13 May 2020
Published last year in the International Journal of Finance and Economics, the paper, 'Do Credit Ratings Affect Spread and Return? A Study of Structured Finance Products' used structural equation modelling to investigate the relationship between credit ratings and spread. The empirical study found out that ratings might have an effective influence on decisions made by investors at the time structure finance products were issued. Findings in this paper support the possibility that ratings directly affect investors' actions in the primary market, which implies that regulators should focus their attention on the credit rating agencies' activities regarding the issuance of new structured products.1
The authors of the paper are Fernando Moreira and Sheng Zhao. This interview was conducted with Fernando, a senior lecturer in Banking and Risk Management, and a Co-director of the Centre for Service Excellence (CenSE) at the University of Edinburgh Business School. The interview covers aspects including research background, motivation and topic selection, the writing process and collaboration, data collection, the review process, contribution and improvement, and finally general advice for doctoral researchers.
Background, Motivation, and Insights Development
Chenzi: We're interested in diving into all aspects of your research and the research process. The first step for anyone is to choose the research question, so maybe you can start by telling us a little bit about how the idea for this paper emerged? In general, how should a PhD student select a topic?
Everything is difficult at the beginning, but Fernando believes people need to think more about the potential conclusion before starting to write. As the old saying goes, a good beginning is half done; the same is true for thesis writing. If you choose a topic and half of the conclusions you may reach are not good enough, this paper is not going to be published.
Fernando: "You should ask yourself if the answer is yes or no: can I publish the paper? Is the result interesting? Because sometimes you try to show something (x and y), if you show that x causes y, you can publish the paper. But especially in the empirical world, if the result shows x does not cause y, or impact y, sometimes it is not interesting and cannot be published, because you are showing something that does not exist. It is a kind of bias."
Nevertheless, there are some exceptions. In some types of appealing questions, people would be interested whatever you conclude.
Fernando: "For example, we discussed the competition before, there are two theories in competition and bank system, competition would increase stability or fragility. If I study that, anything you find in favour or against stability is ok because the two theories will both make sense to someone. So, the first step is to find a question which gives you interesting results."
My ideas come from papers in other areas I was reading.
Chenzi: Can you give some background on the research and how you went about the literature and found a gap? What is the best way to do a literature review?
Fernando believes doing a complete literature review is one of the keys to forming a good idea. Past papers allow us to realise the development trend of the theory, help us to keep up with the pace of the times, and better measure the academic value of new ideas. We can also combine the technique of different areas to fit in our own actual situation.
Fernando: "My ideas come from papers in other areas I was reading. I recommend you read this paper because the technique comes from papers in other areas like economics. There is no paper in this area structured with the same finance product using the same technique, so the idea comes from the combination of different areas."
It may sound like the best academic writers have a shortcut, but they don't. The more you get to read academic papers, the more knowledgeable you become at the end of it. When there are thousands of papers available, a scholar should make sure they have covered everything. Fernando shares his methods to do that:
Fernando: "When you choose your topic, you type keywords in the searching platform like Google Scholar, and you should start reading from the most recent papers with the keywords you think are in your case. Those keywords could be in the Title, Abstract, or Keywords. When the paper cites old papers, you go and read the old papers as well."
Chenzi: What was your initial plan and how did it change with time? Were there any imposed or emergent factors from academia or others? How does this research article fit into your overall research plan?
Fernando stated that feedback is the best way to improve your writing. He made the paper clearer after discussing it with his co-author.
Fernando: "When I finished the first draft, my co-author thought something was not clear and the referees might reject the paper, so I looked for more references to support something I said in the paper. In that case, I needed to explain something better, especially the methods, why it could conclude what I did."
Indeed, writing a research paper is like narrating a story to the other person: it should be neat, clear, and easy to understand. Fernando mentioned that it is meaningful to write a research article that fits into one's overall research plan, and this paper's contribution is more about the application of methods to measure two effects.
