Feedback Pitching Documents 2023
My feedback on the pitches are more direct than I typically did in the past. I do make a distinction between the quality of the pitch, which is reflected in the mark, and the feasibility of your research project, which is reflected in the feedback.
I think the most recurring mistake is that you start from some cool data and then try to retrofit a research question that can be answered with that data. For the proposal, I would all encourage you to think more about what the actual research question is. What is the actual theoretical effect or actual theoretical measurement that you want to detect?
My feedback largely falls in two categories:
- Questions and suggestions about what the research question is.
- Suggestions about how to make sure that the proposed measures are capturing what I think the actual question is.
Lichoo Tay
Tokenization seems to have multiple effects, e.g. fractional trading is possible, digital trading is possible, … I would spend some time on theoretically separating out these effects and see whether there is any literature on these theoretical effects for other markets. For instance, you could learn form papers on debt and mortgage securitisation which allows to fractionalize and recombine existing financial instruments. Maybe the literature on the decrease in tick size on stock exchanges could help as well. More specifically to real estate, are real estate investment firms another vehicle to fractionalize investment in real estate?
Are you mixing up or not connecting tradeability of fractions of the property versus transactions of the property?
This model has the advantage of taking into account the survivorship bias (ie only successfully transacted properties would make it into the dataset)
Only successful properties in the data set means that you have survivorship bias. I do agree that proportional hazard models will deal with the problem though.
Try to get an idea of data access early in your PhD.
Gyles Davies
Thus, while it appears evident spin-offs create value for both parties, the academic consensus is that spin-offs are infrequent which poses the question why do we not see them more often?
Tying this back to the difference between Average Treatment Effect (ATE) and Average Treatment on the Treated (ATT). One explanation would be that the value is only created in a limited number of potential cases, those firms are doing spin-offs, and researchers estimate this ATT. On the other hand, the ATE for the whole potential population of spin-off is negative.
The motivation is really good. I am not sure I follow the hypotheses though. How is past performance related to focus? How is size related to diversification?
For the data, you would want to quickly get an idea of the number of spin-offs in your sample.
For robustness, the same logistics regression model can be applied within Propensity Score Matching approach (Matsumura, Prakash, and Vera-Munoz, 2014).
Why not use this as the main analysis? In general, the comparison between spin-off observations and non-spin-off observations deserves more attention. Or is hypothesis 1 more to do propensity score matching for hypothesis 2? I.e. what is the real research question?
• The significance of this research is founded on the importance of understanding why we observe far more mergers and acquisitions than spin-offs
I am not sure this is true. To answer that question, would you not have to identify firms that did not do a spin-off but should have (based on a prediction model?) and look whether they are underperforming a matching firm who did a spin-off? That is you want to know the ATU, the average treatment on the untreated.
Sami Almukhtar
Research questions that start with “How” make me nervous. It’s rather vague.
Are there no other key papers that investigate CDS in response to geopolitical shocks?
It’s not really clear why downside risk (i.e. CDS) should be the key focus instead of upside risk from the motivation. Practically speaking, CDS are less liquid and trade less. This could mean that the reaction in CDS prices is more noisy.
After the motivation, I am also a bit worried about how you are going to separate the effect of different trade restriction policies from each other and from the effect of the war itself. I do think this question is more interesting than it is formulated in the research question. You are basically testing whether the market believes green transition policies and using the response to the Russian invasion as the research setting.
The three different periods goes towards my question about separating the effect but it does mean that you do not have an event study design.
It provides an understanding of how a political event such as the Russia-Ukraine war can affect the default risk of energy firms differently depending on the region.
A lot of these firms operate in different countries. How do you account for the effect that European energy firms sell oil on a global market? I think the way you identify your industries is going to be important.
Will Chalmers
You position your RQ as global warming has a structural impact on prices. However, you are looking at the impact of basically the weather on traders immediate reaction. Could someone also interpret your findings as irrational overreaction by investors?
If investors respond to abnormal temperature changes, then there is a disconnect between the true extent of climate change and the perceived extent of climate change which educational programs and policies should address to improve the effort to reduce climate change.
This “so what” basically gets at the issue I raised above. What is the mechanism? What would it mean if you find an effect. I am actually glad to read this.
Would it be worthwhile to investigate other classifications of the climate sensitivity? For instance insurers and reinsurers are exposed to flood, fire, and hurricane risks while they might be low-emission. Mining companies might be high emission but if they are more exposed to minerals vs oil and gas might be important.
Just a note: In one of the last seminars, I’ll give an alternative way to deal with these abnormal models which may or may not be relevant to you. Also, I think there are better ways of dealing with the count data.
Jae Rong Chong
For the proposal, you want to think a bit more careful about the potential reactions and trade-off between explaining why the incident will not happen again without given potential attackers more information. This seems undertheorised at this point.
I also think there is a disconnect between the questions the research questions and the research method. I am not sure how the scored and indices can capture what is being disclosed at best they are going to be measures of how many things are being disclosed and maybe how well or in which quantity. There is also no mention of a before-after analysis while that seems to be implied by the research question.
Han Wen
However, there is much concern within the sector and in the accounting profession that the reporting requirements (amongst other requirements) are not appropriate impacting efficiency and effectiveness negatively
This suggest that the appropriate criteria to determine reporting requirements are efficiency and effectiveness. How are these constructs defined? Can they be measured? Is there any literature on the efficiency and effectiveness of reporting requirements? Maybe for the for-profit sector. Or is efficiency and effectiveness different in the not-for profit sector? If so why? Or how?
Why survey accountants and not for-profits? Why are you measuring size and not efficiency and effectiveness directly if that is the appropriate criteria?
