A Review Paper On Pr edication Appr oach For

Stock Mar ket Volatility Based on Time Ser ies

Data

1)PRIYANSH BHALIA
2)Dr. RAJESH MISHRA

1) Student at Gautam Buddha University, Greater Noida

2) Professor of Electronics And Communication Engineering at Gautam
Buddha University, Greater Noida

This present study is a review of an Stock market predication and efficiently predict
the future stock market lows and highs using ARIMA Model.

Abstract:-
Time series analysis and forecasting is of vital significance, owing to its widespread
use in various practical domains. Time series data refers to an ordered sequence or a
set of data points that a variable takes at equal time intervals. The stock market is
considered to be one of the most highly complex financial systems which consist of
various components or stocks, the price of which fluctuate greatly with respect to time.
Stock market forecasting involves uncovering the market trends with respect to time.
All the stock market investors aim to maximize the returns over their investments and
minimize the risks associated. Stock markets being highly sensitive and susceptible to
quick changes, the main aim of stock-trend prediction is to develop new innovative
approaches to foresee the stocks that result in high profit. This research tries to
analyze the time series data of the Indian stock market and build a statistical model
that could efficiently predict the future stocks.

Introduction:-
Any form of prediction is a tough challenge with inside the actual world, specially in
which the destiny could be very dynamic. The inventory marketplace is rather risky
and unpredictable via way of means of nature. Therefore, traders are usually taking
dangers in hopes of creating a profit. People need to make investments with inside the
inventory marketplace and anticipate make the most of their investments. There are
many elements that have an impact on inventory expenses, together with deliver and
demand, marketplace trends, the worldwide economy, company results, ancient
charge, public sentiments, touchy economic information, popularity (together with
properly or terrible information associated with an organization call and product), all
of which can also additionally bring about a growth or lower with inside the quantity
of shoppers etc. Even eleven though one can also additionally examine a lot of
elements, it's miles nevertheless tough to reap a higher overall performance with
inside the inventory marketplace and to are expecting the destiny charge in general.
Predicting the charge of a selected inventory at some point in advance is, via way of
means of itself, a totally complex challenge.

A stock is a financial instrument that represents ownership in a company or

corporation and represents a proportionate claim on its assets (what it owns)
and earnings (what it generates in profits). Stocks are also called shares or a
company's equity.Stock ownership implies that the shareholder owns a slice of the
company equal to the number of shares held as a proportion of the company's
total outstanding shares. For instance, an individual or entity that owns 100,000
shares of a company with one million outstanding shares would have a 10%
ownership stake in it. Most companies have outstanding shares that run into the
millions or billions.

Today's corporate giant likely had its start as a small private entity launched by a
visionary founder a few decades ago. Think of Jack Ma incubating Alibaba
(BABA) from his apartment in Hangzhou, China, in 1999, or Mark Zuckerberg
founding the earliest version of Facebook (now Meta), from his Harvard University
dorm room in 2004. Technology giants like these have become among the biggest
companies in the world within a couple of decades.

However, growing at such a frenetic pace requires access to a massive amount

of capital. In order to make the transition from an idea germinating in an
entrepreneur's brain to an operating company, they need to lease an office or factory,
hire employees, buy equipment and raw materials, and put in place a sales
and distribution network, among other things. These resources require significant
amounts of capital, depending on the scale and scope of the business startup.

Objective :

The main objective in the paper was to develop a less complex and efficiently model.
To analyze the data of the time series in the stock market using the ARIMA Model
So that it can give best approximately solution for the outcome.So,the predicted price
using the ARIMA Model is very close to the Actual price on the given dates. This
indicates the successful implementation of the ARIMA Model on the time series data.
Related Work:-

Armstrong, J.. (1988). Research Needs in Forecasting. International Journal of

Forecasting. 3. 449-465. 10.1016/0169-2070(88)90111-2. :-In this Research
paper,Most cultures have been concerned with forecasting. Sometimes the forecaster
was held in high regard, as was the oracle at Delphi. Often, however, forecasting is
regarded as a necessary evil and is frowned upon.Forecasting is not a respectable
human activity and not worthwhile beyond the shortest of periods." Sometimes it has
been illegal. For example, in Rome in 357 A.D. Emperor Constantine issued an edict
forbidden anyone " to consult a soothsayer, a mathematician, or a forecaster.May
curiosity to foretell the future be silenced forever," In recent years, however,
forecasting seems to have become a respectable activity. Research of forecasting has
also gained respect. The Re paper examines research on forecasting. It includes all
aspects of forecasting in the social sciences.

Generally, there are many approaches to forecast demand among which we find the
exponential smoothing, for example. But, when applying these approaches, we need
to have historical data. In the beginning, there is no information about the past, we use
then an estimation based on similar cases or engineer’s experiences

Velicer, Wayne & Fava, Joseph. (2003). Time Series Analysis.

