An alternative perspective on investing in mining stocks, credit bonds and statistical arbitrage strategies
Citation:LAZZARINO, MARCO, An alternative perspective on investing in mining stocks, credit bonds and statistical arbitrage strategies, Trinity College Dublin.School of Business.BUSINESS, 2018
01. Thesis_20180524 - Deposit.pdf (PDF) 2.201Mb
This thesis is a collection of three distinct essays providing advice to investors in three areas of Finance. The first investigates the sensitivity of mining stocks to metals using multifactor models. The second researches value investing opportunities in credit markets using an analysis of corporates' fundamentals. The third discusses statistical arbitrage, a common financial term for which there is still no generally accepted definition in the literature. In the first study, I investigate the sensitivity of world mining stocks to both precious and industrial metals by adding a metal factor to the CAPM and Fama-French models. I analyse all investible mining firms (421 in total) domiciled in both developed and emerging markets during the period 1990 to 2015. I enhance existing research to include all metals and provide an original comparison of mining stocks' sensitivities across both precious and industrial metals. I use both panel data and time series regressions on equal and value weighted portfolios. I find that metals are fundamental in explaining mining stocks' returns and more influential than Fama-French factors. I also find that metals are more significant for stocks of precious than industrial metals and the effect is stronger for firms domiciled in developed markets. The market factor is more relevant for industrial metal stocks. My results suggest that investors should treat mining stocks differently to other stocks and should also distinguish between precious and industrial mining firms when investing. In the second study, I investigate value opportunities in credit markets across two main geographical areas (U.S.A. and Euro Zone) and using two alternative ratings (Investment Grade and High Yield). By combining spreads with fundamental measures, I create two ratios: the spread over leverage (SL) and the spread over leverage and the reciprocal of interest coverage (SLC). The higher these ratios, the higher is the spread investors receive per unit of leverage and interest coverage. In this sense, SL and SLC normalize spreads by credit quality. I use spreads, SL and SLC ratios as indicators of value opportunities for credit markets in the same way price-earnings and price-to-book ratios are used in equity markets. In particular, I analyse the returns from investing in bonds categorized into quintiles based on various value indicators. I find that average returns are higher when spreads are in the higher quintiles and the effect is stronger over longer time horizons (three to five years). These value strategies perform better if based on SL and SLC ratios than on spreads but the outperformance is not statistically significant. My results indicate that SL and SLC ratios work as well as spreads in identifying value opportunities but can enhance spreads by detecting value opportunities also within investment grade bonds of different ratings. This suggests that researchers should further investigate the use of SL and SLC ratios in building value factors for credit markets. In the final study I investigate Statistical Arbitrage (SA). This is a common financial term for which, however, there is no common definition in the literature while investors use the expression SA for a variety of different strategies. I analyse SA strategies across equity, fixed income and, for the first time, commodity. In total, I review 165 articles on the subject, published between 1995 and 2016. The analysis of strategies' key features indicates that no existing definition fully describes them. To bridge this gap, I identify a general definition and propose a classification system that encompasses the current forms of SA strategies while facilitating the inclusion of new types as they emerge. While distinct, the three topics are connected and show some significant gaps in literature. Factor investing is becoming increasingly popular in the financial industry, with a broad and expanding academic literature. However, existing research focuses on the wider equity market with limited attention to industry specific factors. By studying mining stocks, my first study enhances the literature on factor models for specialized industry groups, which are ever more important in increasingly integrated financial markets. Factors are based on the observations of market phenomena such as the growth and value effects. However, researchers mainly analyse stocks, with fewer studies dedicated to other asset classes and particularly credit markets. My second study expands existing research on value investing for corporate bonds by using both market and accounting measures. In recent years, investors are turning their attention to factor investing and its statistical arbitrage implementations. Statistical arbitrage has been broadly discussed in literature. However researchers either focus on theoretical aspects or on developing and testing investment strategies for selected asset classes. My final study reconciles these two areas of research with a comprehensive investigation of statistical arbitrage strategies across financial literature and asset classes. This thesis makes several contributions to the academic literature on three connected and topical issues of Finance. In the first essay, I extend the literature on factor models to all investable mining firms of both precious and industrial metals. My analysis provides an investigation of metals and Fama-French factors with a novel comparison of stocks of precious and industrial metals. In the second essay, I study the existence of a value effect by originally combining spreads and fundamentals. My investigation also provides an historical analysis and comparison of the value effect over different investment horizons. In the third essay, I provide an innovative analysis of statistical arbitrage connecting academic and financial industry research. I review statistical arbitrage across all asset classes (equity, fixed income and credit) introducing a new definition and classification system. My findings are of practical benefit to analysts and portfolio managers. In the first study I find that investors should treat mining stocks differently from the broader market given their specific sensitivity to metals and Fama-French factors. My results also suggest that investors should differentiate between mining stocks of precious and industrial metals. In the second study, my findings suggest that investors should be aware that the value effect requires longer time horizons to be effective and might have significant drawdowns. My third study brings clarity to statistical arbitrage investing and allows investors to have a common framework to assess existing and emerging investment opportunities.
Author: LAZZARINO, MARCO
Publisher:Trinity College Dublin. School of Business. Discipline of Business & Administrative Studies
Type of material:Thesis
Availability:Full text available