Technical Analysis (TA) is usually based simply on data visualization and on some simple statistical tools (rolling averages). So TA could be a very subjective process. Usually the forecasts based on technical analysis are not so reliable. In every case exist a huge literature on stock market inefficiencies and in this context technical analysis can be useful in order to extract the signal from the noise (and in some way allow useful trading strategies). However technical analysis can generate some data can be analyzed and exploited by a statistical point of view. For example the underlying data which generate the japanese candlestick can be usefully predicted using other technique than simple technical analysis. In this sense it is necessary to build an appropriate data matrix by collecting these particular "complex data" (japanese candlestick consider the intra-day variation so the minima and the maxima related to a single day) and then using some method in order to predict these data over time.

Technical Analysis (TA) is usually based simply on data visualization and on some simple statistical tools (rolling averages). So TA could be a very subjective process. Usually the forecasts based on technical analysis are not so reliable. In every case exist a huge literature on stock market inefficiencies and in this context technical analysis can be useful in order to extract the signal from the noise (and in some way allow useful trading strategies). However technical analysis can generate some data can be analyzed and exploited by a statistical point of view. For example the underlying data which generate the japanese candlestick can be usefully predicted using other technique than simple technical analysis. In this sense it is necessary to build an appropriate data matrix by collecting these particular "complex data" (japanese candlestick consider the intra-day variation so the minima and the maxima related to a single day) and then using some method in order to predict these data over time.

Mahmoud,

In my opinion, these indicators are good to the extent that they are put in the context of the market itself and the underlying causes. While it is true that fundamentals give the state that the underlying company on which the stocks are based on, you have also to take into account that the market moves on speculation, and these technical measures try to evaluate the speculative side of the market. the interesting thing to ask is: are these measures correlated to the real underlying speculation cause? Again, in my opinion this is an open question.

In "Neural network techniques for financial performance prediction: integrating fundamental and technical analysis" by Monica Lam, she combines both types of analysis. One area that could be expanded (I did some searching and could not find information on it) is to increase the number of technical analysis measures not just of the market (which the study concentrated on) but for the particular stock in question (such as moving averages, volume, etc.). The approach I would take on to this type of problem where the fundamentals are not as frequent as technical data on the stock would be to use a mixture of experts model where one would process the high volume of technical indicators and another the slower fundamentals. Some tweaking would be required on the model to account for the decay of importance of the fundamental analysis as it transitions to more volatile events which influence technicals such as recent news etc.

Another interesting area which I would also look into is sentiment analysis for stocks and combine it in the mixture of experts model

Let me know how it goes

I enjoyed Arturo's discussion and review. Agreed that fundamentals are slowly moving data, mostly intrinsic to the underlying company, and usually very domain knowledge-driven. Technical data is reflective of both the fundamentals and quickly changing, perhaps unknown external forces. An ensemble model approach combining both types of data seems intuitively to be a promising approach.

I am agree with Arturo, combining both fundamental and technical indicators will be useful. There are some papers in which author have used both parameters and there results look promising although it depends on the underlying technique used for predictions but still i think combination of both will be better option.

You can take a look at this paper

"Neural Network Applications in Stock Market Predictions- A Methodology Analysis"

by Marijana Zekic, MS.

the technical analysis is the short term forecast whereas fundamental analysis is long term analysis of stocks. however none of them in isolation cannot work. for example if one want to take a long term look of the market than fundamental analysis will be the basic analysis but timings will be decided by the technical analysis.

similarly in case of short term decision one has to see the technical analysis in the light of fundamental analysis it will help you to determine the under and overvalued stock. For this purpose along with PE ratio, Run test can be used to determine the proper price.

your comments are welcomed.

forecasting stock market can never be that easy. many work try to do so:

1- using HMM. just as univariate time series!

2- using NN: it's a black box. how far can you trust results those can NOT be explained.

3- using GP: check EDDIE 8 by Prof. Tsang. (essex university UK). their work seems to be more reliable.

current research by Olsen Ltd. in Switzerland working on global information system that can integrate many resource of information. you may want to check their works.

Technical analysis gives a statistical interpretation with the help of charts, averages etc. However, fundamental analysis is based on the financial results of the respective company, industry or the details about economic environment of the country as a whole. Both, collectively have to be studied as only one can have limitations. However, nowadays, more significance is given to technical analysis. But for long term, fundamentals are more important. In short term, you may make money by using the trends given by technical analysis, but in the long run, fundamental analysis shall sustain for real returns.

As an economist, I am inherently skeptical of technical analysis and my rational self tells me it is akin to predicting the future by the use of tea leaves. However, I see all too much effort being put into it in the financial industry and used by many profitable traders I know. Flags, pennants, head and shoulder patterns, into the more esoteric Fibonacci patterns. So how do I reconcile the two? If it didn't help at all, then why invest all the effort into it. There are several rationales. First, it works because people use it and there is a self fulfilling aspect to it all. People trade off say a 10-day moving average breach of a 50-day or 50-day versus 200-day. It doesn't answer the question of why people use it in the first place. Second, future asset price movements are unknowable and the 'true' model is unknown as well. TA may be a useful rule of thumb. I think of it as mapping out the terrain as we blunder around a mountain in the fog. I don't know when the top of the mountain is reached so I have a rule that says if I take 50 steps downhill, that is significant. Two economists at the Bank of England worked with me at the Bank of England in the 1980s and did an interesting study of the use of technical analysis in the FX market. See https://www.researchgate.net/publication/4801230_The_use_of_technical_analysis_in_the_foreign_exchange_market

Article The Use of Technical Analysis in the Foreign Exchange Market

have a look at CANSLIM analysis originally described by William O' Neal, kind of incorporates both approaches

Dear Rajveer,

Thank you for introducing  CAN SLIM strategy. It seems his book is a very good one on the subject in question. Do you have the link to the book so it can be downloaded?

Both are different and used for different purpose.

Technical Analysis : It is the study of charts (including its patterns and technical indicators like moving average, rsi, adx, etc). This is mainly used by active traders . Being a trader, i myself have traded on the basis of charts for many years. It is useful when used with Multi Time Frame trading, but only for trading (Not Investing).

Fundamental Analysis : It is used by investor, it is beneficial only in case of long term holding (e.g. 3-10 years). Fundamentals dont change on daily basis, its obvious that you will be getting same balance sheet, Profit loss statement till company declare it a quarter. Fundamental analysis have no use for traders, it is only useful for investor.

Feel free to contact me if any further clarification is needed.