is a very wide field and you will be hard pressed to actually find a robust strategy published in a book that is sold for $19.95 on Amazon. To me, your studies prove nothing that I did not already know. Keep searching.
Funny that you never see threads "New study says: Fundamental Analysis is Garbage". Is that because it's not or because it's so obvious it is that there's no point in doing the study? Doing an analysis of 7 billion possible permutations of TA sounds impressive, until you realize that the number of TA rule permutations probably runs into the trillions, perhaps even quadrillions, if you assume that the various technical indicators themselves are worth permutating to begin with. Think about it, there are hundreds of indicators each with potentially thousands of optimized parameter values and if you are modeling that space in a way that each model can have anywhere from 1 to a few hundred indicators in combination and then all of the different combinations possible from that, there's simply no way to do a "definitive" study one way or another because we start getting into factorial math, not simply linear math where we are adding indicators or other technical tools to a static base. With all of those potential moving parts, covering even 1% of the totality of options could be beyond the reach of today's computing power. Personally, I have no idea if TA is garbage or not, but I can recognize a premature conclusion when I see one.