Does local government bond issuance help improve bank funding liquidity?——Quasi-natural experimental evidence from changes in restrictions on bank investment in lead underwritten local government bonds
Since Brunnermeier and Pederson(2009),it has been recognized that financial institutions'funding liquidity plays a key role in maintaining financial market stability.However,the majority of existing studies have focused on the interaction between financial asset market liquidity and financial institution funding liquidity(He and Krishnamurthy,2012;He and Xiong,2012a;He and Xiong,2012b).There has been a paucity of research exploring the financing function played by high-quality collaterals.China's large-scale local bond issuance commenced with the local government debt replacement initiative initiated in 2015.The cumulative volume of local debt issuance from 2015 to 2021 exceeds 36 trillion yuan.Commercial banks have been the primary underwriters of local debt issuance,purchasing more than 80％of local government bonds and holding them for an extended period.A further question arises:as a high-quality collateral security second only to treasury bonds,is the large-scale supply of local government bonds conducive to facilitating banks'collateralized financing transactions,thereby enhancing bank funding liquidity?This question has been somewhat overlooked by many researchers.This paper examines the impact of banks'holdings of local bonds on their participation in collateralized financing transactions and the liquidity effects of bank financing induced by local bond issuance.However,since the breakdown items of bank holdings are not publicly available,in order to overcome the difficulty of data availability,this paper innovatively utilizes the policy change of removing the 20％restriction on bank investment in lead underwritten local bonds issued by China Banking and Insurance Regulatory Commission in August 2018,and conducts a quasi-natural experiment to identify the impact of the change in the proportion of lead underwritten banks'holdings of local bonds on their repo financing transactions.The paper employs a difference-in-difference empirical strategy,utilizing the policy change of the 20％restriction removal.The baseline regression results indicate that the removal of the 20％restriction can significantly increase the proportion of bond repo transactions of local bond lead underwriting banks in the current period as well as in the two subsequent periods.Furthermore,an increase of one yuan in the bank's lead underwriting of local bonds after 2018 can lead to an increase of about 0.12-0.15 yuan in the bond repo transactions in the same period.The bank's lead underwriting of local bonds after 2018 is associated with an increase of approximately 0.12-0.15 yuan in bond repo transactions in the same period.This is a notable and economically significant effect.Two approaches are employed to address the potential endogeneity issue.Firstly,the distance between the city where the bank is located and the provincial capital of the province is selected as the instrumental variables for the share of bank's lead underwriting of local bonds.We then perform DID estimation based on the instrumental variable.Secondly,we use PSM to select a matched sample,and then conduct DID estimation based on the matched sample.The estimation results of both approaches are consistent with the baseline regression.The paper also finds that the lifting effect of the 20％restriction removal on their bond repo transactions is greater in the samples with higher liquidity difficulty,higher liquidity creation,higher sensitivity between interest expense and interest rate,and higher sensitivity between deposit and interest rate.Furthermore,the paper finds that local bonds in regions with higher economic development growth are used more for repo financing transactions;the removal of the 20％restriction does not have a significant effect on the cost of bank bond repo transactions;there is a significant positive effect of increased bank holdings of local bond on bank's future risk-taking.This paper has the following policy suggestions:Firstly,it is imperative to maintain a certain scale of local bond issuance.Secondly,local governments must take every possible step to ensure that local bonds are repaid in full and on time,and to strengthen the protection of the legitimate rights and interests of local debt investors.Thirdly,other qualified investors should be actively encouraged to participate in the issuance and trading of local bond,in order to optimize the investor structure and effectively improve the secondary market liquidity of local bonds.This paper contributes to the existing literature by filling the gap on the causal link between local bond supply and bank funding liquidity.It provides insights into the mechanism of the coordinated operation of fiscal and financial affairs with Chinese characteristics.Furthermore,it offers a valuable reference basis for the public to judge the overall impact of local bond issuance on the Chinese economy.Moreover,there is a paucity of micro-empirical studies on the repo market in China.This paper may be the first empirical research to examine the repo behaviour of Chinese banks utilizing microdata at the bank level.Consequently,this paper also contributes to a full comprehension of the dynamics within the Chinese repo market.Future research should pay attention to the expansionary effects and economic consequences of the large supply of local bonds in the bank funding market.As local bonds are increasingly used for collateral financing in the bank repo market as well as for financing through the central bank's monetary policy instruments,it is worthwhile for researchers to examine in depth the possibility of an expansion of financing by financial institutions that is not planned by the central bank,and whether this poses a threat to the stability of the financial system.
Local government bond issuanceBank underwritingBank funding liquidityQuasi-natural experiments
万方数据
