Harry, Ok here is one more question for you regarding the subject of whether TA is valid or not. First a background and some assumptions: Lets assume that Technical Analysis is statistically unproven and is nothing more than random chatter or noise. Furthermore, lets assume that you have developed an econometric model to trade in the markets that has a core reliability on the inherent principles of mathematics. (I am over generalizing your model for the purpose of this post so, yes I am aware of that). Now last assumption, your model was developed for one singular purpose in mind ultimately, and that is to "beat the market", "take money from the market", "out perform other market participants", however you want to phrase it. Furthermore, I believe that the conflict in this thread arises from the fact that TA is more of an art than a mathematical science and the conclusions of TA differ from user to user unlike in Science where the results will maintain consistency for all. Ok...with that said...... We know there are a population of traders who are accomplishing the singular goal of extracting money from the markets consistently over a long sample of time and at a rate of return that surpasses the historical internal rate of return of the stock market itself. Furthermore, if these traders have been relying on TA (or so they have believed) all this time. We assume TA is false for this question....then what would you attribute their success too? That is the question. Thanks Harry.
In the same S&C mag there is in fact also an interesting and more thorough answer from Don Bright I have OCRised below: Stock & Commodities Q & A <font color=RED>SKEPTICAL OF TECHNICAL ANALYSIS</FONT> I'm looking for objective evaluation of technical analysis â or doesn't. Technical Analysis is a seductive idea: just apply certain rules to historical data, look for certain patterns, invest accordingly, and wait for the profitsto roll in. But <font color=RED>the very fact that I want to believe it makes me cautious. People have devoted their lives to, similar ideas (biorhythms, astrology, gambling systems) with great passion, yet these ideas turn out to be unrelated to any ability to predict events</font>. What is the objective evidence that technical analysis works, such as in a controlled study? And I don't mean a study based only on analysis of historical data. I mean one where analysis is done, predictions are made, and an objective researcher evaluates the data to see the predictive value of the method. Do you know of such a study? ?Phil Quigley, via e?mail Don Bright's Answer First, we must define technical analysis. According to Investor Words Glossary, technical analysis is: "A method of evaluating securities by relying on the assumption that market data, such as charts of price, volume, and open interest, can help predict future (usually short?term) market trends." Now, lets look at the operative word, assumption, in the definition. Using a chart to trigger a response in anticipation of future movement assumes that there will bc enough people relying on the same chart to trigger a self?fülfilling prophecy. Most of the traders at Bright Trading do not use technical analysis to make intraday decisions. They are, however, very aware of the basies, such as support and resistance, as well as S&P numbers such as high, low, support, resistance, and pivot points. Let's discuss that. When a security or an index reaches a support level, is it really caused by the chart, or has it simply reached a price where there has been a great number of buyers? Charting historical data to simplify the viewing of price and volume is beneficial. Using charts to visualize upward or downward trends can be useful as well. The more exode analysis methods tend to be a bit too esoteric for my style of trading. 'Mis is not to say that they do not work for longerterm trading or investing. The best traders use many forms of trading methodology (tape reading, momentum, technical analysis, fundamentals, breakouts, breakdowns, and so on) in their overall attack on the marketplace. They do not get locked into one single concept.
I will answer thoroughly with a VERY LONG thread that's why I postpone it for later . Because until now I and others have answered briefly in some posts but it is as if it were just opinion from me and so not worth.
If a person succeeds in generating profits consistantly for a sustainable period...then the system used is/was valid for that time period for that person only... Trading is an individual matter...and is perhaps impossible to duplicate for this reason... However...is it easy for Wall Street firms play on this hope...and communicate to others by implying that they have something to do with a successful individual's success... If a firm has 1000 traders..and one made $10,000,000 in 6 months....and the rest failed...it is likely that they would still bring in business because of the one successful one they had... Surely...in a fee market...this applies to all business endeavors... not only trading... Being a real player in the trading arena is a lot like a young person in high school wanting to be a player in the superbowl.... ut when you are looking at a person's daily blotter which shows consistant success...then the name of this persons technique called technical analysis or whatever...is valid...
