Conference on Chinas Financial Intermediation October 8 9

Conference on China’s Financial Intermediation October 8 -9, 2015, City University of Hong Kong Discussion on “Reserve Requirements and the Bank Lending Channel in China” By Zuzana Fungacova, Riikka Nuutlainen and Laurent Weill 8 October 2015 Ka-Fai Li Research Department Hong Kong Monetary Authority The views and analysis expressed in this presentation are those of the presenter and do not necessarily represent the views of the Hong Kong Monetary Authority 1

Motivations and Key findings Motivations • China’s monetary policy is complicated, multiple instruments (1. Required reserve ratio (RRR); 2. Benchmark interest rate (IR); 3. Loan quota; 4. Window guidance, etc…) • Interesting to know the monetary transmission mechanism via bank lending channel is applicable for least some instruments Findings • No evidence of bank lending channel through RRR and IR for Chinese banks • However, when bank ownership is explicitly considered, some limited evidences for some type of banks 2

Empirical model • where MP can be either RRR or IR • b should be negative • Bank lending channel is identified if f, g and h are significant and • • positive, meaning banks with lower access to external funding as proxied by lower cap, liq and size react more to MP However, they are either insignificant or significant with the wrong size for both RRR and IR By extending the model by including a dummy for bank type, the authors find limited evidences for Big 5 banks and city commercial banks 3

Comment 1: On the liquidity proxy • The authors use the share of liquid asset to total asset as the proxy for liquidity in the paper • But I think a useful metric is loan-to-deposit (LTD) ratio given the 75% limit imposed by the Mainland authorities • Banks that are close to the limit are more constrained (lower access to external funding) should react more to MP 4

Comment 2: RRR or IR? Or both? • Can MP be a vector of instruments? • Mainland authorities often use both instruments together • If identification permits, interesting to see whether RRR or IR (or both? ) is more conducive in monetary transmission 5

Comment 3: Demand side effect? 6