Fernando: "I have two areas, banking regulation and methods. Methods to measure the consequences, or to measure changing things in banking."
Chenzi: How do you approach writing? Is there a particular process you follow? What are the sections of a paper that require more attention in order to increase the chances of being published?
The great ideologist Confucius had a saying: knowing a thing is not as good as liking it, while enjoying it ranks the best. Fernando also states loving to write is an essential prerequisite for a successful thesis. He shared some tips:
Fernando: "Before I start writing anything, I write some bullet points. First I think about the structure of the paper, how many sections or subsections I am going to have. For example, introduction, literature review, one section about methodology, empirical results, and so on. Then I think about which information I should put in each section or subsection."
In order to increase the chances of being published, Fernando believes the introduction requires more attention. The introduction, as a road map to help readers follow the logic of your ideas, makes it easier for the readers to absorb and understand the details, and it is your chance to draw the attention and interest of your editor.
Fernando: "Especially for the very good journals, the editors don't have the time to read the whole paper."
Scholars usually split the paper; someone writes some sections and others do the rest.
Chenzi: While some papers are written by only one author, in many cases they are the result of a collaboration between two or more scholars. The article was written by you and Sheng Zhao; can you comment on how the writing roles were defined, and what these roles were? When there are several co-authors, how does this process work, and what are the necessary conditions to make sure it is effective for everyone?
For Fernando, there is no rule in collaboration: it all depends on your knowledge of your co-authors to decide who is going to do what. For example, in this special case, Sheng had the database. When Fernando saw it, he knew they could use the database to apply the technique and the modelling. He therefore got the data, ran the analysis, and wrote the draft. He then sent it to co-author Sheng, who made some suggestions. But in most cases, this can be quite different:
Fernando: "When we have three or four coworkers, people share the work. Some people do data analysis, some people collect the data, some people write because they write better. Scholars usually split the paper; someone writes some sections and others do the rest.
"My suggestion is when you look for co-authors, especially after your PhD, co-authors are almost like a marriage; you have to trust the person. I have heard of some cases where you start work with someone, you work hard but they don't do their part, so you have to know the person very well, to trust the person, to get along well with them. Sometimes, of course, people have different ideas and opinions; you have to find a middle point where everyone is happy."
He also mentioned that interdisciplinary specialities benefit collaboration and help to creatively solve some of today's most challenging problems and embrace new ideas:
"Sometimes interdisciplinary speciality can enhance the research quality, because it combines several points of view. In some cases, it is very good for papers because you can find different methods."
Chenzi: In the paper, you observe 895 rating-change events for 328 securities. The data was retrieved from Moody's website. Apart from public resources, have you ever collected data personally through companies? What are the common ways to approach the companies that provide the data? How do you conceive an interesting project for the data provider?
It is widely acknowledged that people find it very hard to get data related to banks. If you just knock on their door, they could not give you data, especially in banking. Providing confidential data results in harsh penalties, so bankers are very afraid to share the data. Luckily, Fernando had experience in using confidential data from companies.
His suggestion is that sometimes, you need someone to introduce you, someone that works in the company. If you have the opportunity to approach the bank, to perform some project, to run some analysis, and write the paper, of course you should tell them why it is important to the bank or company, and what the advantages for them are. You can also promise to deliver a presentation after getting the results, or write a short report, because in the company or bank, they don't read academic papers—especially the long ones—they just need a summary.
He recommended that PhD students take advantage of talking to people when attending conferences:
Fernando: "We have sessions according to topics. If you are going to present a paper, they are going to see your title, and if they are interested in your topic, they are going to your presentation. If they are there because they are interested in the topic, so they could ask you questions, you can remember their face; later in the coffee breaks or lunchtime, you can go and talk to them. People exchange business cards; I think it is the best way to start. Because if they like your topic and presentation, you can keep in touch with them by email, exchange ideas, and ask for suggestions."
Chenzi: How do you go about consolidating reviewers' comments? And any general comments on choosing a journal?