10.1002/0471264385.wei0223:-In this Study they have explained the several major
classes of research questions, including an analysis of basic processes, intervention
analysis, and analysis of the pattern of treatment effects over time.Technical aspects
of ARIMA models are described, including definitions of important terms, statistical
estimation of parameters, and the model identification process.In this the study
explained the research paper.There has been extensive research developing and
comparing procedures to aid the model identification process (Glass, Willson, &
Gottman, 1975; McCleary & Hay, 1980; Akaike, 1974; Bequin, Courieroux, &
Monfort, 1980; Bhansali & Downham, 1977; Grey, Kelly, & McIntire, 1978; Hannan
& Rissanen, 1982; Kashyap, 1977; Parzen, 1974; Pukkila, 1982; Rissanen, 1978;
1986a, 1986b; Schwartz, 1978; Tsay, 1984; Tsay & Tiao, 1984). Unfortunately there
is not yet a clear consensus on a best method for this important task.A simulation
study by Velicer and Harrop (1983) studied the model identification process. Some of
the reasons for difficulty with this step include the large number of data points
required for accurate identification, the complexity of the procedures, and problems
with accuracy and reliability of some methods, even under ideal circumstances.

Din, Marilena. (2015). ARIMA by Box Jenkins Methodology for Estimation and
Forecasting Models in Higher Education. 10.13140/RG.2.1.1259.6888.:-In this further
studies they have used many approaches in the existing literature which have
attempted to treat education enrollments with specific forecasting models such as
moving average, exponential smoothing, Markov chain, regression, Fuzzy time series,
and others, the results were not enough to understand the moving data evolution, or
not enough accuracy.
The empirical study revealed the best ARIMA validated model to be used to forecast
future values for the next eight years.Though such approach generally helps
understanding data or predicts future points, after the steps of identification,
estimation and verification followed here to build the best ARIMA model, the
findings are not providing the foresight of the causes that may influence what will
happen in the future years.

R. J. Hyndman. Forecasting Overview. Accessed: Nov. 8, 2009.

[Online].Available:https://robjhyndman.com/papers/forecastingoverview.pdf:-In this
they explained about how forecasting is required in many situation.They show how
forecasting is used in many widely areas.Forecasting situations vary widely in their
time horizons, factors determining actual outcomes, types of data patterns, and many
other aspects. Forecasting methods can be very simple such as using the most recent
observation as a forecast, or highly complex such as neural nets and econometric
systems of simultaneous equations. The choice of method depends on what data are
available and the predictability of the quantity to be forecast.

They shows it is divided into two parts Quantitative forecasting and Qualitative
forecasting, it explained the basic steps of forecasting in five steps.
All forecasting is about estimating some aspects of the conditional distribution of a
random variable that’s known as forecasting distribution.The issue of measuring the
accuracy of forecasts from different methods has been the subject of much attention.
We summarize some of the approaches here. A more thorough discussion is given by
Hyndman & Koehler (2006).
In the overall discussion they will explain the errors that exist during doing
forecasting which are listed as Scale dependent error,Percentage error and scaled
error.

J. Ranjani, A. Sheela and K. P. Meena, "Combination of NumPy, SciPy and

Matplotlib/Pylab -a good alternative methodology to MATLAB - A Comparative
analysis," 2019 1st International Conference on Innovations in Information and
Communication Technology (ICIICT), 2019, pp. 1-5, doi:
10.1109/ICIICT1.2019.8741475.:-The major goal of this study was to explain the
python libraries.As python we can develop a high performance applications. Python
often utilizes external libraries to perform scientific computing. The most important
libraries used are NumPy, SciPy and Matplotlib to perform scientific and numeric
applications. The Python libraries are open-source tag-on modules, which do
additional frequent mathematical and numerical routines in pre-compiled, high-speed
tasks.
In my model I used to plot the map of all stock market graphs for this i use
Matplotlib.Just like Matplotlib many libaries are used like Numpy,Scipy and many
more.
Pati, Jayadeep et al. “A Comparison Among ARIMA, BP-NN, and MOGA-NN for
Software Clone Evolution Prediction.” IEEE Access 5 (2017): 11841-11851.:-In the
research paper,the main goal is to explain the about how software and algorithm is
developing and gave best accuracy output day by day.Explain how Software is agile
i.e. they require continuous changes to adopt the changing requirements. This rapid
change of requirements leads to software code fragments being slightly modified or
copied to have higher functionalities and This paper presents an approach for
prediction of software clone evolution across different versions of software
applications.
In the Research paper the main aim is to show the difference between
ARIMA,BP-NN and MOGA-NN.
In this paper, we have used ARIMA as well as a Neural Network model for predicting
the evolution of software clone components. The neural network is first trained with
‘‘BFGS quasi-Newton Method’’ based Back-propagation , a standard algorithm used
for training of neural network. We have also used Multi-Objective Genetic Algorithm
based Neural Network training algorithm for improving the accuracy of the model in
prediction. The goal is to have a suitable model for prediction of software clone
evolution which will be helpful in reducing the software maintenance effort. The
detailed description of all the models is described below.
As software evolves, new code fragments are added, deleted or modified to fulfill the
desired requirements. These duplicated, or slightly modified code fragments are called
cloned components and focus is on detection of these cloned components.
The software testing and maintenance are most expensive part in the software
engineering . In this paper, they have discussed on a method for proper monitoring
and predicting the evolution of clone numbers across different versions of the open
source software applications. It is helpful in reducing the effort for software
maintenance.

