A study of the weak-form efficiency of the Philippine stock market using artificial neural networks

Date of Publication

2012

Document Type

Dissertation

Degree Name

Doctor of Business Administration

Subject Categories

Business Administration, Management, and Operations | Technology and Innovation

College

Ramon V. Del Rosario College of Business

Department/Unit

Management and Organization

Abstract/Summary

This study employed artificial neural networks (ANN) to test the weak-form efficiency hypothesis in the context of the Philippine stock exchange index (PSEi) and the seven sector indices. It is a common expectation among researchers and practitioners that stock markets of emerging economies exhibit inefficiencies while the opposite is true of developed markets. Unlike some previous studies that utilized the augmented Dickey-Fuller (ADF) tests, runs tests and variance-ratio tests to test the weak-form efficiency hypothesis, this study used a model comparison approach utilizing the RMSE and MAE statistics to compare out-of-sample forecasts of the ANN models to their counterpart random walk or Naive models for the PSEi and the seven sector indices. Significant out performance of one model vis-a -vis another is validated using forecast encompassing tests. The study proved, once again, the weak-form efficiency of the Philippine stock market using more recent data over a longer period of time. It also resolved the mixed findings from earlier empirical studies of the Philippine stock market concerning the weak-form market efficiency hypothesis. Furthermore, this study showed that both the ANN and ARIMA methodologies failed to encompass the random walk models for the PSEi and the sector indices.

Abstract Format

html

Language

English

Format

Electronic

Accession Number

CDTG005277

Shelf Location

Archives, The Learning Commons, 12F Henry Sy Sr. Hall

Physical Description

1 computer optical disc ; 4 3/4 in.

Keywords

Stock exchanges--Philippines; Neural networks (Computer science)

This document is currently not available here.

Share

COinS