Description: STOCK MARKET PREDICTION USING PROPOSED NUMERIC RANKING TECHNIQUE

FIELD OF INVENTION

[0001] Stock Market prediction is most effective done using technical and fundamental analysis in the world. Fundamental analysis is done by analyzing the balance sheet, seeing the company belief management and the core values, whereas the technical analysis was mainly done using statistical approach. Our approach is basically working on technical analysis using machine learning method. The numeric ranking is given and based on ranking decision is to be taken. High number indicates that its highly recommended and lower number means least recommended.

BACKGROUND OF INVENTION

[0002] A market is a gathering place for people to buy and sell goods. Now again, this market is divided into two categories - primary market and secondary market.
[0003] Securities are created on the primary market and then traded on the secondary market by investors. Securities are created in the primary market. Firms sell new stocks and bonds to the public for the first time in this market. A primary market is exemplified by an initial public offering or IPO.
[0004] In the secondary market, investors trade previously issued securities without the involvement of the companies that issued them. The secondary market is also known as Stock Market. Hence stock markets are venues where buyers and sellers meet to exchange equity shares of public corporations. BSE and NSE is an example of the secondary market.
[0005] Indian Stock Market is again divided into mainly two parts.
? BSE (Bombay Stock Exchange)- The Sensex, or 'Stock Exchange Sensitive Index,' is the Bombay Stock Exchange's stock market index.
? NSE (National Stock Exchange)- The Nifty is the National Stock Exchange's index and stands for 'National Stock Exchange Fifty.
? XMSEI- (Metropolitan Stock Exchange of India)- The SEBI (Securities and Exchange Board of India) recognizes it as a government-owned stock exchange. Banks from the public and private sectors, the stock exchange, financial institutions, and private investors are among the stockholders.
[0006] Now let us discuss some key terms used in the stock market,
1. Issuers: Corporations and other legal entities that are listed on a stock exchange.
2. Trading Clearing & Settlement: Settlement time is up to two days from the day of trade.
3. Clearing House/Corporation: In figure 1 it is shown that clearing house of NSE is National Securities Clearing Corporation Ltd and for BSE is Indian Clearing Corporation Ltd.
4. Depositories and DPs: For security to be eligible to trade in the secondary market, it should be held in electronic or dematerialized form.
i. National Security Depository Ltd [NSDL]
ii. Central Security Depository Ltd [CSDL]

Types of Investors
• RII- Retail Individual Investor: An investor who applies for or bids on assets worth less than Rs 2,00,000
• HNI- High Net worth Individuals: Here, the limit is increased by more than two lacks.
• DII- Domestic Institutional Investor: Insurance firms, mutual fund houses, pension funds, and provident funds are examples of domestic institutional investors. ex LIC
• FII/ FPI – Foreign Institutional Investor/ Foreign Portfolio Investors: Outside entities are investing in India.
[0007] US20050091146A1 In this invention, technical analysis is studied with the help of an intelligent agent. The invention is a stock forecasting system that, over time, learns to profit from short-term stock forecasts and, because of its intrinsic flexibility, remains lucrative in practically all market conditions.
[0008] US8285619B2 In this invention, economical analysis is done based on news sentinel analysis. Natural language processing is used to parse or pattern matching from the current news. It also uses a clustering algorithm to cluster templates into groups that are correlated with the changes in stock price.
[0009] EP2755987 It reveals a neural network for stock selection that only accepts price data as input. The suggested system's core idea is to use a point-and-figure chart to construct runs (sequences) of price trends, increases, and drops, and then use the chart's maximum and lowest values to make a time-series prediction using a neural network.
[0010] INA 202241011193 In this invention, sentinel analysis is done by using machine learning. It trains a model using classification and then the prediction is performed. The input to this system is various parameters that explain the positive and negative aspects also.
[0011] INA 202241009296 In this, technical analysis is performed by machine learning. In this approach support vector machine and random forest approaches are used to evaluate data related to the stock market. The Random Forest model is an ensemble learning method that takes a classification input and generates many decision trees based on random data selections during training, then outputs the mode of classes that these decision trees have produced.
[0012] INA 202211008815 In this invention, artificial intelligence is mainly used for prediction. In AI most prominent method is Artificial Neural Network which can be used to forecast time series or the stock market. The algorithm primarily used are scaled conjugate gradient and Bayesian regularized.

