Capital Adequacy – Page 3 – From the outside

When safety proves dangerous …

… is the title of a post on the Farnham Street blog that provides a useful reminder of the problem of “risk compensation”; i.e. the way in which measures designed to make us safer can be a perverse prompt for us to take more risk because we feel safer. I want to explore how these ideas apply to bank capital requirements but will first outline the basic ideas covered by Farnham Street.

we all internally have a desired level of risk that varies depending on who we are and the context we are in. Our risk tolerance is like a thermostat—we take more risks if we feel too safe, and vice versa, in order to remain at our desired “temperature.” It all comes down to the costs and benefits we expect from taking on more or less risk.
The notion of risk homeostasis, although controversial, can help explain risk compensation.

The classic example is car safety measures such as improved tyres, ABS braking systems, seat belts and crumple zones designed to protect the driver and passengers. These have helped reduce car fatality rates for the people inside the car but not necessarily reduced accident rates given that drivers tend to drive faster and more aggressively because they can. Pedestrians are also at greater risk.

Farnham Street suggests the following lessons for dealing with the problem risk compensation:

Safety measures are likely to be more effective is they are less visible
Measures designed to promote prudent behaviour are likely to be more effective than measures which make risky behaviour safer
Recognise that sometimes it is better to do nothing if the actions we take just leads to an offset in risk behaviour somewhere else
If we do make changes then recognise that we may have to put in place other rules to ensure the offsetting risk compensating behaviour is controlled
Finally (and a variation on #3), recognise that making people feel less safe can actually lead to safer behaviour.

If you are interested in this topic then I can also recommend Greg Ip’s book “Foolproof” which offers a good overview of the problem of risk compensation.

Applying these principles to bank capital requirements

The one area where I would take issue with the Farnham Street post is where it argues that bailouts and other protective mechanisms contributed to scale of the 2008 financial crisis because they led banks to take greater risks. There is no question that the scale of the crisis was amplified by the risks that banks took but it is less obvious to me that the bailouts created this problem.

The bailouts were a response to the problem that banks were too big to fail but I can’t see how they created this problem; especially given that the build up of risk preceded the bailouts. Bailouts were a response to the fact that the conventional bankruptcy and restructure process employed to deal with the failure of non-financial firms simply did not work for financial firms.

It is often asserted that bankers took risks because they expected that they would be bailed out; i.e/ that banks deliberately and consciously took risk on the basis that they would be bailed out. I can’t speak for banks as a whole but I have never witnessed that belief in the four decades that I worked in the Australian banking system. Never attribute to malice what can be equally explained by mistaken beliefs. I did see bankers placing excessive faith in the economic capital models that told them they could safely operate with reduced levels of capital. That illusion of knowledge and control is however a different problem altogether, largely to do with not properly understanding the distinction between risk and uncertainty (see here and here).

If I am right, that would suggest that making banks hold more capital might initially make them safer but might also lead to banks looking for ways to take more risk. This is a key reason why I think the answer to safer banks is not just making them hold higher and higher levels of common equity. More common equity is definitely a big part of the answer but one of the real innovations of Basel 3 was the development of new forms of loss absorbing capital that allow banks to be recapitalised by bail-in rather than bail-out.

If you want to go down the common equity is the only solution path then it will be important to ensure that Farnham Street Rule #4 above is respected; i.e. bank supervisors will need to ensure that banks do not simply end up taking risks in places that regulation or supervision does not cover. This is not a set and forget strategy based on the idea that increased “skin in the game” will automatically lead to better risk management.

Based on my experience, the risk of common equity ownership being diluted by the conversion of this “bail-in” capital is a far more effective constraint on risk taking than simply requiring banks to hold very large amounts of common equity. I think the Australian banking system has this balance about right. The Common Equity Tier 1 requirement is calibrated to a level intended to make banks “Unquestionably Strong”. Stress testing suggest that this level of capital is likely to be more than sufficient for well managed banks operating with sensible risk appetites but banks (the larger ones in particular) are also required to maintain a supplementary pool of capital that can be converted to common equity should it be required. The risk that this might be converted into a new pool of dilutive equity is a powerful incentive to not push the boundaries of risk appetite.

