logo image

Alternative Investment markets

The Gil-Alana (2006) description of time series behaviour of stock market indexes by a fractional representation serves as the foundation for this study of weak form efficiency in the London Stock Exchange. Such a possibility could be due to long-memory or nonlinearities and has serious consequences for the efficient market hypothesis in stock markets. This study utilises the time-series of individual share prices for one hundred companies from the London Main (FTSE) and Alternative Investment (AIM) markets.

A simple Fraction Augmented Dickey-Fuller (FADF) test explores long-memory utilising a range of consistent estimates of the fractional differencing parameter d. Application of random field-based tests and traditional tests for structural breaks, examine the possibility for non-linearity. ? Weak form efficiency posits that the current share price fully reflects the information implicit in the share price history of a company.

This implies that the share price time series, say {Pit}T/t=0 for share i, does not exhibit serial correlation or “memory”. Recent literature puts forward the case for a generic random walk type model, albeit one with less stringent assumptions than those imposed in earlier literature on market efficiency. Traditionally the analysis has been whether {Pit}T/t=0 is I(1) or I(0). In the I(0) case, Pit is covariance stationary and displays the property of short-memory. In the I(1) case, Pit is non-stationary and non-mean reverting.

Fractional integration relaxes the assumption that the differencing parameter d is an integer. Therefore, a time-series process with 0 < d < 1 is defined as a long-memory process. If 0 < d < 0. 5 the series Pit is covariance stationary, but the autocorrelation function is more persistent with a longer decay period, than when d = 0. If 0. 5 ? d < 1, Pit is not covariance stationary, but it is mean reverting, with a persistent, slowly decaying autocorrelation function.

Thus, if Pit is I(d) with o < d < 1 it exhibits memory and is then not consistent with weak form efficiency. While studies focusing on the non-stationary nature of time series have dominated the literature, increasing attention is focusing on the issue of whether series that appear integrated of order < 0 are actually non-linear stationary series. Al-Loughani and Chappell suggests this non-linearity may be the cause of negative results with regards to many tests and variables.

Read full document

Can’t wait to take that assignment burden offyour shoulders?

Let us know what it is and we will show you how it can be done!
×
Sorry, but copying text is forbidden on this website. If you need this or any other sample, please register
Signup & Access Essays

Already on Businessays? Login here

No, thanks. I prefer suffering on my own
Sorry, but copying text is forbidden on this website. If you need this or any other sample register now and get a free access to all papers, carefully proofread and edited by our experts.
Sign in / Sign up
No, thanks. I prefer suffering on my own
Not quite the topic you need?
We would be happy to write it
Join and witness the magic
Service Open At All Times
|
Complete Buyer Protection
|
Plagiarism-Free Writing

Emily from Businessays

Hi there, would you like to get such a paper? How about receiving a customized one? Check it out https://goo.gl/chNgQy