There is a common misconception that when banks trade at a discount to book value they must be cheap. Indeed, on this basis, the universe of global banks, as represented by the Bloomberg World Banks Index, must be cheap – as cheap as it has been at any time since the Global Financial Crisis (see Figure 1).
Figure 1: Bloomberg World Banks Index – Price to Book Ratio: 2005-2020
Source: Bloomberg, daily data from 01 January 2005 to 30 June 2020
A closer inspection of this chart reveals another interesting trend – the multiple to book that investors are prepared to pay for banks globally has been in a structural downtrend for at least the past ten years. Finance theory tells us that the multiple to book that an investor should be prepared to pay for an investment is a function of three key factors:
Since the Global Financial Crisis, a combination of factors has conspired to reduce returns and prospects for growth in the banking sector. These include ever-lower interest rates, pressuring bank margins, increased competition from built-for-purpose “fintech” start-ups, compressing and eroding fees across multiple business lines, cost pressures to adapt legacy systems and infrastructure to new channels of distribution, and a vastly increased regulatory and compliance burden. All of this has occurred within the context of a lacklustre global growth environment.
Investors have become more cognisant of the risks involved in entrusting capital to a (typically) highly-leveraged business model such as a bank, even while banks have been pressured to de-risk and reduce leverage by ever-more-stringent capital and liquidity requirements (the so-called “Basel 3” regulations).
Most recently, the substantial – though at this stage still highly uncertain – impact of the global COVID-19 pandemic on asset prices and credit quality, as well as the future trajectory of the economic recovery, have caused investors to question the integrity of book values, resulting in the sector trading at a deep discount. Nowhere have these trends been more apparent than in Europe, where the banking sector has languished at a deep discount for most of the past twelve years, now trading in aggregate at only 40% of book value (see Figure 2).
Figure 2: Euro Stoxx Banks Index – Price to Book Ratio: 2005-2020
Source: Bloomberg, daily data from 01 January 2005 to 30 June 2020
At the end of March this year, for the first time in recent history, the South African banking sector also traded in aggregate at a discount to book value (see Figure 3). Even at the height of the Global Financial Crisis (which the local banks, for reasons outside the scope of this paper, navigated comparatively well), the South African banking sector traded above book value. Due to the disparate ratings across the banks, the relatively modest aggregate discount to book value hides deeper discounts for certain stocks – Absa and Nedbank, for example, both trade at discounts of more than 20% to disclosed book value – yet all are trading at material discounts to their long-term histories.
Figure 3: South African Banks Index – Price to Book Ratio: 2005-2020
Source: Bloomberg, daily data from 01 January 2005 to 30 June 2020
A signal such as this should naturally alert investors to a potential investment opportunity. Indeed, had one had the prescience to purchase a basket of South African banks during the latter part of 2008 and early in 2009, when their price to book multiples were troughing, one would have been rewarded with solid absolute, and market-beating relative returns (see Figure 4).
Figure 4: South African Banks Index versus the All Share Index: 2005-2020
Source: Bloomberg, daily data from 01 January 2005 to 30 June 2020
If each cycle were the same, and history repeated itself in a predictable manner, investing would be comparatively easy. In order to interrogate this potential investment opportunity, a number of important questions need to be answered:
The epicentre of the Global Financial Crisis lay in the United States residential property market (and multiple derivatives thereof). While the repercussions were felt globally, much of the damage was limited to the banking sector, and its ability to continue to play the essential role of financial intermediation in the global economy. Banks failed, jobs were lost, and global growth suffered a short, yet sharp, hit.
Ancillary industries were negatively impacted by the subsequent global recession, yet the crisis remained fundamentally a banking one. An unprecedented, coordinated global fiscal and monetary response ensured a robust global recovery. Locally, the banks did not escape the rout, although the impact was contained, and as much to do with ill-considered and overly exuberant credit extension during the 2005-2008 credit cycle as the Global Financial Crisis itself.
There is little doubt that the current crisis, brought on by the COVID-19 pandemic, is significantly more widespread, both by geography and industry, than the Global Financial Crisis. Early indications are for a substantially more pronounced impact on growth and employment across multiple industries. While this has again elicited a massive global monetary and (where possible) fiscal response, there are valid concerns over the ability of governments the world over to orchestrate another recovery akin to that post the Global Financial Crisis.
While the current environment is characterised by a high degree of uncertainty, most South African banks anticipate that the impact on credit quality may surpass that of the previous cycle, potentially by some margin. The parlous state of the South African government’s finances precludes the extent of fiscal assistance being provided elsewhere in the world. Recessionary conditions going into the crisis only serve to compound the problem. One mitigating factor may be the somewhat more conservative approach to credit extension adopted by the banks over the past few years.