Tony – From the Outside

Bank dividends

Matt Levine’s “Money Stuff” column (Bloomberg) offers some interesting commentary on what is happening with bank dividends in the US. Under the sub heading “People are worried about dividends” he writes:

So, again, I am generally pretty impressed by the performance of bank regulation in the current crisis, but this is unfortunate:
US banks’ annual capital plans, due to be submitted to the Federal Reserve on Monday, are expected to include proposals to continue paying dividends, reinforcing comments from prominent bank chief executives in recent days, according to people familiar with the situation.
The bankers, including Goldman Sachs boss David Solomon, Morgan Stanley boss James Gorman and Citigroup chief Mike Corbat, argued that they had the means to continue paying dividends and that cutting them would be “destabilising to investors”.
“We’re in a very different position than what we see in Europe,” said Marty Mosby, a veteran banks analyst at Vining Sparks.
“How we set it up [post-crisis capital requirements] was to be able to not have those dividends collapse [in a crisis]. That’s what creates a financial crisis: when dividends start to be ratcheted lower that shakes confidence.”
What is unfortunate is not so much that U.S. banks want to continue paying dividends; for all I know some of them are so well capitalized and so well equipped to weather this crisis that they will actually make a lot of money and have plentiful profits to pay out to shareholders. What is unfortunate is that their explicit view is that cutting dividends would be destabilizing. Common shareholders are supposed to be the lowest-ranking claimants on a bank’s money. The point of equity capital is that you don’t have to pay it out, that it doesn’t create any cash drain in difficult times. But if your view is “we need to maintain our dividend every quarter or else there will be a run on the bank,” then that means that the dividend is destabilizing; it means that your common stock is really debt; it means that your equity capital is not as good—not as equity-like—as it’s supposed to be.
If you take seriously the claim that banks can’t cut dividends in a generational crisis, for fear of undermining investor confidence, then, fine, I guess, but then the obvious conclusion is that when times are good you can never let banks raise their dividends. Every time a bank raises its dividend, on this theory, it incurs more unavoidable quarterly debt and creates a new drain on its funding, one that can’t be turned off in the bad times for fear of being “destabilising to investors”
Bloomberg Opinion “Money Stuff” 7 April 2020

I get the argument that if banks have the means to pay a dividend then they should be free to make a commercial decision. People may however feel entitled to be skeptical given the ways in which some banks were slow to adjust to the new realities of the GFC. There is also a line where the position some US banks appear to be projecting risks becoming an expectation that the dividend should be stable even under a highly stressed and uncertain outlook. It is not clear if that is exactly what the US banks quoted in his column are saying but that is how Matt Levine frames it and it would clearly be a concern if that is their view. That does seem to a fair description of the view some investors and analysts are expressing.

Jamie Dimon seems to be offering a more nuanced perspective on this question. He has advised JP Morgan shareholders that the Board expects the bank to remain profitable under its base base projections but would consider suspending the dividend under an extremely adverse scenario.

Our 2019 pretax earnings were $48 billion – a huge and powerful earnings stream that enables us to absorb the loss of revenues and the higher credit costs that inevitably follow a crisis. For comparison, the Comprehensive Capital Analysis and Review (CCAR) results for 2020 that we submitted to the Federal Reserve in 2019 (which assumed outcomes like U.S. unemployment peaking at 10% and the stock market falling 50%) showed a decline in revenue of almost 20% and credit costs of approximately $20 billion more than what we experienced in 2019. We believe we would perform better than this if the Fed’s scenario were to actually occur. But even in the Fed’s scenario, we would be profitable in every quarter. These stress test results also show that following such a meaningful reduction in our revenue (and assuming we continue to pay dividends), our common equity Tier 1 (CET1) ratio would likely hold at a very strong 10%, and we would have in excess of $500 billion of liquid assets.
Additionally, we have run an extremely adverse scenario that assumes an even deeper contraction of gross domestic product, down as much as 35% in the second quarter and lasting through the end of the year, and with U.S. unemployment continuing to increase, peaking at 14% in the fourth quarter. Even under this scenario, the company would still end the year with strong liquidity and a CET1 ratio of approximately 9.5% (common equity Tier 1 capital would still total $170 billion). This scenario is quite severe and, we hope, unlikely. If it were to play out, the Board would likely consider suspending the dividend even though it is a rather small claim on our equity capital base. If the Board suspended the dividend, it would be out of extreme prudence and based upon continued uncertainty over what the next few years will bring.
It is also important to be aware that in both our central case scenario for 2020 results and in our extremely adverse scenario, we are lending – currently or plan to do so – an additional $150 billion for our clients’ needs. Despite this, our capital resources and liquidity are very strong in both models. We have over $500 billion in total liquid assets and an incremental $300+ billion borrowing capacity at the Federal Reserve and Federal Home Loan Banks, if needed, to support these loans, as well as meet our liquidity requirements (these numbers do not include the potential use of some of the Fed’s newly created facilities). We could, of course, make our capital and liquidity buffer better by restricting our activities, but we do not intend to do that – our clients need us.
JP Morgan Chairman and CEO Letter to Shareholders 2019 Annual Report

