open access

Journal of Mathematics, Physics and Mechanics

P e e r - R e v i e w e d  
Annual (One issue per year)
Comparative Analysis of Correlation-Driven and Statistical Time Series Models for Financial Forecasting
Research Article - Volume: 1, Issue: 1, 2026 (September)
Comparative Analysis of Correlation-Driven and Statistical Time Series Models for Financial Forecasting Download PDF

*Correspondence to: Habibu Muhammad Haris, Department of Mathematics, Mewar University, Chittorgarh, Rajasthan, India, E-mail:

Received: July 27, 2026; Manuscript No: JMPM-26-5600; Editor Assigned: July 31, 2026; PreQc No: JMPM-26-5600 (PQ); Reviewed: August 13, 2026; Revised: August 25, 2026; Manuscript No: JMPM-26-5600 (R); Published: September 29, 2026

ABSTRACT

Financial time series forecasting is important for risk management, investment decision-making, and portfolio allocation. This study com- pares correlation-based analysis with classical statistical time-series modelling using daily NIFTY 50 data from January 2021 to December 2025. Pearson correlation analysis is used to examine contemporaneous relationships among key market variables, while logarithmic returns are analysed using stationarity tests, ACF/PACF analysis, and ARIMA modelling. Candidate ARIMA models are evaluated using AIC, AICc, BIC, and residual diagnostic tests. To assess predictive performance objectively, observations from 2021–2024 are used for model estimation, while 2025 is reserved for genuine out-of-sample one-step-ahead forecasting. The results show that correlation analysis provides useful information on market co-movement but does not itself constitute a forecasting model. The return series exhibits stationarity in the conditional mean, together with heavy tails and volatility-related dependence. ARIMA captures limited linear dependence in returns, but its forecasting performance remains constrained by un-modelled volatility dynamics. The findings support the use of ARIMA as an interpretable benchmark and highlight the potential value of volatility-aware and hybrid models for improved financial forecasting.

Keywords: Financial Time Series; NIFTY 50; Correlation Analysis; ARIMA; Forecasting; Stationarity; Volatility

Citation: Haris HM, Taambu AS, Mehta P, Prajapat K, Basiru IT, Mahmud MD (2026). Comparative Analysis of Correlation-Driven and Statistical Time Series Models for Financial Forecasting. J. Math. Phys. Mech. Vol.1 Iss.1, September (2026), pp:63-81.
Copyright: © 2026 Habibu Muhammad Haris, Aminu Salisu Taambu, Pramod Mehta, Komal Prajapat, Isah Tasiu Basiru, Mahmud Dalha Mahmud. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.