, my study proposes to utilize total assets, number of employees (including volunteers), budget, and annual revenue as the criteria for assessing NFP size
Why will these measures be better at capturing the efficiency and effectiveness of the reporting process?
The whole idea of the paper seems to be that the accountants you will survey know best what the appropriate criteria is to determine which not-for-profit should fall under the different reporting requirements. I think you need to do a better job of explaining why only accountants are relevant in this discussion, what the current reporting regimes are, why you focus on measures of size, and not on general criteria for reporting requirements, why you only focus on measuring size and not on the cut-off values.
Chelsea Graham
How are you going to separate out the effect of institutional attention to climate change? For instance, Larry Fink of Blackrock gives an interview where he emphasizes climate change, managers react to this but it also get discussed on Reddit.
Essentially, the reddit comment data could be collected for a period of 30 days before the earnings call script data is published.
This deals with reverse causality but not with the omitted correlated variable problem. It also assumes that the information is made public in the earnings call is new and was not known before.
This is a good pitch but I don’t really believe that the research design can answer the question. It is too much driven by the availability of innovative measures and not enough by a clear research question. At the least I would want to see it spelled out more how retail attention can drive managerial attention (climate or other topics). Is there other literature on this topic? Second, I think you need to think more about alternative explanations if you find a relation. Is it driven by institutional investors? Is it driven by a general attention to climate change risk in society? Or there situations where you might be able to separate out these effects more? For instance, are there firms or economic circumstances when firms will be more responsive to retail investors?
Simon Xu
To study the difference between web2 and web3 as it is unanimously agreed to be the new revolution.
I don’t agree. So, not unanimously.
Decentralization as a key principal in web3, what it will change traditional industries, and its implied value for decentralized firms by comparing to their centralized counterparts. Moreover, to investigate DAO as a decentralized governance system whether it is a gain or a loss for the firm. And to study the potential values of increasing cross-chain interoperability between major decentralized platforms
You really need to be a lot more critical of this sort of technobabble. If centralization is so bad, why does virtually every for-profit company centralise the ultimate power of decision making in one person, a CEO. Why are for-profit companies not run by a committee with a democratic vote? There is a whole literature in economics and management, management accounting and more on decentralised decision making within firms and the trade-offs it entails. You cannot just set this aside.
valuations comparison between web3 firms and their counterparts.
Why focus on valuation? Why not on quality of decision making? Why not ask the question for which economic transaction a DAO works well and for which ones they should not?
Is the biggest problem not that a lot of market prices in the crypto world are manipulated by insiders and you cannot take them at face value? For instance, according to the court documents Celsius was already functionally insolvent at the time that Terra/Luna crashed. The true value throughout most of 2021 and 2022 for Celsius was 0. I assume that is not the value that you are going to use but I think you should.
In my personal opinion, in 2023 you can no longer credibly pretend that crypto prices are coming from a real financial market and compare them to an actual healthy, financial market. I think a better approach would be to compare different type of crypto projects with each other at least it is a like for like comparison. Maybe you can identify an interesting theoretical distinction between different crypto projects in terms of decentralised decision making or investor protection.
Theresa Santoso
There seem to be a lot of research questions embedded in your motivation and research question. I think it might be wise to focus it more. I think what you are interested in “does it matter when CEOs decide to take a pay cut as a signal that they are a team player”? You can tackle that by looking at this from the determinants or from looking at the effects. You can than ask the question which approach will help you the most to answer the question. For instance, it might be that it is not possible to identify directly whether CEOs are voluntarily taken a pay cut, so it’s hard to look at the consequences but you can hypothesis certain drivers that are more likely to be indicators that CEOs are team players like you plan on doing.
I don’t really see how poor financial performance is an indicator of being a team player, it could also be that the pay cut is punishment for bad performance. Could the same not be said for investor sentiment? Maybe you can look at Glassdoor as a way to measure employee sentiment and see whether it is correlated with the probability of taking a pay cut.
I think you could benefit from thinking a little bit more about the setting. How do you define a pay-cut? Is it just salary or does it include stock compensation? From the previous literature what are the theoretical motivations that are proposed for a pay cut, how would these explanations play out in the COVID era? Would it be different early on when there is a lot of uncertainty? Is it different for later pay cuts when the (macro) economic effects are clearer? Does the size of the pay cut matter? Could you exploit this by looking at the variation in the size of the pay cuts?
Sam Samuels
I am not sure I really “get” the research question?
How do you define the proportion of environmental friendly projects based on number of project (or based on revenues or another outcome)? How can you exclude alternative explanations that environmental friendly projects are financially more (less) attractive? If mining firms can benefit from investing in environmental friendly projects why are they not all doing it? What is the mechanism for the variation in the number of environmental friendly projects?
Yes, Eikon, S&P, and Yahoo are reputable sources for financial and ESG data
Is this true? One of your key papers seems to suggest that this is not true.
With the advent of green bonds is it appropriate to only look at the cost of equity? Is the cost of debt not also relevant?
I think thinking more carefully about the setting would be a good idea. For instance, when you talk about environmental friendly projects it could be increasing the efficiency of production which decreases CO2 emissions and costs. Or it could be a lithium mine which helps with the transition to EVs. Do you count divestments of environmental unfriendly projects such as coal mines?
Furthermore, have you thought about the following dynamic. Most publicly traded firms are to a certain extent exposed to institutional investors which might have ESG mandates. While privately held firms are not and maybe they are more free to take on environmental unfriendly projects.
You might be able to exploit the variation in institutional holdings in public companies to investigate the pressure and to further substantiate your hypothesis that mining companies have investors that are more or less concerned about ESG.