Electronic Theses and Dissertations (PhDs)

Permanent URI for this collection

https://hdl.handle.net/10539/37943

Browse

Browsing Electronic Theses and Dissertations (PhDs) by Keyword "Asymptotic PCA"

Now showing 1 - 1 of 1
  • Thumbnail Image
    Item
    Towards a framework for asset pricing in developing equity markets
    (University of the Witwatersrand, Johannesburg, 2021) Ouma, Wycliffe Nduga; Kodongo, Odongo
    The need to accurately determine risk-return trade-offs in financial markets is an important question that has occupied minds of various players in capital markets for over five decades now. In this study, we addressed this question by attempting to determine the drivers of returns in developing equity markets. Further, we test the ability of the identified factors to command significant and reasonable risk premiums in the returns of developing equity markets. The results of this analysis were compared with those of a select group of emerging and developed equity markets. This study employed monthly stock price data from 20 developing equity markets, starting January 1, 1996, and ending February 1, 2020, resulting in 290 months of observation. Countries are included in the study as their data became available. Monthly stock price data were used to calculate individual stock returns for the first phase of the study, which involved determining the covariance matrix of returns. This was done through a dimensionality reduction technique of Asymptotic Principal Components Analysis and standard PCA whenever data permitted. The underlying drivers of returns variations determined through these procedures resulted in a few significant principal components (PCs) driving overall stock variations. The next step related the identified PCs to the candidate risk factors ensuring that empirical testing of asset pricing factors only included factors with important influence on stock variations. The study has found important empirical results on asset pricing in developing equity markets. First, the study has established that only a few fundamental drivers influence returns in the sampled 20 developing equity markets. At country level, factor identification strategy employed discovered that only a few principal components were significantly related to the covariance matrix of returns. Canonical correlations analysis largely confirmed the results as factors such as excess market returns, book-to-equity and aggregate volatility significantly influenced returns variations in most of the individual markets in the sample. The empirical evaluation of the identified factors further established that excess market returns (MKT), book-to-market (value) factor (HML), profitability factor (RMW), momentum factor (UMD), unanticipated inflation factor (UI), aggregate volatility (VOL), and trade weighted US dollar index (TW$) were significantly priced in the returns of developing equity markets. Although FM regression did not price UI and TW$, the associated 𝑡 −statistic of their coefficients were closer to the threshold value of 2. For instance, UI and TW$ respectively produced 𝑡 −stat 1.952 and 1.929, which are significant at the 10% level. The GMM analysis, used for robustness checks, confirmed that MKT, HML, UMD and VOL were significantly priced, but further found that both UI and TW$ were not only significant drivers of risk variations, but also priced in the returns of developing equity markets Further, DS-LASSO was used to check the robustness of the entire system starting from the factor identification to testing the identified factors. The results largely corroborate the conclusions of GMM, that MKT, HML, UMD and TW$ significantly explain returns variations of frontier equity style portfolios and are priced. The results also indicated that UI is marginally priced with a 𝑡 −stat of 1.958. This study puts forward some recommendations. First, we recommend a favourable policy environment to accelerate capital market development and investment in developing countries. Since accounting information has been established to proxy for common pervasive risk drivers in stock markets, financial reporting standard is, therefore, crucial to the quality accounting information disseminated. Developing equity markets need to institute prudent and harmonised accounting practices and strong corporate governance systems to ensure quality reporting. Debt variables such as corporate bonds and government bonds form a significant component of capital market investment and are important drivers of returns in these stock markets. Unfortunately, developing equity markets do not have well-functioning debt markets rendering the variables irrelevant in their risk-return equation. Thus, these countries should institute deliberate policy measures to ensure growth of debt markets