Since we launched our first ETF in January 2016, there have been four significant macro corrections in four years. None of those large and painful corrections represented a crisis, insofar as the declines did not represent a threat to the solvency of the financial sector, either from a lack of liquidity or the destruction of capital. Rather, they were related to the market rapidly (and, in hindsight, incorrectly) re-pricing stocks to reflect expectations of slowing GDP growth. In this Insight, we explore the implications of the coronavirus (COVID-19).
The current correction relating to COVID-19 is notable in both its severity and speed. Even the “lower-beta” Canadian banks and Australian financials have experienced recent peak-to-trough corrections of ~40% (before bouncing this week). The speed and severity have left many investors wondering about the character of the correction – is it a growth scare, a credit event, or a financial crisis? To answer this question, we believe it is instructive to review the European sovereign debt crisis (2010-2011) and the global financial crisis (2008-2009), as both were legitimate crises representing a direct threat to the solvency of systemically important financial institutions.
Distinctions between a growth scare, a credit event and a crisis are critical
A growth scare is a correction in anticipation of slowing economic growth that is not accompanied by an actual decline in GDP or at least a decline commensurate with the decline in stock prices. This has been the case for the last three macro corrections, all of which represented an overreaction by equity markets to the potential for slowing GDP growth, namely the yuan devaluation (2015-2016), Brexit (2016) and the inversion of the yield curve (late 2018).
A credit event (or credit cycle) is characterized by an abrupt decline in value of an asset that serves as collateral in the banking sector. In the last three major credit cycles, the deteriorating assets causing losses were loans backed by commercial real estate (1991-1992), telecom/media/technology assets (TMT, 2002) and U.S. residential mortgages (global financial crisis 2008-2009), the last of which caused a global financial crisis and recession. The bursting of an asset bubble is almost always coincident with a recession, which causes a rise in credit losses and lower revenues. Historically, credit cycles are relatively short in duration, lasting a few quarters, with share prices starting to recover in advance of peak credit losses. Even a severe credit cycle can avoid any permanent loss of shareholder value if losses can be absorbed by a combination of excess capital, quarterly earnings (net of dividends), and reserves.
A financial crisis goes farther and is often accompanied by regulatory-required recapitalization of (parts of) the financial sector, causing dilution and a permanent loss of shareholder value (often exacerbated by shares issued at depressed prices). In the extreme, rising insolvency risk can result in failures (e.g., the global financial crisis), or can be muted with aggressive policy response (e.g., the European sovereign debt crisis). Basically, the difference between a severe credit cycle and a financial crisis is whether the risk of losses is so great as to potentially overwhelm capital thereby creating insolvency risk.
How does COVID-19 compare to the European debt and global financial crises?
Given the severity and extremely short duration of the correction so far, we believe financials investors should consider how COVID-19 compares against the last two legitimate crises, the European sovereign debt crisis (2010-2011) and the global financial crisis (2008-2009), both of which resulted in powerful central bank responses to ensure functioning markets.
1. Global Financial Crisis: First order impact on the financial sector
The global financial crisis had a first order impact on the financial sector because it was at the center of the crisis. In fact, the housing/mortgage bubble was enabled by the financial sector (thrifts, government sponsored entities or “GSEs” and leverage added by the investment banks). The bursting of this ~US$10 trillion asset class, which was owned globally, was a major cause of the global recession. The resultant losses overwhelmed the U.S. financial sector, creating insolvency risk and/or ultimately failures across a myriad of systemically important firms with trillion-dollar balance sheets – GSEs, giant thrifts, mega-cap banks, global investment banks, and the world’s largest insurer (AIG).
The factors contributing to the failures or government bailouts included: (i) credit losses (thrifts, banks, GSEs); (ii) a catastrophic loss of confidence creating liquidity-driven failures (investment banks funded in overnight repo market); and (iii) leverage losses from derivatives exposure (AIG, investment banks).
How did the crisis end? Ultimately ending the crisis required the nationalization of credit losses in the financial sector (Fannie Mae, Freddie Mac, FDIC auctions) and for the government to become the capital provider of last resort (TARP). These two steps provided capital to absorb the losses, which reduced the risk of insolvency and contagion. Obviously, there were other steps taken to support the economy including significant monetary and fiscal policy steps, which both directly and indirectly supported the financial sector. During this crisis, the U.S. financials index declined over 80%. Because of failures and mandatory recapitalizations resulting in catastrophic EPS dilution, the index did not recover to its pre-crisis high for over 10 years!
