Whether venture capital is capable of improving the survival status of start-ups company is an attractive yet controversial research topic. At present, no consensus has been reached on the positive effect of venture capital, mainly due to endogenous bias and selective bias, which brings enormous difficulties accurately measuring the treatment effect. In recent years a large number of peer-to-peer lending platforms get bankrupt in China after an unexpected rapid growth, which provides a rare quasi-experimental opportunity to systematically assess the effect of venture capital on the survival condition of the platforms. 

This article aims to probe the following three questions: Can venture capital improve platform survival? Is the impact of venture capital differentiated? What is the mechanism of venture capital? We adopt a python-based crawler software to obtain the detailed information of 3176 platforms from a pubic website, including registered capital, business model, security model, venture capital, amount of venture capital, number of rounds of venture capital. We divide the samples into the treatment group and the control group, respectively, and employ the propensity score matching method and generalized propensity score matching method in the counterfactual framework to modify the selection bias and screening effect to get a more robust conclusion. Finally, it explores the mechanism of venture capital. 

Our study on the role of venture capital in the survival profile of start-ups and the evolution of the industry has yielded valuable insights. First, empirical evidence shows that acquiring venture capital helps improve the living condition of invested enterprises. Propensity Score Matching method shows that, on average, compared with the control group, these platforms which receive venture investment reveals a 14% lower mortality rate and 9.13 months longer survival time. These differences remain after excluding endogenous bias and selective bias and stay robust under different matching approaches. Second, the decision-making of venture capital institutions is two-stage, with the first stage deciding whether to inject venture capital into the platform. The second stage is to determine the amount of venture capital. The two-part model demonstrates that registered capital and shareholder background affect the amount of venture capital obtained by the platform. The generalized propensity score method reveals that the impact of venture capital is heterogeneous. More specifically, only venture capital more than 10 million Yuan is sufficient to prolong the survival time of platforms. On average, 10 million yuan of venture capital increases the survival probability of a platform by 42% to 80%. With the increase in the amount of venture capital, the improvement effect is more significant. Third, the mechanism model confirms that venture capital works for start-ups through three channels, namely bringing in social capital, reputation upgrading, and financial support. Access to venture capital will help P2P platforms join official associations, gain credit recognition from third parties, and achieve capital increases. It is also necessary to point out that venture capital plays a role more "perfect" instead of "timely assistance" because of the screening effect. The function of venture capital is limited, which reminds it is unwise to exaggerate the role of venture capital blindly. 

Our research has two policy implications: first, to govern improper advertising. A large number of P2P platforms highlight the equity participation of venture capital institutions, which may lead to abuse of venture capital institution′s reputation, and may also mislead low risk-preference investors to generate the illusion of implicit guarantees and rigid payments. Second, to control aggressive operation. Quite a few P2P platforms have become more aggressive in their business strategies after getting venture capital. They offer discounts, dividends, and other ways to attract ordinary investors. If this "aggressive operation" continues to fester, it will pose a massive risk to the entire online lending industry.