The downside risks to the local banking sector are manifold. The 275 basis points in rate cuts thus far this year will wipe billions of Rands off margin income made from lending; this will be compounded further by muted, and in some categories possibly even negative, credit growth. The economy-wide lockdowns that have been in place since the end of March have decimated many of the annuity fee revenue streams of the banks, as activity levels across branches, points-of-sale and ATMs have declined by more than half. Corporate advisory, insurance and other transaction-related sources of fees have been similarly impacted, with only trading income a bright spot, courtesy of extreme volatility in foreign exchange and rates markets. The extent of the recovery post lockdown remains in question, as consumer behaviour remains cautious, and corporates delay major investment decisions in favour of balance sheet security. Operating costs remain a key focus to try to protect the bottom line, yet recent events have also given rise to incremental costs, as the banks have implemented measures to protect staff and customers, while transitioning thousands of employees to work from home.
By far the biggest risk facing the banks is credit quality. In the context of current projections around the decline in GDP, extent of business failures and job losses, it is very plausible that a “base case” scenario for credit losses in this cycle exceeds the peak of the Global Financial Crisis, and under a “bear case”, exceeds it by some margin. Forecast risk is extremely high, with much depending on the timing and nature of the emergence from lockdown, the ability of the South African government to negotiate unprecedented fiscal challenges, and the extent of support from the global economy. While our central scenario is for the South African banking sector to remain profitable and adequately capitalised in aggregate in 2020, we remain concerned about the very real downside risks.
We have subjected the balance sheets of the banks to multiple stress tests to determine the likelihood of breaching their minimum regulatory Core Tier 1 requirements. While all the banks appear adequately capitalised, FirstRand and Standard Bank stand out due to their superior capital buffers. A little more than a doubling of our forecast bad debts charge for Absa and Nedbank would start to create capital adequacy issues, while in the case of FirstRand and Standard Bank, bad debts would need to amount to four times our forecast level (see Figure 5).
Figure 5: “Big 4” & Capitec South African Banks – Capital buffers under stressed impairment scenario
Source: Ninety One as at June 2020
During the Global Financial Crisis, the price to book ratios of the South African banks troughed in March 2009, although most maintained a rating of parity to book or greater (see Figure 6). A similar trough may have been reached in May of this year, albeit with ratings slightly lower (see Figure 2).
Figure 6: “Big 4” South African Banks - Price to Book trough during the Global Financial Crisis
Source: Bloomberg, 2008 and 2009
After the trough in 2009, the banks experienced a period of strong absolute and, over most time periods, market-relative returns. Investors were typically rewarded for investing in the banks at depressed multiples. However, it is apparent from Figure 7 that there is little correlation between the relative rating at trough multiples, and the relative performance thereafter.
In fact, over the ensuing five-year investment horizon, the best performing bank by some margin was the one that had troughed with the highest price to book multiple, whilst the worst performer had screened as second cheapest. The thesis to buy the “cheapest” bank, as measured by the deepest discount to book value, in the expectation of the highest return as ratings recover, was not vindicated in practice.
Figure 7: “Big 4” South African Banks - Total Returns in the Five Years post Price to Book trough
Source: Bloomberg, Based to 100 at start of March 2009
There are typically two reasons why a bank would maintain a higher rating even at trough levels – a lower risk premium being accorded to the bank providing the greatest protection in the event of a “bear case” scenario and/or superior prospects for long-term returns and growth post-crisis. Factors driving the former include capital and liquidity buffers, conservatism in provisioning and asset valuations on balance sheet, pre-provision profitability and, most importantly, franchise quality and management track record. The latter would be driven by return on equity, market share gains and ongoing investment for growth.
Post the Global Financial Crisis, FirstRand and Capitec’s ROEs far exceeded their peers (see Figure 8) and whilst we expect the ROE for the sector to be cyclically lower over the next three years, we still expect these two banks to deliver a superior ROE.
Figure 8: “Big 4” South African Banks – Return on Equity (ROE)
Source: Bloomberg and company financials, January 2005 to June 2020
In recognition of the potential investment opportunity in the South African banks, we currently have a neutral weight in the sector, and are selective in terms of our stock-specific exposures. Reflecting the uncertainty as to the depth and duration of the crisis, and the ability of the local economy to stage a credible recovery in 2021 and beyond, we have selected stocks based on balance sheet strength and ability to withstand a worse-than-expected outcome, as well as those stocks that we believe will be able to sustainably generate returns in excess of their cost of equity in a post-crisis environment. In this regard, our core positions are in FirstRand and Capitec/PSG.
Next article
The pandemic has accelerated key themes in mining such as declining mine production supply and heightened ESG risk.