Stock Selection, Visualization, and Forecasting
Ayush Pareek
22BDS011
Financial Data Analytics
Abstract—This project aims to analyze the stock performance • Relative Strength Index (RSI): Measures whether a
of various companies listed on the NIFTY 50 index. A set of stock is overbought or oversold, indicating potential buy
stocks was selected, their historical performance explored, key or sell signals when values exceed 70 or fall below 30,
financial ratios analyzed, and predictive models used to forecast
stock prices for the next month. Various statistical and technical respectively.
approaches were combined to select, visualize, and predict stock
prices, assisting in better investment decisions. III. S TOCK S ELECTION AND A NALYSIS
A. Exploration of Financial Ratios
I. I NTRODUCTION
The analysis of financial ratios provided an in-depth under-
Stock market analysis involves a combination of financial standing of each stock’s financial health. These ratios served
ratios, market data, and statistical models. This project focuses as filters for selecting stocks with solid fundamentals:
on understanding the performance of NIFTY 50-listed stocks
• P/E Ratio: Stocks with low P/E ratios relative to industry
through stock price analysis, financial ratio evaluation, and
peers were considered undervalued.
predictive modeling. The overall objective is to create a robust
• P/B Ratio: Stocks with a low P/B ratio were considered
framework that helps in predicting the price movements of
for value investing.
these stocks, thus assisting investors in making informed
• Dividend Yield: High dividend yield stocks were se-
decisions. The analysis follows two phases:
lected for generating passive income.
• Stock Selection: Key financial metrics such as P/E ratio,
• EPS Growth: Stocks with consistent EPS growth were
P/B ratio, dividend yield, and EPS growth are employed prioritized for their stability and potential for profitability.
to select potential stocks.
• Prediction: Time series models like ARIMA are applied
to predict stock prices, enabling short-term forecasts for
better trading decisions.
II. S TOCK P RICE P ERFORMANCE A NALYSIS
After selecting stocks, historical price data was used to
evaluate the movement of stock prices over a specific period.
This analysis was complemented with technical indicators that
provide insights into market trends and potential reversals.
A. Stock Price Movement
Stock price movement analysis involved the visualization
of closing prices over the past year. By analyzing trends,
we identified upward and downward movements, along with
periods of high volatility, allowing for better market timing
decisions.
B. Technical Indicators Fig. 1. Financial Ratios: P/E, P/B, Dividend Yield, and EPS Growth
Various technical indicators were calculated to provide
additional insights into stock trends: IV. S TOCK S ELECTION P ROCESS
• Simple Moving Average (SMA): Used to analyze long-
term trends by averaging stock prices over specific peri- The stocks were divided into three categories based on their
ods, such as 20-day and 50-day periods. market capitalization:
• Exponential Moving Average (EMA): A weighted ver- • Large-Cap Stocks (60%): These stocks are relatively
sion of the moving average, which places more impor- stable with lower volatility, ideal for risk-averse investors.
tance on recent prices, useful for tracking shorter-term – TITAN.NS, TCS.NS, DIVISLAB.NS,
trends (Fig 1 and 2). ASIANPAINT.NS, ULTRACEMCO.NS,
JSWSTEEL.NS, TECHM.NS, HDFCLIFE.NS, R EFERENCES
SUNPHARMA.NS [1] G. E. P. Box and G. M. Jenkins, Time Series Analysis: Forecasting and
• Mid-Cap Stocks (30%): These stocks offer a balance Control, Wiley, 2015.
[2] Moneycontrol, ”Stock Market News, Financial News, Market
between growth and stability. Analysis, and Investment Tools,” 2025. [Online]. Available:
– BAJFINANCE.NS, M&M.NS, MARUTI.NS, https://www.moneycontrol.com. [Accessed: Jan. 22, 2025].
[3] Overleaf, ”Overleaf: Online LaTeX Editor,” 2025. [Online]. Available:
ITC.NS https://www.overleaf.com. [Accessed: Jan. 22, 2025].
• Small-Cap Stocks (10%): These stocks are high-risk,
high-reward, selected for their growth potential.
– SHREECEM.NS, Swiggy
V. S TOCK P RICE P REDICTION FOR THE N EXT M ONTH
To forecast future stock prices, an ARIMA model was im-
plemented. The ARIMA model was trained on historical data,
capturing trends, seasonality, and noise to predict future prices.
The performance of the model was evaluated by comparing
predicted values with actual values, which demonstrated the
model’s predictive capabilities.
Fig. 2. Stock Price Prediction for the Next Month
A. Model Performance Evaluation
The model’s performance was assessed using several met-
rics, such as Mean Absolute Error (MAE), Root Mean Squared
Error (RMSE), and R-squared. The ARIMA model was found
to perform well for short-term predictions, but additional
techniques like LSTM (Long Short-Term Memory) networks
could improve performance for longer-term forecasts.
VI. C ONCLUSION
This project demonstrates a comprehensive approach to
stock analysis, combining financial ratios, technical indicators,
and predictive modeling. The ARIMA model provided valu-
able insights into stock price trends, though further refinement
through machine learning models like LSTM could improve
prediction accuracy. The results suggest that a blend of fun-
damental and technical analysis, combined with predictive
modeling, can provide actionable insights for investors.
