7,649
edits
Changes
Jump to navigation
Jump to search
The Effect of Leverage Buyouts on Innovation (view source)
Revision as of 15:33, 10 July 2020
, 15:33, 10 July 2020→Addressing Concerns
Addressing his concerns:
*A paragraph on the many and varied endogeneity issues is a very good idea. We should have done this more properly.
*The paper (in the dropbox folder) seems to use an event study approach to isolate the stock market reaction (in terms of market cap) to a patent issue. We can't do that - it's too much work. Also, almost all (85%+) of patent applications result in at least one issue, and the 'news' occurs when the patent application is made public (usually 18 months after filing), so that whole paper is pretty questionable. But then the team is outstanding and includes Amit Seru. They also solicited comments from the who's who of innovation economics.
Kogan, L., Papanikolaou, D., Seru, A., & Stoffman, N. (2017). Technological innovation, resource allocation, and growth. The Quarterly Journal of Economics, 132(2), 665-712.
===JF Rejection===
The paper was rejected by Amit Seru, though he was nicer than the AE or the reviewers. His comments were as follows (emphasis added):
Thank you for submitting your paper, MS 2019-0353, "The Effect of U.S. Public-to-Private Leveraged Buyouts on Innovation," for publication in the Journal of Finance. I have now received excellent feedback on your paper from two referees. The referee reports are attached to this email. Both referees recommend that I reject the paper. The AE also recommends a rejection. The main concerns of reviewing team center around the marginal contribution of the study and its empirical execution.
The AE summarized the sentiment of the reviewing team in the note to me as "I largely agree with the referees. I have concerns about the methodology (there is a lack of transparency in the way the paper is written) and even taking the results at face value I don't find the contribution large enough to publish the paper in JF":
*The topic of innovation and LBOs is already well occupied by existing papers. To make a contribution vis a vis Lerner-Sorensen-Stromberg, we would have to think a bit deeper about the economic mechanism that links research and capital structure. Why would PE target destroy value ? In LSS, LBOs seem to be creating value through some form of disciplining effect and better innovation. But '''LSS have a different sample, mostly made of private to private deals.'''
*Even if I take the results at face value, the decrease in patent stock mostly comes from the target selling patents right in the year when takeovers take place. This is the main result of the present paper. But is this value destroying ? We do not know. The million $ question in this literature is whether LBO targets really destroy value. But it's a hard one.
*'''The methodology is opaque.''' Is it Diff in Diff or matching ? There doesn't seem to be pair fixed effects (so it's closer to DD, than close to the LSS methodology than the authors claim). Even though '''the authors never actually write down the econometric equation.''' In any case, using matching is not a big innovation in the literature, many papers already do it, and in a more transparent way. The true innovation in this literature would be to find an IV.
*To make a paper on LBO publishable in JF, I think one of two things need to be done: (1) having a real source of variation or (2) providing clear and unambiguous proof that LBO target destroy value by being too short-termist. The paper unfortunately does none of the two (but these are both very hard problems, so I really don't blame the authors)
*Like the referees, I would recommend that the authors rewrite the paper to make much more transparent. Settle on a sample (not removing data points every time you run a different regressions). Use it to replicate existing finding (LSS seems like a good template, perhaps others). Confirm existing results, and then slowly build from there.
*Something I was confused about: How do the authors get information on post-LBO R&D spending? After the LBO the firm is not in compustat anymore. perhaps there is an explanation somewhere but I missed it.
'''I have also read your paper and found a lot to like about it.''' However, I was not able to convince myself to overrule the concerns raised by the reviewing team. I have therefore decided to reject your paper.
'''Select comments from Reviewer 1''' (uses terminology like 'forward citation' and doesn't understand 'synthetic controls'):
#The paper currently reads as if the main sales point is the methodology
#The exact method they used however is not clear. In the introduction they describe it as “difference-in-difference”, but in the text and particularly in the tables is hard to understand the empirical strategy (it doesn’t help that the equations are not written in the text).
#Instead, I would focus on what I think is the strength here, which is understanding more deeply how corporate innovation changes after LBOs, an in particular the distinction between in-house innovation and patent purchases, as well as the more strategic considerations of managing patent portfolios (expiration, renewals etc.)
#the authors argue that “…the entrepreneurial theory of LBOS stands alone in its positive connotations for innovation…” (pg. 12). Instead, I would argue that at least two other theories could predict a positive effect on innovation from LBOs. There is a long-standing hypothesis that LBOs (specially public to private—the focus here) can relieve managers from short‐term pressures from public shareholders. Similarly, LBOs can also alleviate agency issues between shareholders and managers, which can include the avoidance of innovation investments by CEOs with career concerns.
