I conducted two research projects during my undergraduate studies and have also accumulated 2.5 years of investment research experience whilst working as a Fixed Income and Currencies Analyst at an Investment Brokerage Firm.
My first undergraduate research project was a survey research I conducted in 2014 during my third year of undergraduate studies. The purpose of the research was to evaluate the impact of the then recently adopted International Financial Reporting Standards (IFRS) on the financial reporting quality of the listed Nigerian Banks. For the purpose of the research, I surveyed a cross section of institutional and retail investors in the listed Nigerian banks to discover if they found the new financial reporting format by the banks to be more useful for their investment decisions. The results of the survey demonstrated that the investors found the new financial statement format to be more useful, especially due to the fair value requirement of the IFRS, which required the banks to test their assets and liabilities annually for impairment, thus ensuring that they were carried in the financial statements at a value closer to their true economic values.
My second undergraduate research project was a cross sectional time-series analysis I conducted in my final year (2015), to evaluate the impact of dividend policy on the share price performance of firms listed on the Nigerian Stock Exchange. For the purpose of the research, I collected and analysed a 20-year cross sectional time-series data on the dividend, earnings and share price of 20 listed Nigerian firms. The results of the analysis demonstrated that Nigerian investors favoured companies that paid regular dividends, thus supporting the dividend relevance hypothesis.
Whilst working as a Fixed Income and Currencies Analyst, I prepared daily, weekly and monthly research reports covering the Sub Saharan Africa Fixed Income and Currencies market for institutional clients of my firm. I also contributed regularly to several local and international financial news platforms, including Bloomberg and CNBC Africa.
Proposed Research Plan
Please describe your research plan for the proposed doctoral research and include in particular detailed objectives and methodology. The description should also include the general field of the research; the specific research question(s); the preliminary research conducted and any results that may support the feasibility of the work (if applicable); the significance, originality and/or anticipated impact of the work. Please also provide references or citations, if applicable (up to 600 words).
Aims
The aim of this research is to provide a framework for the implementation of big data and machine learning strategies in the discretionary investment management industry. The study specifically aims to achieve the following objectives:
- To identify and analyze the different types of big and alternative data sets applicable in security analysis and to assess their relevance for different asset classes and discretionary investment styles.
- To assess the efficacy of the different machine learning methods used in analyzing the big and alternative data sets.
This study aims to amongst other things answer the following major questions:
- How can discretionary portfolio managers make use of alternative data in enhancing their portfolio returns?
- Which machine learning methods are most appropriate for analyzing the alternative data sets?
Methods
To fulfil the objectives of this research, several big data sets from the different alternative data categories identified (such as: social media data, news and reviews, web searches, business process data, satellite imagery, Geo-location and other sensor data types) would be sourced from different alternative data vendors and platforms and would be analyzed using the different machine learning techniques (including: Supervised Learning (regressions, classifications), Unsupervised Learning (factor analysis, clustering) as well as novel techniques of Deep and Reinforcement Learning), with the main aim of determining which of the alternative data categories and machine learning methods are most suitable for predicting future asset returns and generating alpha in discretionary portfolios.
Scope
This research study falls within the domain of Financial Economics and Data Science and the sub-domains of Security Analysis, Portfolio Management and Machine Learning.
Significance
Recently published research shows that algorithms can be highly effective when paired with skilled investment professionals, because by themselves, machines have trouble anticipating the complicated human responses of politicians and central bankers that can drive market regime changes. However, when operating within the framework of an economists’ hypothesis, algorithms can forecast expected returns with much-welcomed precision far better than traditional statistical methods, where forecasts remain deeply shrouded in approximation and estimation errors.
Also, as smartphones and devices multiply, cameras and other sensors boom, and organizations increasingly ground their business processes in data, new kinds of analysis are opening up for traders and investors to make more informed decisions about the world beyond traditional data sources like stock price activity or earnings reports.
The ‘Quantamental Investing’ approach stated in the research topic, is a new approach to investing where the results from machine learning and artificial intelligence are used by fund managers to improve fund performance. A Quantamental approach will rely heavily on machine learning and artificial intelligence to analyze vast amounts of data which would then be used by the portfolio manager to identify an attractive set of securities based on a number of fundamental criteria.
As discretionary investment funds are now coming under pressure due to under-performance relative to their passive and systematic counterparts, this research study would be of immense benefit to the discretionary investment industry, as it seeks to explore how discretionary portfolio managers can leverage the recent disruptive tools of big data and artificial intelligence to enhance their portfolio returns above their passive and systematic counterparts through a 'Quantamental Investing’ process.
This research, by adopting a machine learning methodology, would also enhance existing literature in asset pricing which have mostly focused on the application of traditional statistical techniques for forecasting future asset returns.
