Search

We study non-performing loan (NPL) resolution using proprietary data on NPL transactions in China. Rather than representing orderly NPL resolution, the transactions appear to reflect banks concealing non-performing assets from regulators as (i) transaction prices do not reflect underlying credit risks; (ii) banks fund the NPL transactions and remain responsible for debt collection, and (iii) after two years, over 95% of NPL packages are re-sold at a profit to bank borrowers who do not appear to be NPL resolution specialists. Our findings suggest banks remain at least partially exposed to the NPLs that were removed from their balance sheets.

We demonstrate the negative impact of air pollution on China’s Initial Public Offering review process. We find that the pass rate of firms with connection to the review committee is increased by 10 percentage points when reviewed on a polluted day. The effect only exists for air pollution during working hours and pollutants near the workplace of the reviewers. The result is unchanged when using the wind speed as an instrumental variable and when controlling for a variety of potential missing factors. Using LDA (Latent Dirichlet Allocation) model, we show that the reviewers intentionally ask trivial questions during the review, supporting the idea that the committee members become more reckless during hazy days. This effect translates to an economic loss amounting to RMB 200 billion (USD 33 billion).

Why are certain communities more prone to anti-government protests than others? Do past rebellions lead to more protests today? We study the historical roots of social unrest using the experience of China. We document that counties with higher incidence of peasant uprisings against local government officials during the Qing dynasty period (1644-1912) also have higher incidence of anti-government protests in present-day China. To generate plausibly exogenous variation in the incidence of past protests, we exploit differences in the visibility and magnitude of solar eclipses during the Qing dynasty period. In the Confucian tradition, solar eclipses are considered a negative divine signal on the legitimacy of the ruler, facilitating the coordination of protest actions. We test this mechanism using detailed data on the timing and location of anti-government rebellions extracted from local chronicles. Counties within the totality zone of a solar eclipse were significantly more likely to experience a rebellion in the eclipse year. Leaders of early anti-Qing rebellions were recorded in popular culture and celebrated in temples, favoring the transmission of the memory of their actions across generations. The persistent effect of past protests is stronger in counties with such temples and memorials, consistent with a long-term memory of revolutions.

This paper considers lending to finance projects in a setting where repayment enforcement appears impossible. The loan was illegal and thus legally unenforceable. Creditors were incapable of applying private coercion to force repayment. Borrowers lacked both collateral and reputation capital. Project cash flows were unobservable. The projects were the acquisition of Imperial administrative posts by scholars in nineteenth century Qing China. The lending mechanism was the “trusted-assistant loan.” Our model of trusted-assistant lending shows that it is a renegotiation-proof implementation of efficient state dependent financing. Empirical analysis of officials’ diaries and bank records shows that the employment of trusted-assistant lending and the performance of trusted-assistant loans conforms roughly with the model’s predictions.

This paper utilizes a novel instrument variable to solve the endogeneity problem that exists in identifying a causal effect between property rights protection and investment. In China, the replacement of a city mayor, the key figure lobbying for a higher local land supply, has a substantial negative impact on the total land quotas for local government. However, the impact of the replacement of city mayors on other investment-related policies, such as tax and fiscal expenditure, is quite weak. This exogenous decrease in land supply is used to verify the impact of obtaining land with property rights protection on investment. The result of instrumental variable regression shows that for each US Dollar that is invested in entitled land purchasing, the induced investment is 1.05 US Dollars. This indicates that the investment promoted by property rights protection accounts for 43% of economic growth in China for the period between 2008 and 2011. This impact is exerted through both a security channel and a collateral channel.

In this paper I will explore the cyclical nature of political connections by investigating market reactions to news of a unique and effective method of monetary contraction in China: the adjustment of reserve requirements. I illustrate that politically-connected firms, when the ultimate controller is either the central or the local government, demonstrate an additional negative cumulative abnormal return (CAR) when the reserve requirement is increased. There is also a lower level of real activity on the part of those connected firms during the tightening period. I argue that this impact is likely to be exerted through the bank-lending channel: specifically, banks are not willing to lend to connected firms due to the concern of either a higher probability of them defaulting, or a lower probability of a bailout from the government. This effect is less likely to result from other factors, such as changes in the value of collateral, decreases in investment opportunities, or government intervention in regard to banks.

This paper utilizes a natural experiment to examine the role of the protection of property rights in promoting investment. In order to explore a title-granting scheme in Shenzhen, China, I collect a sample of 83 listed SOE firms, with 32 of them holding about-to-be-entitled lands. Those landholders exhibit both a sharp short-term 7.8% additional increase in stock market price and a long-term 63% extra increase in investment, when compared with non-landholders, despite that there is no pre-event structural difference between the two. These increases in value are a result of having solved hold-up problems rather than a result of increased collateral values because those politically connected SOEs under analysis are financially unconstrained. Cross-sectionally, those firms with weaker pre-event protection against “hold-up” are associated with greater increases in share price and investment. Potentially, solving the hold-up problem of all unentitled land would bring about value of 2.2 Trillion RMB, almost triples Shenzhen's GDP in 2009.