Banks are cyclical investments – who knew?

Stress testing models must of course be treated with caution but what I think this mostly illustrates is that banks are highly cyclical investments. That may seem like a statement of the obvious but there was a narrative post GFC that banks were public utilities and that bank shareholders should expect to earn public utility style returns on their investments.

There is an element of truth in this analogy in so far as banks clearly provide an essential public service. I am also sympathetic to the argument that banking is a form of private/public partnership. This pandemic is however a timely reminder of the limits of the argument that banks are just another low risk utility style of business. Bank shareholders are much more exposed to the cyclical impacts than true utility investments.

In the interests of full disclosure, I have a substantial exposure to bank shares and I for one need a lot more than a single digit return to compensate for the pain that part of my portfolio is currently experiencing. The only upside is that I never bought into the thesis that banks are a low risk utility style investment requiring a commensurately low return.

The higher capital and liquidity requirements built up in response to the lessons of the GFC increase the odds that banks will survive the crisis and be a big part of the solution but banks are, and remain, quintessentially cyclical investments and the return bank investors require should reflect this. I think the lesson here is not to worry about the extent to which dividend cuts would be destabilising to investors but to focus on what kind of return is commensurate with the risk.

I will let APRA have the final say on what to expect …

APRA expects ADIs and insurers to limit discretionary capital distributions in the months ahead, to ensure that they instead use buffers and maintain capacity to continue to lend and underwrite insurance. This includes prudent reductions in dividends, taking into account the uncertain outlook for the operating environment and the need to preserve capacity to prioritise these critical activities.
Decisions on capital management need to be forward-looking, and in the current environment of significant uncertainty in the outlook, this can be very challenging. APRA is therefore providing Boards with the following additional guidance.²
During at least the next couple of months, APRA expects that all ADIs and insurers will:
– take a forward-looking view on the need to conserve capital and use capacity to support the economy;
– use stress testing to inform these views, and give due consideration to plausible downside scenarios (periodically refreshed and updated as conditions evolve); and
– initiate prudent capital management actions in response, on a pre-emptive basis, to ensure they maintain the confidence and capacity to continue to lend and support their customers.
During this period, APRA expects that ADIs and insurers will seriously consider deferring decisions on the appropriate level of dividends until the outlook is clearer. However, where a Board is confident that they are able to approve a dividend before this, on the basis of robust stress testing results that have been discussed with APRA, this should nevertheless be at a materially reduced level. Dividend payments should be offset to the extent possible through the use of dividend reinvestment plans and other capital management initiatives. APRA also expects that Boards will appropriately limit executive cash bonuses, mindful of the current challenging environment.
“APRA issues guidance to authorised deposit-taking institutions and insurers on capital management”, 7 April 2020

Tony (From the Outside)

ECB acknowledges the potential for IFRS 9 to amplify procyclicality

This ECB press release lists four initiatives to deal with impact of Covid 19

ECB gives banks further flexibility in prudential treatment of loans backed by public support measures
ECB encourages banks to avoid excessive procyclical effects when applying the IFRS 9 international accounting standard
ECB activates capital and operational relief measures announced on March 12, 2020
Capital relief amounts to €120 billion and could be used to absorb losses or potentially finance up to €1.8 trillion of lending

This guidance on flexibility is helpful (arguably necessary) but it would have been better if the relationship between loan loss provisioning and capital buffers was more clearly thought through and built into the design of the system before it was subject to its first real test.

