The lending channel of monetary policy in Indonesia

Abstract

This study evaluates the bank lending channel of monetary policy in Indonesia by using quarterly bank-level data over the period of 2005-2016. I find that the lending channel of monetary policy works for all banks, both large and small. The results suggest that higher capital buffers and better liquidity positions moderate the impact of changes in monetary policy on credit growth for large banks, while capital buffers and liquidity positions do not alter the strength of the lending channel for small banks. The findings indicate that the central bank can use prudential instruments affecting capital buffers and liquidity positions for managing the strength of adjustment in the monetary policy interest rate on bank credit growth.

Introduction

The 2008-09 Global Financial Crisis (GFC) shed light on the importance of banking stability and its critical role in supplying credit to the economy. As the GFC caused a global economic recession, central banks in affected countries pursued a monetary loosening to stimulate bank lending in an attempt to accelerate economic recovery. This policy response renewed interest in the lending implications of monetary policy, or the lending channel of monetary policy.

In the aftermath of the crisis, the Indonesian economy’s growth rate declined from 6.35 % in 2007 to 4.70 % in 2009, while credit growth contracted from 31 % in 2008 to 10 % in 2009. To support Indonesian economic recovery, Bank Indonesia (BI) lowered its policy interest rate, the BI rate, over the period Q1 2009 to Q1 2013. This loosening in monetary policy was intended to stimulate higher lending growth. However, from Q2 2013 to Q4 2014, the BI increased its policy rate in an attempt to reduce capital outflows due to the Federal Reserve’s tapering. This tightening regime contributed to an unprecedented low level of credit growth in 2014. The BI has since 2015 resumed loosening its monetary policy by lowering the BI rate. Nevertheless, credit growth was sluggish and stood at just 7.8 % by the end of 2016, the lowest rate of growth since the 1997- 98 Asian Financial Crisis. Fig. 1 illustrates the development of the BI rate, credit growth, and economic growth in Indonesia.

This experience of Indonesia highlights the importance of understanding the impact of monetary policy on bank lending and the factors that may alter the strength of the relationship between monetary policy and bank credit. Theoretically, the lending channel of monetary policy works if a tightening in monetary policy reduces banks’ reservable deposits, thereby reducing lending growth (Bernanke & Blinder, 1992; Kashyap & Stein, 1994; Morris & Sellon-Jr, 1995). Conversely, a loosening in monetary policy increases lending growth.

Peek and Rosengren (2013) suggest that the effectiveness of the lending channel depends on the ability of banks to adjust their reservable deposits following changes in monetary policy. Banks’ liquidity and capital positions play a critical role in the ability of banks to adjust their reservable deposits. Less-liquid and less-capitalised banks are less able to access alternative sources of funding to replace their lost reservable deposits following a tightening in monetary policy (Kashyap & Stein, 1994; Kishan & Opiela, 2000). Investors will demand higher external finance premiums, as these banks are considered to be riskier (Bernanke, Gertler, & Gilchrist, 1999). As a result, the higher funding costs will lead less-liquid and less-capitalised banks to limit their lending to a greater extent than more-liquid and better-capitalised banks would do. In other words, the loan growth of less-liquid and less-capitalised banks is more responsive to changes in monetary policy, while the loan growth of more-liquid and better-capitalised banks is less sensitive to these changes (Kashyap & Stein, 1994, 2000).

Against this backdrop, the objective of this paper is to examine the role of monetary policy in managing bank lending growth in Indonesia. I address two research questions: (1) to what extent do changes in the policy rate, the BI rate, impact the lending of commercial banks? (2) Do bank liquidity and capital buffers alter the strength of the lending channel of monetary policy? The case of Indonesia is relevant for a study on the lending channel of monetary policy, since Indonesia is a bank-based economy where banks hold around 70 % of total assets of the financial sector (Park, 2011). The limited market-based financing implies that Indonesian commercial banks rely on deposits as the main source of funding. Meanwhile, borrowers depend on bank loans as the main source of external finance, given that the asymmetric information (lack of accounting disclosures and creditworthiness indicators) restricts their access to raising finance in the capital markets.

