Xinxin Wang

University of North Carolina - Chapel Hill
4115 McColl Building
Chapel Hill, NC 27599

E-Mail: EmailAddress: hidden: you can email any NBER-related person as first underscore last at nber dot org
Institutional Affiliation: University of North Carolina, Chapel Hill

NBER Working Papers and Publications

August 2020Investor Tax Credits and Entrepreneurship: Evidence from U.S. States
with Matthew R. Denes, Sabrina T. Howell, Filippo Mezzanotti, Ting Xu: w27751
Angel investor tax credits are used globally to spur high-growth entrepreneurship. Exploiting the staggered implementation of these tax credits in 31 U.S. states, we find that while they increase angel investment, they have no significant effect on entrepreneurial activity. Tax credits induce entry by inexperienced, local investors and are often used by insiders. A survey of 1,411 angel investors suggests that a “home run” investing approach alongside coordination and information frictions explain low take-up among experienced investors. The results contrast with evidence that direct subsidies to firms have large positive effects, raising concerns about using investor subsidies to promote entrepreneurship.
July 2020The Leveraging of Silicon Valley
with Jesse Davis, Adair Morse: w27591
Early-stage firms utilize venture debt in one-third of financing rounds despite their general lack of cash flow and collateral. In our model, we show how venture debt aligns incentives within a firm. We derive a novel theoretical channel in which runway extension through debt increases firm value while potentially lowering closure. Consistent with the model's mechanism, we find that dilution predicts venture debt issuance. Empirically, treatment with venture debt lowers closure hazard by 1.6-4.4% and increases successful exits by 4.3-5.3%. Back-of-the-envelope calculations suggest $41B, or 9.4% of invested capital, remains productive due to venture debt.

National Bureau of Economic Research
1050 Massachusetts Ave.
Cambridge, MA 02138

Twitter RSS

View Full Site: One timeAlways