DETAILED FIELD OF INVENTION

[0013] Stock selection is the process of choosing one or more stocks (or shares) based on a set of criteria in order to increase the likelihood of achieving trading or investing goal. There are over 5000 equities that can be traded or invested in. Nobody can trade or invest in multiple markets at once. As a result, a manageable quantity of stocks must be chosen.
[0014] For stock selection, the most popular tools are economical, fundamental, and technical analysis and it is also visible in figure 2. It entails evaluating or investigating subjects or concerns from the standpoint of an economist. This allows investors to examine the market from a broad perspective to specific stocks. The current market is determined by various economic parameters like Nifty and Sensex indices, GDP, Unemployment rate, inflation rate, war, economic issues etc.
[0015] Fundamental analysis helps to find the real evaluation of a company by various parameters like total net worth, revenue, debt, loan on a company, return on equity, and PE ratio. The major goal of this strategy is to find organizations that are fundamentally sound to make long-term investments in them. Fundamental analysis is again divided into two parts -Qualitative and Quantitative analysis also shown in figure 2. Qualitative Analysis considers information that cannot be expressed numerically. Its factors are company core beliefs, management, corporate governance and relation, competition in industry etc. Quantitative analysis has to do with the information found in a company's financial statements. This study aids you in determining when to buy and sell assets, as well as other investing opportunities. Its factors are company’s revenues, profit margins, return on equity, future growth potential, and financial ratios.
[0016] Technical analysis looks at previous data, primarily price and volume, to forecast future fluctuations. This analysis is based on pre-historical data. The underlying concept is that the market price already reflects the fundamentals of any given stock, hence it can be overlooked.
[0017] For performing technical analysis, there are mainly two broad categories of this prediction–statistical technique and artificial intelligence technique. The statistical technique is based on various statistical formulas. Statistician believes that knowing a market's statistics allows us to beat it and profit. When performing statistical analysis of a stock, it's much more effective to focus on the returns rather than the price.
[0018] The percentage change in the closing price between two days is referred to as the return from one day to the next.
Eq. 1

The PCT change function on series objects in python allows us to determine this number. The lag to use is the point of contention. In figure 3 we can directly use the above-mentioned formula.
[0019] It displays the relationship between the high & low and opening & closing prices of a stock. The body of the candle represents the opening and closing prices during the period. Above and below the body are vertical lines called wicks or shadows that show the lows and highs of the traded prices.
[0020] The description of key terms are as follows:
• Open: it is the first price at which stock is traded. It is represented at top or bottom of Candlestick
• Close: it is the last price at which stock is traded. It is also represented at either top or bottom of candlestick.
• High: It simply means the highest price of that specific time or interval set for that candlestick. It is represented by top of shadow which is shown above the candlestick. This shadow is also known as upper shadow. This high can be equal to open or close depending upon the cases.
• Low: It means the lowest price at which stock is traded in that specific interval. It is represented by shadow at the bottom of candlestick. this shadow is also known as bottom shadow. The low can be equal to open or close depending upon the cases.
Open close high and low is clearly mentioned in figure 4. the current state of candlestick is shown in figure 5 without using our approach.

