In this comprehensive time series analysis of Procter & Gamble's stock prices, we employed various statistical models and techniques to forecast future performance. Leveraging historical stock data obtained from Yahoo! Finance, we explored different approaches including linear trend and seasonal models, Holt-Winter's exponential smoothing, ARIMA models, and auto ARIMA. Each model provided valuable insights into the underlying patterns and trends in P&G's stock prices and offered predictions for the upcoming period.
Key Findings: Linear Trend and Seasonal Model: Provided insights into overall trend and seasonal fluctuations, but its accuracy was limited.
Holt-Winter's Exponential Smoothing: Offered a sophisticated approach capturing trend, seasonality, and volatility, yielding promising results.
ARIMA Models: Explored different ARIMA configurations, such as ARIMA(2,1,2) and auto ARIMA, demonstrating reasonable forecasting performance.
Regression Models: Investigated regression models with linear, quadratic trend, and seasonality, showing varied performance levels.