The banking sector has transitioned from being at the epicenter of the 2008 global financial crisis (GFC) to being part of the potential solution following the 2020 COVID-19 outbreak. We believe that the impact of COVID-19 on banks has been moderate in most APAC countries because of the supportive stance of governments and financial institutions, along with relatively better initial conditions. Accommodative government policy, coupled with less stringent regulation, has provided banks with adequate time to absorb the potential shock. Despite that, the spreads for APAC financial institutions are wider than their U.S. peers, largely for technical reasons.
This study assesses how banks across Asia Pacific have been affected by COVID-19 and how they performed under our stress test analysis. Based on this analysis, discussed below, we conclude that Japan, China and Korea may offer attractive investment opportunities with the potential for excess return and defense. On the other hand, we maintain a cautious approach towards Australian and Indian banks considering the risks of rising nonperforming loans (NPLs) related to structural issues and the ongoing pandemic.
PIMCO’s Regional Bank Stress Test
Our stress test scenario assessed the current financial strength of banks in the region by analyzing how they might perform under a severe global financial crisis, such as the GFC in 2008-2009 and the Asia crisis in 1997 – as discussed above, so far Asia has avoided such a scenario. Our analysis suggests that in this scenario, we should expect major Asia Pacific banks to suffer an average loss of 50-200 basis points (bps) on their common equity tier 1 (CET1) capital—the core capital banks are required to hold to absorb unexpected losses.
It is worth noting that even in our risk case scenario, all regions retained at least high single digit CET1 ratios, so we believe the risk of capital impairment is remote. We saw a slightly smaller negative impact on banks in Japan, Korea and China in the stress test, but Australian and Indian banks underperformed their regional peers. This is not because Japan, Korea and China are immune from the impact of COVID-19, but rather because we expect widespread use of forbearance (reducing payment amounts) and payment holidays to delay the recognition of potential losses.
Banks’ net interest margins (NIMs) are under pressure. However, Japanese banks have been experiencing pressure on NIMs for decades and offsetting the impact by diversifying their revenue sources. Korean banks are also pursuing revenue diversification and 1H results demonstrate that they are making good progress. Chinese banks’ earnings are likely to decline this year, but they maintain relatively higher NIMs and return on equity (ROE) that may offer an adequate cushion to absorb losses from NPL write-offs.
We also expect Australian and Indian banks may see negative net income under our stressed scenario, while Japanese, Korean and Chinese banks should be less affected. We believe that the divergence mainly stems from different initial conditions. While NPLs in Japan, China, and Korea remain near historical lows, NPLs in Australia and India were increasing even before the COVID-19-outbreak. Authorities in India and Australia have cut interest rates, extended debt repayment moratoriums and offered fiscal stimulus packages. However, we do not view these measures as sufficient to adequately offset the negative asset quality impact from COVID-19, as well as existing structural issues. When banks report large negative net income, the risk that banks skip paying their coupons on their additional tier 1 (AT1) capital is expected to rise because it may trigger capital distribution constraints.
Key Highlights for Each Market
Australian banks have started to make preemptive provisions for the impact of COVID-19. The management teams of these banks estimate a credit loss of 100-200 bps based on a range of macroeconomic outcomes (normal credit loss is about 20 bps per year). According to our analysis, there is further downside to these management assumptions since the forecasts incorporate a fairly high probability weighting to a V-shaped recovery.
Australian banks are expected to face higher credit costs in the future since housing sentiment is weak and unemployment is rising. Mortgages make up the majority of Australian banks’ loan exposures. The underlying loan-to-value distribution of these mortgage loans will likely be the primary driver of loss given default and is also an important factor contributing to the probability of defaults (negative equity borrowers are more likely to default all else being equal). Banks can extend the initial six-month debt relief, from March to September 2020, by up to four months.
Our stress test assumed a moderately more negative macroeconomic scenario than prevailing consensus forecasts. While we expect no principal loss from capital instruments, our analysis indicates that it is possible banks may skip paying coupons on their AT1 capital as the CET1 ratio nears 8% in our stressed scenario.
In 1Q 2020, we saw significant divergence between listed banks’ reported NPL ratios and the implied NPL ratios reflected by the listed corporate universe. At the end of 1Q, the NPL ratio for the 10 main listed banks we track was almost flat compared with the level at year-end 2019. However, the implied NPL ratio that we calculated based on a sample corporate portfolio surged to 18% in 1Q, up from an average of 6.2% for 2017-2019. This divergence reflects the massive scale of forbearance loans that Chinese banks have provided to pandemic-impacted borrowers, in addition to the ‘evergreen’ loans they generally provide to state-owned borrowers. Work resumption in the second quarter could narrow the gap between reported NPLs and implied ones.
