Analysis Of Stock Market Price Behaviour: A Markov Chain Approach

Research Article
Aparna Bairagi and Sarat CH. Kakaty
Markov chain, Expected number of visits, Steady state probabilities, Expected first reaching time.

Stock market prediction has proved to be of vital importance in the present day economic scenario and stochastic process can be very effectively applied in forecasting the market trend. This paper attempts to analyse the behaviours of stock price of State Bank of India, one of the leading commercial bank of India for 1035 days covering the period from 21st March 2011 to 20th March 2015. The secondary data on daily closing price of shares for are collected from Historical price of share– Yahoo Finance. In order to meet the objectives of the paper, the investigators will propose to find out the long term behaviour of the share, expected number of visits to a particular state, and expected first reaching time of different states. The Markov Chain model is applied to analyse and predict the stock behaviour considering three different states, ‘up’– when the share price increase, ‘down’– when the share price decrease and ‘remain same’– when share price gets unchanged. The Markov Chain model is a probability model based on transition probability matrix and initial state vector. By observing the number of transitions from one state to another, the transition probability matrix has been obtained. The study reveals that regardless of bank’s current share price steady state probabilities of share ‘up’, ‘down’ and ‘remain same’ for SBI are 46.99%, 49.81% and 3.19% respectively. It is observed that if the closing value of SBI share is in the state ‘up’ in the day one then it can be expected to return to the state ‘up’ for the first time at the third day.