Box A: Stress tests of the New Zealand banking system
Stress testing is a tool to assess the resilience of financial institutions to a hypothetical adverse event, usually a severe but plausible economic downturn. Introducing a comprehensive stress testing framework for the New Zealand banking system is a strategic priority for the Reserve Bank, and this is discussed in more detail in chapter 6.
During this year, the Reserve Bank and the Australian Prudential Regulation Authority (APRA) collaborated on stress tests that included the four subsidiaries of the major Australian banks. The Reserve Bank also conducted a smaller exercise with the five domestically owned banks.1
The major bank stress testing exercise featured two adverse economic scenarios over a five year time period. In scenario A, a sharp slowdown in economic growth in China triggers a severe double-dip recession. Real GDP declines by around 4 percent, and unemployment peaks at just over 13 percent. House prices decline by 40 percent nationally, with a more marked fall in Auckland. The agricultural sector is also impacted by a combination of a 40 percent fall in land prices and a 33 percent fall in commodity prices. The decline in commodity prices results in Fonterra payouts of just over $5 per kilogram of milk solids (kg/MS) throughout the scenario.
Scenario B features a significant increase in interest rates. Initially this is due to a stronger-than expected economic recovery. However, from the second year of the scenario there is a material slowing in global economic growth accompanying a severe disruption to global oil production. The consequent inflation pressures delay monetary policy easing, despite a domestic recession, a 30 percent fall in house prices and unemployment of just under 12 percent. Rising global bank funding costs increase lending rates by a further 200 basis points, accentuating tight monetary policy and resulting in floating mortgage rates peaking at 11-12 percent. Fonterra payouts fall to around $6 kg/MS from the second year.
These scenarios were intended to be severe but plausible, and in the case of scenario A, broadly matched the experience of some of the more severely affected economies in the GFC.
The stress test was conducted over two phases. In phase one, banks were asked to provide estimates of credit losses, based off their own internal stress testing models. Results in this stage were subject to a high degree of variability due to differences in modelling approaches between banks. In the second phase, banks were supplied with common credit loss estimates, which were estimated on the basis of Reserve Bank and APRA stress testing models, international experience during similar scenarios, and bank responses from phase one. In both of these phases, banks were initially asked to model the effects of these scenarios on balance sheets and profitability without assuming any management actions to mitigate their impact.
The most significant effect of the scenarios was to generate a large increase in loan losses, with impaired asset expenses peaking at 1.3 and 1.5 percent of assets, respectively, in the two scenarios in phase two (figure A1). Higher credit losses, combined with a decline in net interest income due to increased costs for bank funding, resulted in a significant decline in bank profitability. However, reflecting strong underlying earnings in the New Zealand banking system, these factors were only sufficient to cause negative profitability in a single year in each scenario (figure A2).
Figure A1: Impaired asset expenses (percent of gross loans)
Figure A2: Return on assets
While positive average profitability meant that banks were able to conserve capital levels, deteriorating asset quality resulted in an increase in the average risk weight of exposures, causing a decline in regulatory capital ratios. Common equity Tier 1 (CET1) capital ratios declined by around 3 percentage points to a trough of just under 8 percent in each scenario, but remained well above the regulatory minimum of 4.5 percent (figure A3). Banks are also required to maintain a 2.5 percent conservation buffer above all minimum regulatory capital requirements, or else face restrictions on dividends. On average the banking system fell within this buffer ratio in both scenarios, due to total capital ratios falling close to minimum requirements (figure A4). Average buffer ratios reached a low of 1 percent in both scenarios. As a result, some banks would have been faced with restrictions on their ability to issue dividends. The intention of the buffer ratio is to provide a layer of capital that can readily absorb losses during a period of severe stress without undermining the ongoing viability of the bank. Given the severity of the scenarios, capital falling within buffer ratios was an expected outcome.
Figure A3: CET1 capital ratios (percent of risk-weighted assets)
Figure A4: Total capital ratios (percent of risk-weighted assets)
Faced with a scenario of this magnitude, banks’ management would be expected to undertake a number of actions to restore capital buffers. A second part of phase two involved banks outlining these management responses and to model their effects. For most banks, the key response was to significantly reduce new lending to reduce risk weighted assets and increase regulatory capital ratios. Some deleveraging is to be expected in a severe recession, especially on corporate exposures where defaulted assets will generally not be replaced. However, continued supply of new lending is essential to prevent markets from becoming disorderly and to allow banks to resolve defaults by selling foreclosed assets to new owners. The proposed average reduction in exposures of around 10 percent in each scenario was large, and may have resulted in significantly worse macroeconomic and credit loss outcomes.
Most banks also report that they would reduce costs by cutting bonuses, reducing staff numbers and lowering discretionary expenditure. Some banks also reported measures to re-price credit to recoup margins, while one bank reported a capital injection from their parent.
The result of these mitigating actions is to return the banking system to profitability in all years of the scenarios, and to boost CET1 capital ratios to 9.5 percent at the low point of each scenario. International experience suggests that it can often be difficult to implement significant mitigating actions in the midst of a severe crisis. Therefore, in interpreting results, the Reserve Bank’s emphasis tends to be on ensuring that banks have sufficient capital to absorb credit losses before mitigating actions are taken into account. The results of this stress test are reassuring, as they suggest that New Zealand banks would remain resilient, even in the face of a very severe macroeconomic downturn.
The Reserve Bank also engaged the five domestically owned banks in a basic credit stress test in April this year. This test assessed the balance sheet resilience of these banks to a large increase in credit losses due to a severe domestic recession and large falls in asset prices. All participating banks appear to be resilient to a major downturn of this nature, with no bank breaching minimum capital requirements. This test was designed as an introduction to stress testing and is part of a multi-stage process to enhance stress testing capacity at these banks.
The Reserve Bank views stress testing as an important component of sound risk management within banks. The banks are expected to continue to develop their stress testing frameworks, and to use the results to inform their capital management and risk appetite setting processes.
1 The domestically owned banks are: Co-operative, Heartland, Kiwibank, SBS, and TSB.