#the authors’ description of the “myopic based theory” is contradictory. I don’t see how they conclude that the treatment effect of an LBO would be to curtail innovation activities. One of their main points leading to the argument is that Bernstein (2015) shows lower innovation in public firms.
#The writing is at times confusing and contradictory. For example, in pg 13 it is not clear what the authors mean by “We also require each LBO to be matched to a control firm in the same industry, operating at approximately the same tie, and holding within 20% of the LBO target firm’s patent stock…” Do you mean to say that matched firms co-own 20% of the patents of the LBO target?
Select comments from Reviewer 2 (likely one of Lerner, Sorenson, or Stromberg):
*'''This is a well-written paper that investigates an important question.''' While other papers have tackled this same question... there is room in the literature for a paper that revisits this question using rich new data and a careful identification strategy. The submitted paper does analyze richer patenting data than prior studies, and some of its descriptive findings are certainly interesting. However, I am afraid that '''the paper’s identification strategy''' and '''discussion of identification challenges''' falls well below the standards now expected by readers of the JF.
*the authors do not appear to fully appreciate the endogeneity challenges they face in trying to identify a causal relationship between LBOs and innovation... private equity (PE) firms do not choose LBO candidates randomly. PE firms do extensive due diligence on their target firms... '''the concern remains that the matched firms likely differ from LBO firms on dimensions that the authors do not observe (and even some that they do observe) and these differences, and not the LBO itself, might drive the post- LBO differences in their investment activities.'''
*To be sure, addressing this endogeneity challenge is a tall order, and there is no straightforward solution to it. But at the very least I would advise the authors to acknowledge the challenge and avoid overstatements such as “[w]e use a difference-in-differences approach to address issues of causality” (p. 3). Davis et al. (AER 2014) is a helpful example of how the endogeneity challenge can be discussed.
*It would also be helpful if the authors could explore the sensitivity of their results to their matching choices
*it would be helpful if the authors could provide a more thorough discussion of ... why are LBOs associated with a decrease in patenting?
*the “myopia-based theory claim[ing] that the market is inefficient and does not correctly value long term investment projects” (p. 9) has often been cited as a reason why PE firms can generate value by taking firms private and thereby allowing them to focus on long-term investments
*The authors appear to hold a more negative view of these cuts, seeing them as “worrisome” (p.
35) and thus implying that they harm the long-term value of the firms. However, it is not clear that cutting productive innovative investments will help PE firms increase the value at which they can sell their portfolio companies five years down the road. In particular, the so-called “non-productive balance sheet restructuring” that the authors discuss in page 11 would only be optimal if PE firms expect to sell their companies at a multiple of the book value of the companies’ assets
*I thought the introduction is written in a very casual tone, with several non-obvious statements that are not backed by any references. Some examples are:
**“[P]ublic-to-private leveraged buyouts remain particularly important and controversial.” (p. 1)
**“The empirical literature also leads one to strongly suspect that public-to-private leveraged buyouts cause reduced innovation. An empirical study in the late 1980s and early 1990s established the stylized fact that increased leverage is associated with reduced R&D expenditures.” (p. 2)
**“Our sample consists solely of material, innovative U.S. public-to-private buyouts that were leverage financed, and it is much bigger than previous samples of its kind.” (p. 3). I am not sure what the authors mean by “material” or what other previous samples they are comparing their sample to.
*Also, I thought that the theoretical discussion in Section 2 of how innovative activities might be affected by LBOs is biased toward theories that view LBOs as negatively impacting innovation. In particular, I am not sure the statement in page 12 that “[t]his entrepreneurial theory of LBOs stands alone in its potential positive connotations for innovation” is correct. For instance, as discussed above, short-termism and other agency theories can predict that private firms will invest more in innovation than public ones
*a) The authors write in p. 4: “Like Lerner et al. (2011), we find that patents held by innovative LBO firms receive more citations after their buyouts. However, we are able to attribute this solely to citation inflation: we find no statistically significant difference in citations received between LBO firm patents and control firm patents before or due to an LBO.” '''Lerner et al. (2011) are aware of such citation inflation''', and they address it by scaling citations by the citations received by matching patents (pp. 459-460).
*b) The sentence in p. 19, “In our sample, firms that experience a leveraged buyout are large, and those that have patents are about half as large again as those that do not” is confusing
*c) The sentence in p. 23, “However, including either firms without patents in the year of the buyout, or firms with a large number of patents does drive our results” appears to be missing “not” (as in “does not drive…”).
=Version 2=
