Fahiz Baba-Yara started his PhD journey in Economics and Finance in 2015. His research focus on finance, specifically how machine learning models help us make better this decision and how we can improve the forecasting ability of these machine learning models. Fahiz is now heading forward to Indiana University as Assistant Professor from next semester. This is our first placement in a top 20 business school in the USA. Congratulations Fahiz!

‍

Fahiz's research interest lies at the intersection of Return Predictability, Machine Learning, and Financial Econometrics. He is particularly interested in answering the question of whether or not portfolios that leverage our improved ability to estimate future returns are priced by existing asset pricing models.

‍

Read the full interview below:

‍

Why did you start this Ph.D. program at Nova SBE?

Since purchasing my first stock in high school, I have been fascinated by how the typical investor decides which stock to include and exclude from their portfolio. This interest informed both my choice of a bachelor's and master's degree.

However, it was in deciding which Ph.D. program to pursue that I finally elected to make this interest central to my studies.

I joined the Ph.D. program at Nova SBE following a recommendation from my masters' thesis supervisor, Prof. Michael Kisser. He advised that finding a prospective Ph.D. supervisor who shares an interest in what you want to work on is crucial in having a successful Ph.D. I was lucky enough to find such a supervisor at Nova, Prof. Martijn Boons. At the time, Prof. Martijn was working on a paper related to the commodity futures market, which was also my masters' thesis topic. Additionally, Nova SBE is one of the strongest finance departments in Europe in terms of publishing in top academic journals. Thus getting a Ph.D. at Nova puts one in the right environment that should mold you into a researcher capable of publishing in similar journals.

‍

What was the most valuable experience during your Ph.D.?

My most valuable experience during the Ph.D. will have to be the friendships I formed during this period. Although the Ph.D. can seem depressing at times, I was lucky enough to have very social colleagues who went out of their way to make each interaction a fantastic experience. I continue to look forward to each lunch break, not because of the food but because of the conversations.

Other exciting experiences include presenting my work to colleagues, visiting LSE as a visiting researcher, and getting the chance to be the lead instructor in a master's course.

‍

In looking back, did Nova SBE provide you with enough knowledge and skills for entering the job market?

The toolset Nova provided was very useful in helping prepare me for the job market. We were made aware of the academic job market as early as the first year. This helped a lot in setting expectations and putting things in perspective from the get-go.

Once I settled on a research question for my job market paper, I received many valuable comments from faculty and other Ph.D. colleagues, which helped shape the paper.

Finally, I also got the necessary practice that helped me compete with students from top schools on the market. This involved practice interview sessions with Professors who partake in recruitments on the academic job market.

‍

What would you say to students starting their Ph.D.?

I would advise them to be true to themselves and find topics they really care about. Research involves contemplation and thinking through the same issues over and over. Hence, working on topics that you actually care about is important because it helps keep you motivated.

Also, they shouldn't pidgin-hole themselves into a particular topic too early. Although you may start the Ph.D. with a clear idea of what topic you want to work on, be open to other ideas and engage with the literature much more broadly. There might be strands of the literature slightly outside your field that you may find fascinating once you are exposed to.

‍

What is your research about?

My research is broadly concerned with how we can use information observed today to decide which companies to include and/or exclude from our stock portfolios.

In my job market paper, I ask specifically whether we can use machine learning models to help us better make this decision. Additionally, I show that we can improve the forecasting ability of these machine learning models by allowing results from the economic literature to guide us in designing these models.

‍

How can the results of your research be applied to everyday life?

To see how my results can be applied in everyday life, put yourself in the shoes of an individual who wants to change his or her portfolio. You will want to include companies that are most likely to experience an increase in price over time and exclude companies that are most likely to fall in price.  And this is where the forecasts from the machine learning models come into play.

My research also provides stock return forecasts for multiple future dates tied to a company's characteristics.  These forecasts (discount rates) are essential to businesses when evaluating new investment opportunities.

‍

How did you come up with your research idea?

I learned about machine learning informally and was using them in my hobbies. In a conversation with my supervisor, Prof. Martijn Boons, he proposed that I try and bring these methods into my research. Given that my primary interest revolves around return predictability and machine learning models are fundamentally predictive, it was an easy union.

‍

Are you working on any project now?

Another ongoing project that I am a part of seeks to show that the academic literature and investment managers (practitioners) have overlooked some fundamentally important information set regarding stock return predictability. For a long time, both academics and practitioners have focused on the latest fundamental information reported by companies as being the only informative piece of data about security pricing. The implied consensus has been that if we are interested in pricing companies, then focusing on their latest reports is enough. In this project, we show that some of the fundamental information reported as far back as three to five years ago still has a lot to say about a company's valuation today.

‍