Probability Distributions of Hidden Markov Models for Stock Closing Prices

Authors

  • Tirupathi Rao Padi, Sarode Rekha, Gulbadin Farooq Dar

DOI:

https://doi.org/10.17762/msea.v71i4.1337

Abstract

This study is on developing a Hidden Markov Model (HMM) for a daily closing price of the National Stock Exchange (NSE) and Industrial Credit and Investment Corporation of India (ICICI) bank. The changing states of NSE are considered hidden and the influencing states of change in the visible states of ICICI bank. Obtained the parameters of HMM namely Initial Probability Vector (IPV), Transition Probability Matrix (TPM) and Observed Probability Matrix (OPM) by assuming the discrete Markov chains among (i) within hidden states (Gain & Loss in NSE) and (ii) between hidden and observed states (Rise & Fall) respectively. Separate probability distributions for a sequence of n days’ transits on Rise & Fall are formulated. The behaviours of Rise and Fall states in the closing prices of ICICI bank are explored through explicit mathematical relations of different statistical measures and the related Pearson’s coefficients. Numerical illustrations are considered for a proper understanding of the developed models. These models will be useful for short-term business investors for getting the indicators on when to sell and when to purchase by observing the chances of emission states. Computer automation will make this tool more user-friendly in understanding the day-to-day share market contexts.

Downloads

Published

2022-12-28

How to Cite

Gulbadin Farooq Dar, T. R. P. S. R. . (2022). Probability Distributions of Hidden Markov Models for Stock Closing Prices. Mathematical Statistician and Engineering Applications, 71(4), 7174–7198. https://doi.org/10.17762/msea.v71i4.1337

Issue

Section

Articles