China has experienced one of the largest and longest credit booms in history, with total credit to the non-financial sector more than quadrupling between 2008 and 2019 to RMB 255.9 trillion (259% of GDP by 2019), double the leverage ratio of other emerging markets. China’s banks allocate 80% of loans to state-owned enterprises (SOEs), which accounted for nearly half of China’s debt by 2018, despite SOEs showing higher levels of bad debt than private firms. Bank branch managers are evaluated based on lending targets, raising the question: how do these quotas contribute to bad debt accumulation and affect economic efficiency? 

The data. Researchers analyzed over 300,000 business loans issued by 1,500 branches of a large Chinese state bank between 1997 and 2010. Firms receiving loans were grouped by asset size. The researchers then selected over 20% of firms from each of the three groups and obtained complete loan histories for each loan contract, including issuance dates and time of the month, values, types, quality classifications, and borrowers’ income statements. Loan quality was assessed using the bank’s five-class system, where the bottom three categories indicated a higher likelihood of default. The research team then compared lending patterns and default rates between SOEs and non-SOEs. 

When month’s end looms, lending booms. The researchers find that bank managers doled out 92% more loans in the final days of each month compared to days in the middle of the month. The increase in loans occurred both in the number of loans issued as well as the amount loaned. While there was a lower rate of lending at the beginning of each month, this shortfall did not come close to offsetting the end-of-month increase. End-of-month increase in bank lending appears for both SOE and non-SOE borrowers.

Monthly average value of new lending in sample

Risky loans to SOEs drive decline in end-of-month loan quality. A month-end loan is about 8% more likely to eventually be classified as a bad loan compared to loans in the middle of the month. While the end-of-month increase in bank lending appears for both SOE and non-SOE borrowers, the quality decline is driven primarily by loans to SOEs. 

Eighty two percent of loans in the sample went to SOEs. SOEs with past bad loans constitute 37% of SOE borrowing, as compared to just 6% of borrowing by non-SOEs with past bad loans, suggesting more stringent loan criteria for private firms. SOE loans are also more prone to default: 16.2% of sampled SOE loans are ultimately classified as bad, nearly double the 8.4% rate for non-SOEs. However, the end-of-month increase in bad loan rates is three times higher for SOEs than non-SOEs, while there is no increase in bad loan rates among non-SOEs with clean credit histories. Given that the average beginning-of-month bad loan rate is 14%, this implies an 18% month-end increase in the already high rate of bad loans among SOEs. This finding is consistent with the interpretation that state banks do not discriminate against SOEs with poor credit history when they increase lending to meet targets at the end of each month. 

Relaxed lending standards account for rise in riskier SOE loans. The standard tools that bank managers have for meeting quantity targets are greater effort and relaxing standards. For non-SOE lending, the increase appears to come from an increase in effort, as there is no increase in default. Since SOE lending quantity increases and quality declines, month-end loans to SOEs appear to be due to a decline in  bank manager standards to meet month-end quotas. 

Monthly average share of bad loans in sample

Lending quotas hurt economic productivity. The researchers evaluated the impact of increased month-end lending to risky SOEs on the productivity of China’s economy as measured by total factor productivity (TFP). They estimate that quotas raised loan amounts to SOEs by approximately 50%. Previous research has estimated that the TFP of SOEs is 41% lower than that of private firms. Using these estimates as benchmarks, they calculate that TFP for China’s economy is 2.5% lower as a result of bank managers lending to risky SOEs to fulfill their monthly lending quotas. They note the TFP loss could be larger depending on how one defines the proportion of China’s investment that was financed by bank loans in the past two decades. 

Quotas exacerbate capital misallocation. The evidence suggests that bank managers turn to higher-risk SOEs to meet loan quantity targets, while maintaining relatively high standards for non-SOE lending. This in turn suggests that not only do branch managers face little consequence for making bad loans to SOEs, but they may even benefit from having this ready outlet for lending available given their own quantity-based incentives. Because loan officers in China have little discretion over interest rates, riskier month-end lending cannot be recovered via higher interest rates. This incentive structure may exacerbate capital misallocation in China’s economy.