No abstract

The Discussion Paper series is intended to make the results of the current economic research within the Reserve Bank available to other economists. Its aim is to present preliminary results of research so as to encourage discussion and comment. Views expressed in this paper are those of the authors and not necessarily those of the Reserve Bank. Use of any results from this paper should clearly attribute the work to the authors and not to the Reserve Bank of Australia. The contents of this publication shall not be reproduced, sold or distributed without the prior consent of the Reserve Bank of Australia.

Macrofinancial stress testing is a tool to help policymakers better understand the key systemic vulnerabilities in a financial system. The Reserve Bank of Australia's (RBA) macrofinancial bank stress testing model is an example of this, enabling the RBA to analyse potential financial risks to Australia's banking sector, such as those arising during the COVID-19 pandemic. The model projects how economic shocks may influence a bank's profitability, dividends, loan growth and capital position, primarily using decision rules and accounting identities that are uniformly applied to profit and balance sheet data for the nine largest banks operating in Australia. It is designed with a focus on understanding systemic vulnerabilities and a philosophy of prioritising transparency over complexity. The key advantages of this model are its ability to quickly produce estimates of the capital loss in response to various macroeconomic scenarios, model various forms of contagion across banks, and allow the modeller to undertake 'reverse' stress tests. The paper sets out the key features of this model, how it was used during the past two years and the areas in which further work is required.

This paper studies the role of collateral in credit markets under stress. Australian interbank markets at the time of the Lehman Brothers failure present a platform for identification, because the collateral is liquid and homogenous across borrowers (unlike in retail credit markets), the shock is large and exogenous (unlike in countries with bank failures), and there is comprehensive administrative collateralised and uncollateralised loan-level data. After the exogenous shock, collateralised and uncollateralised borrowing compositions diverge. Uncollateralised borrowing declines for <i>ex ante</i> riskier borrowers while collateralised borrowing increases for borrowers <i>ex ante</i> holding more high-quality collateral. Moreover, riskier banks with sufficient high-quality collateral substitute from uncollateralised to collateralised borrowing. In aggregate, collateralised borrowing expands substantially, predominantly collateralised against second-best (but still high quality) collateral, while interest rates on loans against first-best collateral fall substantially, indicating scarcity of the most-liquid safe assets. This liquid asset demand encourages collateralised lending, contrary to cash hoarding.

The Australian Prudential Regulation Authority implemented 2 credit limits between 2014 and 2018. Unlike similar policies in other countries, these imposed limits on particular mortgage products – first investor mortgages, then interest-only (IO) mortgages. With prudential bank-level panel data, we empirically identify banks' credit supply and interest rate responses and test for other effects of these policies. The policies quickly reduced growth in the targeted type of credit while total mortgage growth remained steady. Banks met the limits by raising interest rates on targeted mortgage products and this lifted their income temporarily. The largest banks substituted into non-targeted mortgage products while smaller banks did not. Practical implementation difficulties slowed effects of the (first) investor policy, and led to some disproportionate bank responses, but had largely been overcome by the time the (second) IO policy was implemented.