Nice post. Throughout this thread people attribute uses for TA. Primarily "prediction" comes up. This common theme is probably where most people go wrong using any approach. Another facet of this thread is backtesting and forward testing. The tie in is prediction usually. That is some one thinks up an investing approach and then they try to see if it works. One alternative that has not been delt with here is the history of TA. It is fairly orderly in terms of the logic of what came first and also the fact that the big pieces were dealt with first and then as science provided convenient support devices more and more refinements were made. It all went from the general to the specific and at one point branched out into two main streams. The biggest single shift was from reporting the staus of markets to "predicting. The bracnching occurred when one emphasis became attention to details in a macro context. The other emphasis was a micro focus on making money in individual's portfolios. I am just an amateur who have used TA for portfolio operations. But I have done this using TA since 1957. I learned from the general to the specific since many detailly things did not exist when I began. Luckily throughout my investment career I have been in TA whereas most others who invest have not used TA. My financial forte, academically, is a combo of problem solving an inovation in corporate prowess. By not being employed in the work, job, salary sense, I have stayed out of the conflict of interest and reporting requirement game throughout most of my life. To also stay apolitical and be available to the political arena has also been a goal of mine. So for about 50 years I have done two basic thnigs: used TA to invest and done anything I wanted to with my time. Here are the following opinions I have about investing with TA. 1. Historically, the advent of support technology, has screwed the financial industry. They have gone to macro analysis and stuff. This branch has several basic misconceptions that they do not want to prove are wrong. THe consequences are that the big guys cannot beat the averages. historically the averages they can't beat ere the first use of TA and you can check out Dow for that. This is a very difficult situation for the brainpower of the industry. It is approaching a dilemma status. 2. The shift from maintaining a market status effort to a "predictive" orientation, was an unretractable development because of it's orientation to "beating" the competition. There are two major facets of this: "predicting" does not work; aand "predicting" is not necessary. These two items are very difficult to understand by most people. The "history repeating itself" and many other liberal arts things, drive this irrationality. 3. the progress of TA into the financial industry is a slow process. currently we are experiencing MA's for index volumes becoming commonplace. secondly A/D indicators are now being offered on software (see Qcharts' recent inovation). TA is no where as yet in the mainstream. I went through the introduction of TA in one of the major brokerages. They periodically did stuff to turn my selections into theirs for national distribution and as a consequent they added staff in NYC as a TA department. (my personal accounts traded (in a day) 1.89 times Gary smith's largest daily trade as reported in his book (I will not provide the LOC #)) so I was influential at that brokerage house. To intorduce TA into the financial industry really wrecks much of their operating philosophy. This issue is the key one for what role TA plays in making money. Employees in the industry vis a vis investing are sales people and TA is not a sales asset. 4. To make TA work best you may not use it in a macro manner. As in any science or maths based application, the mor narrowly defined the applicationis the more significance it's results will be. The economic development of the US is reflected in mass production and education. Everyone gets the same and it is called, strangely:"opportunity". In making money the opposite strategy is applied vis a vis "competition". To make money potential investments need to compete for scarce funds. TA affords the best analytical tool for this. Technology finally gave us two TA tools: RS and EPS as percentiles (99 groups, they just missed by one and it is okay). Cullling and selecting a repeatable universe is the first major TA tool for success. TA has two applications: analysis and making investments (Monitoring and Trading). Analysis is as important as any TA factor. 5. Making money with repeatable equities and indexes. TA shows how markets work for selected excellent universes. In the context of the market operating points, you choose the trading strategy. All investments are constrained to making money via change. TA is the only and best method for enabling the two effeiciencies deployed to be optimised: the market efficiency to deliver capital and the trder efficiency to extract it. TA is the bridge that joins the two and this is what is optimized in a strictly micro manner. You can simply regard this as bullshit. IT isn't. It is 50 years work where the yield is excellent. what makes it most powerful is it's simplicity and transferability to anyone of any potential. there are no special training or brainpower requirements to be successful. What makes this strange and incomprehnesible is that it is not the mainstream way of thinking. It is uncomplex and logical and based on the market principles of operation just using the logical historical basic theormens that have been laid aside in favor of machinations done by powerful macro machines and maths which try to encapsulate whole markets. No one plays whole markets. People who get rich focus on the places where money can be made most efficiently. In the next 18 months a lot of people are going to be in a different place. TA is what is going to take them there.
There are many types of TAs. Some are obviously targeting prediction like Elliott Wave. Others are just decision aid. But it depends also on what you mean by prediction. Many people think that because they use a probabilistic framework they don't predict whereas statisticians would consider that it is prediction. Because saying that the price will be either at this level down or at another level up - even if you don't know if it is up or down - is a prediction that can just be wrong later since the price can just never touch both levels for example. Saying that when you buy at one's level and put a stop means that you favor a direction and so you make a prediction although you know that you can be wrong and for sure for a prediction wouldn't be such if you was always right. Deciding step by step - which exists also in statistical decision theory - doesn't elude prediction between steps. So you are always making prediction in the sense I illustrated above. So you need prediction whatever procession decision you chose, but it will be more or less harder, more or less profitable or not profitable at all depending on the tools you chose but also on your mastering of the tools. It is sure that I can't use Elliott Wave as for me because it is too much subjective, intuitive and unprecise for me, it doesn't mean that it is completely useless. There is the problem of scientific validity also of elliott waves for example which is in fact the question of the thread. For that see http://www.elitetrader.com/vb/showthread.php?s=&threadid=24469&perpage=6&pagenumber=7 Didier Sornette and Jean Philippe Bouchaud - IBM laureate of young Researcher 10 years ago - who said in "Stock Market Crashes, Precursors and Replicas": "It is intriguing that the log-periodic structures documented here bear some similarity with the "Elliott waves" of technical analysis [citation Elliott Wave Principle by Frost and Prechter]â¦. A lot of effort has been developed in finance both by academic and trading institutions and more recently by physicists (using some of their statistical tools developed to deal with complex times series) to analyze past data to get information on the future. The "Elliott wave" technique is probably the most famous in this field. We speculate that the "Elliott waves"â¦could be a signature of an underlying critical structure of the stock market."
It is not the proof per se that is really interesting, it is that in the path of proving something you get a better understanding of how the things really works in details and that will conduct to more performance and reliability. For example everybody knows that with traditional TA if you try to improve the reward risk ratio you will degrade exponentially the ratio of good trades so that it seems impossible to improve both. And it is impossible as long as one doesn't have the true equations of the market or at least something that is near. Whereas with my fundamental approach I have these equations and I will (try) to prove statistically that it would be possible to have consistlenty a high risk reward ratio (>3) coupled with a high probability of success (>75%). I have settled a statistical page here although it's just the beginning: http://www.econometric-wave.com/statistics/1/home.html.html