[0021] Our proposal as a classifier uses technical indicators as input variables to forecast financial price movement tendencies across parameters open, close high and low.
[0022] Our prediction model is based on numerical ranking from 1 to 5 depending upon the pattern occurring:
Numerical Ranking 5
the numerical ranking 5 is given when following patterns occur and listed:
• Dozi:
? Body: long shadow, no body is also shown in figure 7
? Meaning: Trend Coming to end
ABS(O-C) <= 0.02*(H-L) and (H-C) <=0.3*(H-L) and ((H-L)>= (AVGH10- AVGL10)) and (H>L) And (L=MINL10)
AVGH10 = 10 days high price moving average
AVGL10= 10 days low price moving average
MINL10= Lowest low over the last 10 days
MAXH10= Highest high over last 10 days
• Hammer Candlestick: in figure 8 hammer is represented
? Body: Small Body, Long tail
? Meaning: Bullish Trend Reversal
(((H-L)>3*(O-C)AND((C-L)/(.001+H-L)>0.6)AND((O-L)/(.001+H-L)>0.6)))
Numerical Ranking 4
When 20 * ABS (O - C) <= H - L
C>O and H=C and O=L then level 4 is given it indicates upward trend .
Numerical Ranking 3
Spinning Top Candlestick:
Body: Small
Meaning: Indecision 50-50
Interpretation: Reversal
Shooting Star:
Body: Small body, long tail, looks like inverted hammer
Meaning: Trend Reversal
(((H – L) > 4 * (O – C)) AND ((H – C) / (.001 + H – L) >= 0.75) AND ((H – O) / (.001 + H – L) >= 0.75)))
Numerical Ranking 2
Piercing Line
Meaning: short term reversal
((C1 < O1) AND (((O1 + C1) / 2) < C) AND (O < C) AND (O < C1) AND (C < O1) AND ((C – O) / (.001 + (H – L)) > 0.6))
Numerical Ranking 1
Three Black Crows
Meaning: Market bearish
Interpretation: trend downwards
(O > C * 1.01) AND (O1 > C1 * 1.01) AND (O2 > C2 *1.01) AND (C < C1) AND (C1 < C2) AND (O > C1) AND (O < O1) AND
(O1 > C2) AND (O1 < O2) AND (((C – L) / (H – L)) < .2) AND
(((C1 – L1) / (H1 – L1)) < .2) AND (((C2 – L2) / (H2 – L2)) < .2)
[0021] On applying our approach we get the output as shown in figure 6.a bubble shape with number filled is generated. Based on numeric ranking decision and prediction can be done.

[0022] 000 is representing the input from stock market data which is taken. 001 means data selection operation is performed means that we must select the time frame or time interval and then apply data selection. then in 002 removal of noisy data is done to improve accuracy of the invention and reduce time execution time also. 003 means now we have applied data transformation which is required for model training in 004. After this step we are using various type of classifier 005 to predict the numerical ranking. 005 depending upon numerical ranking decision is taken to hold or buy and sell the stock.

FIGURES

Figure 1: Clearing house of NSE, BSE and XMSEI

Figure 2: Types of analysis in stock market

Figure 3: ITC stock data for five years

Figure 4: Candlestick (left for bear market and right for bull market)

Figure 5: In current state-of-candlestick pattern

Figure 6: Our proposal to current state-of-candlestick pattern

Figure 7: Dozi Candlestick

Figure 8: Hammer candlestick

Figure 9: Flowchart for proposed technique

CLAIMS

[1] A novice can learn or invest by observing a numeric ranking from 1 to 5, where the higher number represents a high recommendation for buying and selling.
[2] The prediction can be visualized on the chart in the circle shape which helps in decision making.
[3] This system will also show the appropriate price which will be used for buying and selling equity.
[4] Depending upon various parameters, it will also suggest holding stock for long-term profit.
[5] Our Classifier will also help in Mutual Fund Prediction based on this stock prediction.

April 26, 2022 Mr. Harsh Vardhan et al.


ABSTRACT

The present invention is a system and process for optimization of stock prediction in share market. Stock prediction is mainly done either by fundamental analysis or by technical analysis. Our approach is based on technical analysis in which numeric ranking method is established using the machine learning. Numeric 5 means that it is of top priority whereas 1 means low priority, the present approach works on numeric method techniques based on ranking the decision is to be taken to hold, buy or sell the stock.

April 26, 2022 Mr. Harsh Vardhan et al.

, Claims: [1] A novice can learn or invest by observing a numeric ranking from 1 to 5, where the higher number represents a high recommendation for buying and selling.
[2] The prediction can be visualized on the chart in the circle shape which helps in decision making.
[3] This system will also show the appropriate price which will be used for buying and selling equity.
[4] Depending upon various parameters, it will also suggest holding stock for long-term profit.
[5] Our Classifier will also help in Mutual Fund Prediction based on this stock prediction.