Non-performing Loans in Housing Bubbles
Faculty of Economics,Fukuoka University,Japan
It was found that housing bubbles caused non-performing loans (NPLs), using bank data in China from 2007-2015. At first, by default risk classification and ownership, we found that all the NPL ratios decreased till 2011 then began to increase. Additionally, using a bubble test and the ratio of monthly price to rent housing bubbles were found in 36 major cities between 2004 and 2015.By panel estimations on 1,117 and 2,832 pooled bank-year data, it was found that housing prices had significantly negative impacts on NPL ratios after controlling for macro- and micro factors, and that decreasing housing prices since 2011 have raised NPL ratios. Furthermore, treating “restriction of speculative housing purchase (RSHP)” in 50 cities from 2010 to 2014 as a quasi-natural experiment, and using 10,043 observations for monthly housing price and rent in 70 cities from 2000 to 2015 to adopt the difference in difference estimation, then using this restriction as an instrument for housing price to regress NPLs, it was found that RSHP had significantly raised NPLs. Lastly, quarterly housing price Granger caused NPLs but the opposite was not supported. Therefore,controlling housing prices would be important for micro- and macro-prudential policies of preventing systematic financial risk in China.
JEL: G21, C23
Keywords: Non-Performing Loans, Default Risk, China, Bubble, Housing Policy, Restriction of Speculative Housing Purchase, Reserve Ratio, Inefficiency, GDP, Bank Size, Interest Margin,Dynamic Panel, Difference in Difference, Granger Causality