Fernando believes it is possible reviewers do not understand the content. Sometimes the referees are not experts; they accept and review the paper but they are not experts. Sometimes the writers are confident that what they wrote is very clear, however the reviewer said it is not very well explained or they said something that does not make any sense. If the comments show that they did not understand because it is too technical, it is the authors' responsibility to explain it. Fernando emphasised that sometimes you have to make it really clear, that is, to write in a more accessible way.
You may cite lower-level papers, but most of them should be at the level you are going to publish.
In terms of selecting a journal, Fernando mentioned the importance of references. When editors receive papers or manuscripts, they check the reference lists. Therefore, the first place you should look to decide your target journals is your reference list. They should be in the same area and at the same level (ranking). Of course, people can cite some lower-level journals, but PhD students should start reading level 4 journals if possible.
Fernando: "Because of that, naturally, the topic you are going to work on would be at the same level you are going to publish, and at least in terms of the same interest. You may cite lower-level papers, but most of them should be at the level you are going to publish."
Contribution and Improvement
Chenzi: The main objective of this paper is to test whether causality between ratings and ABS performance is plausible. But when the results do not reject the models, they cannot prove the causal associations analysed; they just assure the possibility of causality. Moreover, the tests face the risk of omitted variables bias, so the conclusions are limited to the samples. Can you recommend a few future paths to complement this research? Are you planning or already collaborating on any of these? How could the findings contribute to beyond academia?
Fernando came up with an innovative data analysis method to complement this research:
Fernando: There are other measures that I am using now, not with the same database but at least to address the problem of omitted variables. For example, there is something that might be small for empirical people, instrumental variables that address the problem and is a much more powerful technique. But the problem is much harder, we need to find some variable that complies with the condition, but when we do that, it has a good chance of being a good paper to publish."
Findings in this paper support the possibility that ratings directly affect investors' actions in the primary market. Therefore, the research has potential implications for regulators. Since the policymakers don't necessarily turn up to conferences, it is recommended that a personal website is an ideal and effective channel to pass the message to the regulators.
Fernando: "One way is to create a blog, or even to take advantage of the personal website at university. You can put a summary of the conclusion you reach because it is easy for someone to read. Even if they are googling for some information in their area, if it is very relevant, for example like this paper in ratings, and financial products, they will find the website. They will know your name."
Chenzi: Lastly, is there a piece of advice or any lessons you would like to share with readers and PhD students?
Fernando provided three main pieces of advice: be wise choosing a topic, do something you like, and be prepared for difficulties.
Fernando: "As I said before, when you choose your topic, especially if you are just starting your PhD, you can save time when you choose a more promising topic. Because if you choose a topic now and then you realise it is not a good choice, you may waste a year of your time."
He also provided his personal experience of finding a promising field:
Fernando: "By reading the literature review, when you find a gap you can start thinking 'what contribution could I make here?' and 'why is it important?'. 'No one did it before' is not enough; we need to ask more questions. Why did no one do it before? Sometimes because there is no data or it is not possible to do. For empirical work, it is very common; you should think about if it looks possible and how important it might be to someone."
You should be motivated, and be prepared, because in most cases, if we don't face the rejection of papers, at least they might ask us for revisions, changes, or corrections.
He mentioned that students should choose something that makes them happy and motivated. Academic papers need not be boring if you feel happy and take the time to construct your narrative in a way that is easy and interesting for readers to grasp the information.
Fernando: "Of course I have several ideas for topics, but my students should come up with at least a basic idea, because it should be something that they are really curious about. Sometimes it is quite boring and stressful when you collect data, or when your results are not favourable and you try something else. So you should be motivated, and be prepared, because in most cases, if we don't face the rejection of papers, at least they might ask us for revisions, changes, or corrections. Be prepared for that, because everyone faces that. It is not pleasant but it is part of our lives."
1 Moreira, F. and Zhao, S. (2018). Do credit ratings affect spread and return? A study of structured finance products. International Journal of Finance and Economics, 23(2), pp.205-217.