Due to the outbreak of COVID-19, the People’s Bank of China (PBoC) and the China Banking and Insurance Regulatory Commission (CBIRC) have required banks to revitalize China’s economy and concede their profits to the corporate segment. We see banks delivering on these dual requirements by keeping credit policies lenient for broad-based enterprises and, at the same time, increasing NPL write-offs and reporting over 20% profit contraction in 2Q 2020. We expect Chinese banks to continue writing-off significant numbers of NPLs into end 2020 or even 1H 2021. This should allow them to keep their reported NPL ratio stable. The cost of doing so will likely be a widening capital gap, with loose credit policy boosting risk-weighted assets while earnings contraction is expected to erode core capital generation. That said, the PBoC is committed to securing onshore bond funding channels for banks to replenish capital.
Prior to the outbreak of the pandemic, the Indian financial sector was struggling with weak corporate asset quality and a shadow banking crisis. Capital and provision levels remain low, and there is still meaningful downside for bank profits. In addition, state-owned banks require regular capital injections from the government. Only a few more conservative banks have started to make provisions for COVID-19 related losses. State-owned banks generally have not done so.
Following the Reserve Bank of India’s directive, banks offered an initial six-month moratorium for all borrowers until the end of August 2020, with the option to selectively restructure loans after that.
The take-up rate was around 20% across prime retail and corporate term loans at stronger banks. Weaker lenders and micro, small, and medium enterprise borrowers have shown a substantially higher take-up rate. Our stress test scenario expects a material share of the loans under moratorium to become impaired over the next few quarters, which will likely translate into elevated credit costs.
Japanese banks are experiencing a relatively mild impact from the pandemic, reflecting better initial conditions (recourse loans, moderate loan growth, cash at hand) and ample liquidity support from the government (including economic measures worth Yen 108 trillion (USD 1 trillion), or 20% of GDP). In addition to that, the CET1 buffer of Japanese banks is quite high compared with peers because they maintain around a 12% CET1 ratio, and the regulator has set a low ‘point of non-viability’ (PoNV) trigger, which is when banks have negative net worth (a CET1 ratio of less than 0) instead of CET1 hitting 5.125%, which is the trigger level specified under Basel III.
We believe that debt instruments, such as total loss-absorbing capacity (TLAC) notes, are the most attractive means of investing in Japanese banks considering strong external support and favorable regulation. In our view, it would be difficult to pursue excess returns by investing in Japanese bank equity since profitability could remain low for a considerable period of time under our stress test scenario.
We did not see a significant increase in credit costs in 1H 2020, partly because of the ample liquidity in the market. Korean banks increased retail loans by KRW 40.7 trillion (USD 34 billion) in the first half of 2020, double the KRW 21.4 trillion (USD 17.8 billion) supply in the first half of 2019. Relief loans, backed by government guarantee, are driving this growth. During the same period, non-bank financial institutions reduced their lending by KRW 4.4 trillion (USD 3.6 billion). So in essence, credit risk is being transferred to the government balance sheet.
Under our risk case scenario, we would expect Korean banks’ CET1 ratio to drop to around 11%, which would leave them in a stable financial condition and unlikely to suffer principal impairment. Similar to Japan, the PoNV trigger for Korean banks is when they fall into negative net worth.
With COVID-19 leading to unprecedented disruption for the global economy, we are taking a cautious approach, focusing on careful security selection as uncertainty over the economic recovery remains high. Overall, we have a constructive view on Japanese, Chinese, and Korean banks. Despite the COVID-19 outbreak, banks in these countries have maintained low NPLs and delinquency ratios, thanks to better initial conditions and strong government support. While credit spreads have come down since the March 2020 peak, we believe these financial institutions still offer attractive valuations compared with their U.S. peers due to the absence of (or modest, if any) large scale corporate bond purchase programs by their central banks. Bonds issued by these banks are less correlated to global indexes and remained resilient during the market turmoil in March. We believe that banks in these countries offer attractive investment opportunities for those who seek additional yield along with defensive attributes. On the other hand, we do not expect much spread tightening in Australian and Indian bank bonds considering their deteriorating credit fundamentals.