Green, Shakira, "Time Series Analysis of Stock Prices Using the Box-Jenkins
Approach" (2011). Electronic Theses and Dissertations. 668:-In the research paper,the
main aim is to explain apply the box jenkins, using the BoxJenkins approach, is there
an ideal model and sampling interval when looking at modeling stock prices, with
regard to specific industries.
it is possible to forecast values based on an ARIMA (0,1,0) model, it would be not
wise to do so, especially when using the model as an investment tool. The danger in
using a random walk model to forecast is that they have no way of estimating t within
the model.then they will take the several instances to explain the how box Jenkins
method is applied in the stock and using box Jenkins method how they predict the
future outcomes.By using ACF and PACF they calculated the auto regression
Parameter(p) because with the auto-regression(p) parameter we can check whether
our data is seasonal data or nonseasonal data.Once the ACF and PACF have been
calculated and the behavior of them has been examined to determine the number of
autoregressive parameters (p), and/or the moving average parameters (q) and an
appropriate model has been selected.
For checking the model weither it is good fit for model they listed some properties
the residuals should be approximately Normal, all the parameter estimates should
have significant p-values, and the model should contain as few parameters as possible.
The last step is to use the final model to forecast future time series values.

Cesario, Eugenio & Catlett, Charles & Talia, Domenico. (2016). Forecasting Crimes
UsingAutoregressiveModels.795-802.10.1109/DASC-PICom-DataCom-CyberSciTec.
2016.138. :-In this paper the main approach is to focus on steadily increasing
urbanization, by 2030 more than sixty percent of the global population will live in
cities.this paper explain how it affect both socially and economically transformation
both side on positive as well as negative.Talking about positive side such as it
increased opportunities in urban areas.
On other side Negative,the crime is increases there is a deflection in the economy of
cities budgets.Nevertheless, new technologies are enabling police departments to
access growing volumes of crime-related data that can be analyzed to understand
patterns and trends.
The main aim to design the model to decrease the level of cybercrime for this we need
to train the model with the help of Auto regression(p).
New technologies are enabling police departments to access growing volumes of data
related to crime, that can be analyzed to understand patterns and trends. Such
knowledge is useful to anticipate criminal activity and predict/prevent crimes. This
paper presented an approach, based on auto-regressive models,for reliably forecasting
crime trends in an urban area.

Experimental Setup:-
In the paper the data set used is from NIFTY50 and BSE companies and the model is
trained on the platform Anaconda-Jupyter.The Whole model is written in Python
Language.

MODEL ARIMA ESTIMATION:-

ARIMA, short for ‘Auto Regressive Integrated Moving Average’ is actually a class
of models that ‘explains’ a given time series based on its own past values, that is, its
own lags and the lagged forecast errors, so that equation can be used to forecast
future values. Know more about parameters of ARIMA and its limitations, in this
free video tutorial
 Any ‘non-seasonal’ time series that exhibits patterns and is not a random white noise
can be modeled with ARIMA models.
An ARIMA model is characterized by 3 terms: p, d, q
where,
p is the order of the AR term
q is the order of the MA term
d is the number of differencing required to make the time series stationary
If a time series, has seasonal patterns, then you need to add seasonal terms and it
becomes SARIMA, short for ‘Seasonal ARIMA’. More on that once we finish
ARIMA.
So, what does the ‘order of AR term’ even mean? Before we go there, let’s first look
at the ‘d’ term.

