Analyzing the Efficient Market Hypothesis with the Structural Break and Nonlinear Unit Root Tests: An Application on Borsa Istanbul
Müge ÖzdemirThe literature on finance defines the concept of an efficient market as a market where information about securities is instantly reflected in prices. Investors who trade on efficient markets cannot obtain abnormal returns. All market participants are assumed to have instant access to the information coming to the market, with everyone knowing the same information. The aim of this study is to test within the framework of information efficiency the weak-form efficiency market hypothesis using statistical tests with daily, weekly, and monthly frequencies for the BIST100 index over the period of October 2011-October 2021. The study uses the runs test, variance ratio test, unit root tests, structural break unit root tests, and nonlinear unit root tests to test weak form efficiency for this period. According to the findings, the BIST100 market index has been concluded to show a random walk at all frequencies in the relevant period (i.e., it has weak form efficiency).
Etkin Piyasa Hipotezinin Yapısal Kırılmalı ve Doğrusal Olmayan Birim Kök Testleri ile Analizi: Borsa İstanbul Üzerine Bir Uygulama
Müge ÖzdemirFinans literatüründe menkul kıymetlerle ilgili bilgilerin anında fiyatlara yansıdığı piyasa kavramı etkin piyasa olarak tanımlanmaktadır. Etkin bir piyasada işlem yapan yatırımcılar anormal getiri elde edememektedir. Tüm piyasa katılımcılarının piyasaya gelen bilgilere anında ulaştığı ve herkesin aynı bilgiyi bildiği varsayılmaktadır. Bilgi etkinliği çerçevesinde, bu çalışmanın amacı, piyasa türlerinden biri olan zayıf formda etkin piyasa hipotezini BİST100 endeksi için 2011:10-2021:10 döneminde günlük, haftalık ve aylık frekanslarda istatistiki testler ile test etmektir. Çalışmada, söz konusu döneme ait verilerin zayıf formda etkinlik sınaması için runs testi, varyans oran testi, geleneksel birim kök testleri, yapısal kırılmalı birim kök testleri ve doğrusal olmayan birim kök testi kullanılmaktadır. Çıkan bulgulara göre, ilgili dönemde BİST100 piyasa endeksininin tüm frekanslarda rassal yürüyüş gösterdiği yani zayıf formda etkin olduğu sonucuna varılmaktadır.
Fama (1970)’s efficient market hypothesis argues all the information that occurs in this market to be instantly and correctly reflected in the prices of financial assets, with no investor able to obtain abnormal returns. According to this view, no asymmetric use of information occurs, and all research on efficient market theory is concerned with whether prices fully reflect certain subsets of information. Market efficiency is a theory that has been the subject of much debate and study in the finance literature.
This study analyzes market efficiency in accordance with the information taxonomy, which is categorized into three increasingly inclusive subsets, with the first considering the weak form of market efficiency, in which the information reflected in the prices consists of already observed historical prices. In this market type, an investor investing in the market is argued to be unable to expect abnormal returns in the future using historical data because the prices already reflect all that information, and it doesn’t provide a framework for any systematic abnormal returns. The second taxonomy considers the semi-strong form of market efficiency, in which prices reflect all past and current available information. As such, the information set is extended by including things such as current firm announcements and analysis reports into the precious information set. The third taxonomy conducts the strongform efficiency test to statistically prove or disprove whether investors with access to private or insider information are able to systematically achieve abnormal returns.
This study will test weak form efficiency by analyzing whether the BIST1000 stock market index complies with the random walk process. The data range covers the period of October 2011-October 2021, with the data being composed of the closing index values at daily, weekly, and monthly frequencies.
This study conducts the runs test, variance ratio test, traditional linear unit root test, structural break unit root test, nonlinear unit root test in order to test whether the series complies with the random walk hypothesis. The linear unit root and stationarity tests include the augmented Dickey-Fuller (1981), Phillips-Perron (1988), and Kwiatkowski–Phillips–Schmidt–Shin (KPSS, 1992) tests, while the unit root tests with structural breaks include the the Perron (1989), Zivot-Andrews (1992), LeeStrazicich (2003), and Kapetanios et al. (2003) nonlinear unit root tests. The runs test deals with whether the residuals are positive or negative, with the assumption that the residuals are independently distributed. The variance ratio test is used to test whether a random walk occurs in a series that is robust in terms of both constant variance and non-constant variance. The study uses the programs RATS (Regression Analysis of Time Series), EViews, and Stata for the tests.
According to the results, the price series at each frequency contain unit roots, and the variance ratio and runs tests support the unpredictability of the returns. As a result, analysis of the Bist100 index at daily, weekly, and monthly frequencies reveal the market to have a structure that supports the weak form of the efficient market hypothesis. Although this study found no difference in the results regarding different frequencies, no surprise would be had if divergent results were observed even for the same time series, as has been discussed in the literature. The ups and downs within each month were not observable due to the closing daily observation at the end of month being recorded as a monthly BIST100 data point. For example, the Perron test allows for a unit root with a structural break and revealed the BIST100 to exhibit its lowest reading as well as a structural break on March 23, 2020, which also coincides with the official start of the COVID-19 pandemic in Turkey.
For this reason, having individuals deal with the frequencies of the price series according to the maturity structure they have planned while investing would be a good decision. In this way, investors will be able to measure return and risk more accurately in the portfolios they have created. Investors who are unable to find a profitable opportunity in an efficient market can choose their investment preferences using an analysis that will detect the deviations in the price series with regard to their shortterm investment strategies. Checking the dynamics of international stock markets as well as BIST100 stocks would be additionally useful. Moreover, this study believes that future research should focus on nonlinear situations in the stock markets of other developing countries in order to better understand the extent to which information plays a role in portfolio strategies. The fact that investors in stock markets that do not comply with the random walk process and therefore do not provide the efficient market hypothesis are able to benefit from profitable opportunities in these markets through the arbitrage process is also worth mentioning. In summary, the results show that investors are not likely to systematically make profits using historical information on the BIST100; therefore, investors should base their investment strategies on this assumption.