Time Series Analysis of Bank Data

This repository contains an R script for performing time series analysis on bank loan data. The analysis includes model estimation, stationarity testing, parameter estimation, residual analysis, and forecasting.

Table of Contents

• Installation
• Usage
• Script Overview
  • Step 1: Install Packages and Import Dataset
  • Step 2: Model Estimation
  • Step 3: Test for Stationarity
  • Step 4: Parameter Estimation
  • Step 5: Fitted Values
  • Step 6: Residual Analysis
  • Step 7: Forecasting
• Notes
• License

Installation

To run this script, you need to have R installed on your system. Additionally, you need to install the following R packages:

• tseries
• forecast

You can install these packages using the following commands in R:

install.packages("tseries")
install.packages("forecast")

Usage

1. Clone the repository or download the script.

2. Ensure you have installed the required packages:

   install.packages("tseries")
   install.packages("forecast")

3. Download the dataset and place it in the specified location:

   Make sure the bank_case.txt file is located at C:\\Users\\hakhamanesh\\Downloads\\.

4. Run the script:

   source('path_to_your_script.R')

Script Overview

Step 1: Install Packages and Import Dataset

The script installs the necessary packages and imports the dataset as a time series object.

install.packages("tseries")
install.packages("forecast")

library(tseries)
library(forecast)

bank_case <- as.ts(scan("C:\\Users\\hakhamanesh\\Downloads\\bank_case.txt"))
print(bank_case)

Step 2: Model Estimation

The script plots the time series data, autocorrelation function (ACF), and partial autocorrelation function (PACF).

par(mfrow=c(1,3))
plot(bank_case, main = "Time Series of Loans")
acf(bank_case, main = "ACF of Loans", lag.max = 10, ylim = c(-1,1))
pacf(bank_case, main = "PACF of Loans", lag.max = 10, ylim = c(-1,1))

Step 3: Test for Stationarity

The script tests the time series for stationarity using the Augmented Dickey-Fuller (ADF) test and performs differencing if necessary.

adf.test(bank_case)
bank_case_d1 <- diff(bank_case)
adf.test(bank_case_d1)
bank_case_d2 <- diff(bank_case_d1)
adf.test(bank_case_d2)
par(mfrow=c(1,3))
plot(bank_case_d2, main = "SOD Time Series of Loans")
acf(bank_case_d2, main = "ACF of SOD Loans", lag.max = 10, ylim = c(-1,1))
pacf(bank_case_d2, main = "PACF of SOD Loans", lag.max = 10, ylim=c(-1,1))

Step 4: Parameter Estimation

The script estimates the parameters of the ARIMA model.

bank_fit <- arima(x=bank_case, order = c(0,2,1))
bank_fit

Step 5: Fitted Values

The script calculates and prints the fitted values of the ARIMA model.

fitted(bank_fit)

Step 6: Residual Analysis

The script performs residual analysis to check the adequacy of the model.

par(mfrow = c(1,3))
plot(bank_fit$residuals, ylab = "Residuals")
acf(bank_fit$residuals, ylim = c(-1,1))
pacf(bank_fit$residuals, ylim = c(-1,1))

checkresiduals(bank_fit)

Step 7: Forecasting

The script forecasts future values using the ARIMA model and plots the forecast.

bank_pred <- forecast(bank_fit, h=24)
bank_pred
par(mfrow = c(1,1))
plot(bank_pred)

Notes

• Ensure that the dataset path in the script matches the location of your dataset.
• The script assumes the dataset is a univariate time series stored in a text file.

License

This project is licensed under the MIT License. See the LICENSE file for details.