1 of 14

Discussion of Homroy and Mkrtchyan (2025)

Mikael Paaso

Erasmus University Rotterdam

​

ENTFIN 2025

2 of 14

This paper studies partisan homophily in VC investment

  • VC-CEO political alignment increases the likelihood of an investment

​

  • Alignment is associated with better firm outcomes
    • Higher likelihood of IPO
    • More IPO underpricing (??)
    • Quicker exits
    • Fewer failures + lawsuits
    • Lower CEO turnover
    • More patents

3 of 14

Main comments

  • Relevant paper!

​

  • Contribution
    • Dig deeper into mechanism

​

  • Execution / plausibility tests
    • I have a hard time with the current interpretation
    • Economic magnitudes often seem implausible (e.g. 100% increase in p(IPO)) 🡪 need for perfect execution

​

  • Minor notes

4 of 14

There is a relatively large literature on partisan homophily in economic decision-making

  • Colonnelli, Neto and Teso (2025 AER) -> employment

​

  • Fos, Kempf and Tsotsoura (R&R JF) -> executive teams

​

  • Dasgupta, Guo, Ren and Shu (2025) -> executive-board homophily

​

  • Duchin, Farroukh, Harford and Patel (R&R JF) -> employee partisanship and M&A

​

  • McCartney, Orellana-Li and Zhang (2025 JF) -> residential decision

​

  • Siriwardane, Tang and Vallee (2025) -> pension fund allocations to PE

​

  • VCs:
    • Wu, Symitsi and Bozos (2025)
    • Qu (2025)
    • Wang (2025)

​

5 of 14

How can you contribute to the literature?

  • The elusive question in partisan homophily remains whether homophily is driven by affect or views about the world
    • The paper is currently pitched almost entirely about affect or interpersonal homophily

6 of 14

People from different parties have different views on how the world should run

  • Initial VC-CEO match and future success may be determined by granular industry specialization
    • Goes beyond fixed effects

​

  • Possible to use a survey + experiments to dig deeper into mechanisms? (a la Colonnelli et al. 2025)

​

  • Qualitative analysis of funded company descriptions?

7 of 14

Some minor execution comments

  • P. 11: “We identify 777 unique startups […] and 2,210 unique startups” -> Sample size indicates 7,814 startups

​

  • Table 2: Supposedly based on seed / series A funding, but N=19,000 when median number of rounds = 4

​

  • “We include x fixed-effects” -> not in most tables

​

  • CEO coded as D if donations(D)>1.2*donations(R) and vice versa

​

​

8 of 14

The CEO coding could induce some bias

  • CEOs with no donations are coded as neutral
    • Neutral-neutral matches are not considered aligned

​

  • But what if donating to a political party correlates with quality?
    • Non-donating firms are all by definition mismatched
    • Foreigners are not allowed to donate

​

  • Text mentions excluding non-donators, but I don’t see it in the tables
    • What about using voter registration as the proxy for partisan affiliation?

9 of 14

You should show that donator-VCs and CEOs are not different

  • Simple summary statistics by party should help with this
    • The main tests in the table should include VC (partner)-year FE

​

  • Estimating the results without non-donators will also help

​

  • It would also be important to also test if your results on performance also hold for RR and DD matches separately

10 of 14

Summary

  • The paper got me thinking!

​

  • It’s a crowded field and importantly, mixed results
    • All papers -> higher p(match)
    • Wang (2025) and Wu, Symitsi and Bozos (2025) -> worse performance
    • This paper and Qu (2025) -> better performance
    • Can we learn something from this?

​

  • Some results look implausibly large -> focus on great execution!

11 of 14

Other minor notes

  • From the summary stats table:

​

​

​

​

  • It seems there is a typo here?

​

​

12 of 14

Other minor notes

  • We include the same control variables as in our previous tests. In addition, we include an indicator variable that equals one if a startup is located in CA or MA (p. 20)

​

  • Why not just used state-year FE like in other tests?

​

​

13 of 14

Other minor notes

  • Why is the 4-week return considered a positive outcome? IPO underpricing is generally seen as a negative

​

​

14 of 14

Other minor notes

  • Standard errors should account for the fact that your counterfactual deals may be repeated (e.g. imagine a Democrat VC fund invests in two Democrat CEO firms in the same industry 🡪 the counterfactual firms will be identical)

​

​