Do banks respond to capital requirements? Evidence from Indonesia

Abstract

Using dynamic panel data models, and addressing a common inappropriate use of simultaneous equation models in the literature, we examine the effect of capital requirements on banks’ behaviour in Indonesia. We find that banks tend to comply with capital requirements by increasing their capital ratios when the ratios are lower than, or falling towards, the 8% regulatory minimum. However, our results are mostly driven by large private-domestic banks and heavily undercapitalized banks that were closely monitored by the regulator in the aftermath of the 1998 crisis. Therefore, whether in normal circumstances banks in Indonesia comply with capital requirements remains questionable.

Acknowledgements

We are grateful to Hans Degryse, Jung Hur, Changhui Kang, Basant Kapur, Shandre M. Thangavelu, Julian Wright and participants of the Singapore Economic Review Conference 2007, Singapore, for their helpful comments. We thank Riza Haryadi, Dian Oktariani and Makin Toha of Bank Indonesia for providing the dataset.

Notes

¹ See Basel (2003) for a detailed description of Basel's risk-based capital requirements.

² See Santos (2001) for a theoretical literature review of bank capital regulation.

³ See, for example, Milne and Whalley (2001) and Milne (2002).

⁴ For this line of literature, see, for example, Jacques and Nigro (1997) and Aggarwal and Jacques (2001), which – like Shrieves and Dahl (1992) – examine US banks. Rime (2001) analyses banks in Switzerland, while Kleff and Weber (2004) and Heid et al. (2004) look at Germany's banks.

⁵ The PCA follows the 1991 US Federal Deposit Insurance Corporation Act.

⁶ See Pangestu and Habir (2002) for a brief summary of the 1998 banking crisis and the subsequent bank restructuring.

⁷ If the estimates of α₁ in Equation 13 and that of α₂ in Equation 14 are positive and negative, respectively, and their magnitude is large, we know that the regulatory pressures do induce banks to increase CAR. If the signs of the estimates of as are not as expected or their magnitude is small, then we would need to estimate Equation 15 to find out whether banks comply with the capital requirements.

⁸ A few banks do not submit financial reports for a number of quarters. These missing observations, however, make up a small proportion of our sample and therefore would not affect our results much. At the time we process the dataset, some banks had not reported their financial statement for the last quarter in our sample, the second quarter of 2005.

⁹ Large private-domestic banks may trade foreign currencies while small private-domestic banks may not. Regional-development banks are owned by provincial governments, and therefore are like state banks though they are typically small. Joint venture banks are joint ventures between domestic and foreign owners. Foreign banks are owned by foreign investors.

¹⁰ Bank Indonesia's definition of capital follows that of the Basel Accord (Basel, 2003).

¹¹ Again, we do not use the risk ratio, the ratio of risk-weighted assets to total assets, for the same reason that we do not use the capital ratio.

¹² Normally, the minimum requirement is 8%. However, in the aftermath of the crisis up to 2001, the central bank introduced a 4% minimum requirement for some banks.

¹³ It should be noted that in this risk equation, unlike in the capital equation, we introduce the dummy for public banks, TPublic, which equals one for banks that have gone public and zero otherwise. We include TPublic because using only Size and Income as control variables leads to very similar coefficients of lagged dependent variable Risk_ti for the OLS, fixed effect and system GMM. Our estimates in the capital equation, on the other hand, are about the same whether we include TPublic or not.

¹⁴ It should be noted that, unlike previous studies in this line of literature, we introduce lagged values of capital and risk as additional explanatory variables, not their current values.

¹⁵ Table 6 presents the results for the system GMM for capital, risk and CAR equations. There is no state-owned or foreign-owned bank which is under regulatory pressure if we use RegPCA, and therefore the interactive terms in Columns (1), (3) and (5) drop out of the regressions. Moreover, the interactive terms for joint-venture banks are dropped due to collinearity. All bank types present in regressions using RegPROB in Columns (2), (4) and (6).

¹⁶ The results are available from the authors upon request.