Training Phase :-
There are usually five basic steps in any forecasting task which are listed below:-
Step 1: Problem definition Often this is most difficult part of forecasting. Defining the
problem carefully requires an understanding of how the forecasts will be used, who
requires the forecasts, and how the forecasting function fits within the organization
requiring the forecasts. A forecaster needs to spend time talking to everyone who will
be involved in collecting data, maintaining databases, and using the forecasts for
future planning.
Step 2: Gathering information There are always at least two kinds of information
required: (a) statistical data, and (b) the accumulated expertise of the people who
collect the data 2 and use the forecasts. Often, a difficulty will be obtaining enough
historical data to be able to fit a good statistical model. However, occasionally, very
old data will not be so useful due to changes in the system being forecast.
Step 3: Preliminary (exploratory) analysis Always start by graphing the data. Are
there consistent patterns? Is there a significant trend? Is seasonality important? Is
there evidence of the presence of business cycles? Are there any outliers in the data
that need to be explained by those with expert knowledge? How strong are the
relationships among the variables available for analysis?
Step 4: Choosing and fitting models Which model to use depends on the availability
of historical data, the strength of relationships between the forecast variable and any
explanatory variables, and the way the forecasts are to be used. It is common to
compare two or three potential models.
Step 5: Using and evaluating a forecasting model Once a model has been selected and
its parameters estimated, the model is to be used to make forecasts. The performance
of the model can only be properly evaluated after the data for the forecast period have
become available. A number of methods have been developed to help in assessing the
accuracy of forecasts
Testing Phase:-
With the help of model we have to calculate the value of Mean Square error
Root mean square error ,overall accuracy of the model

For finding the result we have to apply Arima Model.Before apply the model we have
to integrate the database of stock market.
This differencing of the time series is represented by the ‘d’ value of the ‘‘ARIMA
model’’. There are some situations that may arise with the value of ‘d’, likewise
below:

When ‘d’ = 0, it signifies that the series under consideration is stationary, so we don’t
require to take the difference of it.
If ‘d’ =1, it signifies that the current series is not stationary, we need to take the first
difference of the series.

If ‘d’ = 2, it signifies that the series under consideration has been differenced two
times.

Result and Discussion :

The main aim of the paper is to give accurate predication of stock market with the
help of past open and high values of the companies.it can follow the pattern to give
best outcomes.
For better outcomes we have done differencing at least 12 times in the base paper they
have done 6 times but after differencing.The main question arise why 12th times not
more than or less than?here is the answer because after 12th differencing it shows the
value “nan”or zero,which don’t predict the accurate value.
It’s also found that these small small changes made the huge effect in the accuracy of
the model.The accuracy of my base paper is 95% but using Sarimax our accuracy is
increased to 97%.

Reference:-

1. Armstrong, J.. (1988). Research Needs in Forecasting. International Journal of

Forecasting. 3. 449-465. 10.1016/0169-2070(88)90111-2.

2. Velicer, Wayne & Fava, Joseph. (2003). Time Series Analysis.

10.1002/0471264385.wei0223

3. Din, Marilena. (2015). ARIMA by Box Jenkins Methodology for Estimation and
Forecasting Models in Higher Education. 10.13140/RG.2.1.1259.6888.
4. R. J. Hyndman. Forecasting Overview. Accessed: Nov. 8, 2009.
[Online].Available:https://robjhyndman.com/papers/forecastingoverview.pdf

5. Pati, Jayadeep et al. “A Comparison Among ARIMA, BP-NN, and MOGA-NN

for Software Clone Evolution Prediction.” IEEE Access 5 (2017): 11841-11851

6. J. Ranjani, A. Sheela and K. P. Meena, "Combination of NumPy, SciPy and

Matplotlib/Pylab -a good alternative methodology to MATLAB - A Comparative
analysis," 2019 1st International Conference on Innovations in Information and
Communication Technology (ICIICT), 2019, pp. 1-5, doi:
10.1109/ICIICT1.2019.8741475

7. Green, Shakira, "Time Series Analysis of Stock Prices Using the Box-Jenkins
Approach" (2011). Electronic Theses and Dissertations. 668.
https://digitalcommons.georgiasouthern.edu/etd/668

8. Cesario, Eugenio & Catlett, Charles & Talia, Domenico. (2016). Forecasting
CrimesUsingAutoregressiveModels.795-802.10.1109/DASC-PICom-DataCom-C
yberSciTec.2016.138.
9. Adhikari, Ratnadip & Agrawal, R.. (2013). An Introductory Study on Time series
Modeling and Forecasting. 10.13140/2.1.2771.8084.
10. Palanisamy, Karthika & Parthasarathy, Karthikeyan. (2016). Stock Returns and
Volatility: Evidence from Nifty 50 Index. Asian Journal of Research in Social
Sciences and Humanities. 6. 434. 10.5958/2249-7315.2016.00127.1.
11. Manh, Ha & Duong, & Siliverstovs, Boriss & Manh, & Duong, Ha. (2006). The
Stock Market and Investment.
12. Meek, Christopher & Chickering, David & Heckerman, David. (2002).
Autoregressive Tree Models for Time-Series Analysis.
10.1137/1.9781611972726.14.