2. European Sovereign Debt Crisis: Second order impact on the financial sector
The European sovereign debt crisis had a second order impact, meaning that the potential for sovereign defaults in large southern European economies created insolvency risk within their banking sectors. Following the creation of the common currency (euro) in 1999, many southern European countries saw their cost of borrowing decline significantly for many reasons (including lower risk of currency depreciation). This precipitated a massive and unsustainable build up in sovereign debt for several countries. While Greece was the country initially at the centre, contagion rapidly spread to Ireland, Portugal, and more ominously, to the much larger economies of Spain and Italy.
This risk of sovereign default began to create a negative feedback loop, including the possibility of direct losses on sovereign bonds held within each countries’ domestic banking sector, which were (and remain) substantial holders of their own countries’ debt. This crisis enveloped some of the largest banks in the world, with trillion-dollar balance sheets.
Europe’s significance to the global economy – accounting for one-third of global GDP and a banking sector among the largest in the world (total assets of ~€25 trillion) – made this a global event/crisis. Moreover, the origins of this crisis were much more complicated than a simple asset bubble bursting because the eurozone’s single currency prevented the most impacted countries from using currency devaluation as a tool to mitigate the negative impact.
How did the crisis end? The resolution of this crisis was due to a combination of several actions, including nationalization of losses (Spain), intergovernmental fiscal transfers (eurozone/Germany to Greece), and aggressive central bank action, primarily by the ECB. Central banks took two critical steps to help end the crisis. First, massive bond buying pushed down sovereign yields and preserved access to sovereign debt financing (breaking the negative feedback loop). Second, providing the banks with access to low cost funding facilitated carry trades that generated earnings/revenues used to re-liquify/recapitalize the banks. Throughout this crisis, the European bank index declined almost 50% (from its peak) over a period of 19 months and took another 31 months to reach its pre-crisis high.
How does COVID-19 compare to these crises?
In our view, the current correction should not be considered a financial crisis since the market does not believe large systemically important institutions are at a risk of failure as a result of events. Rising insolvency risk (and/or loss of market confidence) was a key ingredient in both the global financial crisis (which experienced a significant number of failures and/or catastrophic dilution) and the European sovereign debt crisis (where failures, but not dilutive capital raises, were averted).
One reason market confidence has been retained is that the global financial sector has essentially doubled its loss-absorbing capital in the last decade while also adding significant liquidity. These measures give the sector the strength to absorb losses under some highly stressed scenarios.
COVID-19: An Emerging Credit Cycle
So, if not a crisis, then does COVID-19 represent a growth scare or a credit cycle? Despite the share price volatility in recent years, loan losses have generally remained stable, at below long-term trends, as prior corrections relating to yuan devaluation (2015-2016), Brexit (2016), and yield curve inversion (2018) represented growth scares and did not correctly foreshadow a material slowdown in GDP growth and decline in financial sector earnings.
However, this is not going to be the case for COVID-19.
The steps taken to mitigate the spread of the virus will obviously negatively impact financial sector earnings. The lower earnings will result from both negative GDP growth (higher credit losses, lower volume growth/market activity, higher insurance claims) and the impact of the term structure of interest rates (lower net interest margins/asset yields, higher actuarial reserves).
Most credit cycles are characterized by large losses in a single loan category. What makes this looming credit cycle so different and potentially more severe is that loan losses are not likely to be confined to one loan category. Because the steps taken to limit the spread of the virus are impacting all businesses and consumers, losses will probably be broad-based. That said, with the support of the government, global banks are also likely to take steps not seen before, including loan deferments, interest-only and other pro-active measures to bridge the gap from the quarantines to a resumption of economic activity. Nevertheless, a rise in bankruptcies is inevitable.
Given the wide range of possible growth outcomes, the market has very abruptly priced in a severe recession as its baseline scenario with declines in financial sector indices of 40-50% in less than six weeks.
Hiding in plain sight; capital forbearance
There are variables that are not normally part of the regulatory “tool kit” that could be significant to investors, including the potential for capital forbearance.
The global financial sector has never been better capitalized. Given these extremely high capital levels, the sector can absorb unprecedented losses under highly stressed scenarios without creating insolvency risk. In fact, we believe it is more likely bank regulators and policymakers seek to “spend” this capital in the form of “de-facto” regulatory forbearance – i.e., allowing minimum capital levels to decline from their current high levels to facilitate a recovery in the economy through lending. Most governments’ focus this downturn will be on working with the financial sector to provide a bridge to consumers and various stressed sectors, like energy, airlines, and hotels to get through this turbulent period.