To answer these research questions, I follow the work of Kashyap and Stein (1994) by modelling bank credit growth as a function of changes in monetary policy. I use quarterly bank-level data from 90 Indonesian commercial banks between 2005 and 2016 to capture variations in bank lending growth. I use changes in the BI rate as a proxy for changes in monetary policy. I control variation in bank lending due to the dynamics of the demand-side as well as the supply-side of loans to estimate the effect of changes in monetary policy on lending growth. I run regressions for a sample of all banks as well as for subsamples of large and small banks.

I find strong evidence that the lending channel of monetary policy works for all Indonesian banks and for both large and small banks, since an increase in the BI rate reduces their lending growth. Conversely, a reduction in the BI rate expands such lending. Variations in capital buffers and liquidity positions significantly alter the strength of the lending channels for large banks. Higher capital buffers and liquidity positions moderate the impact of the BI rate on the credit growth of large banks. However, variations in capital buffers and liquidity positions do not play a significant role in determining the strength of the bank lending channel for small banks.

Fig. 2 shows that the liquidity of large banks was at the lowest level in 2014 and their non-performing loan ratios had increased since 2013. The rising credit losses reduced their capital buffers to the lowest level during this period. Consequently, the loan growth of these less-liquid and less-capitalised banks became more responsive to a tightening in monetary policy, which explains why an increase in the BI rate resulted in an unprecedented low level of credit growth in 2014.

Since 2015, large banks have responded to their increasing vulnerability with higher capital buffers. Their choice of bank stability over lending growth during a contraction phase of the credit cycle indicates a risk-averse behaviour, as higher capital buffers imply a higher opportunity cost of bankruptcy for equity holders thereby reducing their risk-taking (Keeley, 1990). Fig. 2 shows that large banks have improved their liquidity position since 2015, thereby supporting this risk-averse behaviour. As the banks became more liquid and better capitalised, their loan growth was less responsive to a loosening in monetary policy, which explains why the substantial cut of the BI rate from 2015 did not stimulate higher lending growth.

This research is related to the work of Agung (1998), but differs from his work in some important dimensions. First, Agung (1998) evaluated the bank lending channel by using data for 1983-1995. This period followed financial deregulation in the early 1980s, prior to the 1997-98 Asian Financial Crisis (AFC). The AFC contributed to a large scale Indonesian bank default and in turn, costly bail-outs. The devastating impact of the crisis led the BI to impose stricter supervision and prudential standards on intermediation activities to promote banking stability. The crisis was a lesson on the importance of adopting prudent risk management practices for Indonesian commercial banks. Banks’ consideration of their own risks and resilience, their borrowers’ risks and creditworthiness, and general macroeconomic prospects affects their willingness to extend loans. My research examines the lending channel of monetary policy under this new banking landscape by utilising bank-level data from 2005 to 2016. Second, Agung (1998) does not analyse how bank capital buffers and liquidity positions affect the strength of the lending channel. By contrast, my work considers how these bank-specific factors may alter the strength of the lending channel. Lastly, in the environment of financial deregulation before the AFC, Agung (1998) finds a small impact of changes in monetary policy on lending growth, since the lending channel worked only for small banks. Utilising instead a sample covering the period after the AFC under a new banking environment, I find that the lending channel of monetary policy works for both large and small banks and that bank capital buffers and liquidity positions alter the strength of the lending channel for large banks.

As such, this study is the first to investigate how bank-specific factors affect the strength of the bank lending channel of monetary policy in Indonesia. The results demonstrate the significant role of capital buffers and liquidity positions for the lending channel of large Indonesian banks, which hold about 75 % of total banking assets. The finding suggests that regulators may encourage large banks to build up capital buffers and liquidity positions during an expansionary phase of the credit cycle and release them during a contractionary phase. The higher capital buffers and liquidity positions, which are built up during the expansionary phase of the credit cycle, are useful in preventing a credit crunch, should the BI tighten monetary policy during the contractionary phase. Lowering capital buffers and liquidity positions strengthens the lending channel of a loosening monetary policy to stimulate higher lending growth during the contractionary phase of the credit cycle. As such, this study points out that the BI can use relevant prudential instruments (i.e., capital buffer and liquidity) to manage the strength of the monetary policy impact on loan growth. This implies a need for the BI to closely observe the credit cycle to judge the stength of the lending channel of monetary policy.

This paper proceeds as follows. Section 2 explains the development of the hypotheses and the relevant literature. Section 3 discusses the data, variable construction, identification strategy, and estimation technique. The results and robustness checks are presented in section 4. And the conclusion and policy implications are discussed in section 5.