No, variation in stock prices are highly random.
The only thing we could hope to do is to make an educated guess by analyzing the available data from the past few decades(or any reasonable timeline, for that matter) and with an appreciable plausibility, decide where we should(and should not) invest our money for gaining the maximum profit.
# What should we do for predicting the stock prices?
In our modern socially connected world, various sources of social media consistently plays a major role in predicting the image or relevance of any person or any company in general, thereby determining the stock prices. By extracting relevant data from a myriad of these sources, we can make good educated guesses regarding the prices of the concerned stocks. A few of these sources include:-
- News
- Company tweets
- Company History
- Industry Trends
...and many more.
As the economist Burton Malkiel (in his book A Random Walk Down Wall Street) says :
"A blindfolded monkey throwing darts at a newspaper's financial pages could select a portfolio that would do just as well as one carefully selected by experts"
If that is the case, then why do the companies like Morgan Stanley, Goldman Sachs and Citigroup hire data scientists for building predictive models? Why do all the companies in Wall Street use Huge data analytics to begin with?
In the beginning, people in Wall Street used to randomly guess the probability of price variation on each stock, haphazardly calling, yelling on their phones, trying to bet on the best of odds.
Now, many top analysts calmly sit in front of their computers and try to get the best bet, using machine learning techniques for predictive analysis. Still a better way than teaching a monkey to throw darts.
| Earlier ways of predicting prices | Current method using algorithms |
|---|---|
 |
In this age of algorithms, we can obtain a reliable guesstimate for stock market prices using accurate machine learning models and Quantitative analysis (the study of how certain variables correlate with stock price behavior).
In the light of the above fact, my objective in this competition will be:-
- To develop the technology, which will be open to public, that will help them decide where to invest and where not to
- To build an end user product, where a user can upload a CSV data file of the stock prices (spanning a reasonable enough timeline for accurate prediction) of any given company in order to train an algorithm that will predict the closing price and highest stock price for a given opening price of the company stock, for that particular day.
- Visualization of Growth in the company's share price using Chart.js(graphical visualization library in Javascript)
- To provide a platform which helps the users to compare the growth of share prices of two different companies
- Python
- Tensorflow
- Pandas
- numpy
"Stocks are the only thing that people are happy to buy when the price goes up" ~ Warren Buffet