Which financials will fare the best? Three variables to consider
In our view, the financials that will fare the best during this downturn had a combination of characteristics before the virus roiled the markets:
- Firms operating in developed economies with higher forecast GDP growth (before the virus) generally have lower recession risk. For example, financials operating in countries where forecast GDP growth was safely above 2.0% (including Australia, U.S. Southeast/Southwest, Nordic countries) are likely to fare better than those in countries where pre-virus forecast GDP was below 1.5% (including Canada, U.S. Mid-Atlantic/Midwest, Europe and Japan). Emerging markets with less developed/robust healthcare systems are likely to face greater difficulty coping with a virus outbreak, regardless of pre-virus expected GDP growth (e.g., India, Latin America, Central and Eastern Europe).
- Firms with higher EPS growth forecasts will have more “cushion” compared to those with lower EPS growth, although certain industries will be more impacted than others. Business mix will also be important, as some fee businesses could have a counter-cyclical impact.
- Firms with higher ROEs will also fare better as they generate greater earnings/capital to absorb any rapid rise in credit losses. One quarter’s earnings (net of dividends) is the first cushion for unexpected losses, so firms with higher profitability and capital generation are in a better position to absorb an abrupt rise in losses. Higher capital levels and excess capital are also mitigants to the potential for losses to reduce core capital.
What is the market pricing in and how might this correction end?
In the coming weeks, as we approach first quarter earnings season, analysts will update their earnings estimates to reflect the impact of the virus. In the meantime, the markets have clearly priced in a very severe outcome, with significant declines in price-to-earnings (“P/E”) and book value multiples. In fact, a number of financials now trade below tangible book value (“TBV”) – i.e., net asset value – implying the market believes those firms could lose money and/or issue equity (note: no Hamilton ETFs trade at a portfolio-weighted TBV below 1.0x).
Using our ETFs as examples, the table below highlights that P/Es have declined dramatically and indiscriminately versus their 5-year averages. Despite Hamilton ETFs owning some of the best financials in the world, with strong balance sheets, track records of stable dividends and operating in higher growth economies/regions, two of our ETFs have experienced P/E compression of 4.5 multiple points or higher (HBG, HFMU.U), while the other three have seen declines of over 3.0 multiple points (HFA, HFY, HCB).
Coming out of this correction, we expect the financials that outperform will be those that experienced the greatest multiple expansion or are more resistant to EPS downgrades.
When will this correction end?
The market, and financial stocks, will likely begin to rebound once investors are confident the negative impact of the virus has peaked, and/or the situation has improved enough for governments to allow for a resumption of economic activity. The sooner economic activity resumes, the less damaging the impact on the financial sector and the faster its recovery. The most likely indicator for this would be a deceleration of the daily increase in infections or fatalities or other evidence of “flattening the curve”. It is also possible that a restarting of the economy is accompanied with aggressive measures designed to protect the most vulnerable. It is reasonable to assume these dates will differ between countries and hence some markets could start to recover before others. An important development, in our view, is the recent comment from the U.S. President about his goal of restarting the U.S. economy by Easter (April 12).
More will be known in the coming weeks, including how successful each country is in “flattening the curve”.
A word on trading liquidity for ETFs …
Hamilton ETFs are a highly liquid ETF that can be purchased and sold easily. ETFs are as liquid as their underlying holdings and the underlying holdings trade millions of shares each day.
How does that work? When ETF investors are buying (or selling) in the market, they may transact with another ETF investor or a market maker for the ETF. At all times, even if daily volume appears low, there is a market maker – typically a large bank-owned investment dealer – willing to fill the other side of the ETF order (at net asset value plus a spread). The market maker then subscribes to create or redeem units in the ETF from the ETF manager (e.g., Hamilton ETFs), who purchases or sells the underlying holdings for the ETF.
 During these three credit downturns, provision ratios for the Canadian banks peaked at progressively lower levels (~130, ~100 and ~80 bps respectively). Of interest, peak losses from the TMT credit cycle were higher than those of the global financial crisis.
 Troubled Asset Relief Program. A program initiated by the U.S. government in 2008 to help stabilize the U.S. financial system and provide a boost to the U.S. economy. The most significant component to the banks was the investment of the U.S. government in financial institutions via preferred shares. https://www.treasury.gov/initiatives/financial-stability/TARP-Programs/Pages/default.aspx#
 S&P 500 Financials Net Total Return Index, from April 2007 to March 2009.
 As of March 24, 2020.