Tony

IFRS 9 loan loss provisioning faces its first real test

My long held view has been that IFSR 9 adds to the procyclicality of the banking system (see here, here, and here) and that the answer to this aspect of procyclicality lies in the way that capital buffers interact with loan loss provisioning (here, here, and here).

So it was interesting to see an article in the Financial Times overnight headlined “New accounting rules pose threat to banks amid virus outbreak”. The headline may be a bit dramatic but it does draw attention to the IFRS 9 problem I have been concerned with for some time.

The article notes signs of a backlash against the accounting rules with the Association of German Banks lobbying for a “more flexible handling” of risk provisions under IFRS 9 and warning that the accounting requirements could “massively amplify” the impact of the crisis. I agree that the potential exists to amplify the crisis but also side with an unnamed “European banking executive” quoted in the article saying “IFRS 9, I hate it as a rule, but relaxing accounting standards in a crisis just doesn’t look right”.

There may be some scope for flexibility in the application of the accounting standards (not my area of expertise) but that looks to me like a dangerous and slippery path to tread. The better option is for flexibility in the capital requirements, capital buffers in particular. What we are experiencing is exactly the kind of adverse scenario that capital buffers are intended to absorb and so we should expect them to decline as loan loss provisions increase and revenue declines. More importantly we should be seeing this as a sign that the extra capital put in place post the GFC is performing its assigned task and not a sign, in and of itself, indicating distress.

This experience will also hopefully reinforce the case for ensuring that the default position is that the Counter Cyclical Capital Buffer be in place well before there are any signs that it might be required. APRA announced that it was looking at this policy in an announcement in December 2019 but sadly has not had the opportunity to fully explore the policy initiative and implement it.

Tony

Capital Rules Get Less Stressful – Matt Levine

Nice quote from Matt Levine’s opinion piece on the change in US bank capital requirements

“Everything in bank capital is controversial so this is controversial. Usually the controversy is that some people want higher capital requirements and other people want lower capital requirements. Here, pleasantly, part of the controversy is about whether this is a higher or lower capital requirement.“

https://www.bloomberg.com/opinion/articles/2020-03-05/capital-rules-get-less-stressful

Confusing capital and liquidity

I have been planning to write something on the relationship between capital and liquidity for a while. I have postponed however because the topic is complex and not especially well understood and I did not want to contribute to the body of misconceptions surrounding the topic. An article in the APRA Insight publications (2020 Issue One) has prompted me to have a go.

Capital Explained

The article published in APRA’s Insight publication under the title “Capital explained” offers a simple introduction to the question what capital is starting with the observation that …

“Capital is an abstract concept and has different meanings in different contexts.

Capital being abstract and meaning different things in different contexts is a good start but the next sentence troubles me.

“In non-technical contexts, capital is often described as an amount of cash or assets held by a company, or an amount available to invest.”

I am not sure that the author intended to endorse this non-technical description but it was not clear and I don’t think it should be left unchallenged, especially when the casual reader might be inclined to take it at face value. The fact that non-technical descriptions frequently state this is arguably a true statement but the article does not clarify that this description is a source of much confusion and seems to be conflating capital and liquid assets.

The source of the confusion possibly lies in double entry bookkeeping based explanations in which a capital raising will be associated with an influx of cash onto a company balance sheet. What happens next though is that the company has to decide what to do with the cash, it is extremely unlikely that the cash just sits in the company bank account. This is especially true in the case of a bank which has cash flowing into, and out of, the balance sheet every day. The influx of one source of funding (in this case equity) for the bank means that it will most likely choose to not raise some alternate form of funding (debt) on that day. The amount of cash it holds will be primarily driven by the liquidity targets it has set which are related to but in no way the same thing as its capital targets.

Time for me to put up or shut up.

How should we think about the relationship between capital and liquidity including the extent to which holding more liquid assets might, as is sometimes claimed, justify holding less capital.

Liquidity risk is mitigated by liquidity management, including holding liquid assets, but this statement offers no insight into the extent to which some residual expected or unexpected aspect of the risk still requires capital coverage (All risks are mitigated to varying extents by management but most still require some level of capital coverage)
So the assessment that holding more liquid assets reduces the need to hold capital is open to challenge
One of the core functions of capital is to absorb any increase in expenses, liabilities, loan losses or asset write downs associated with or required to resolve an underlying risk issue; the bank may need to recapitalise itself to restore the target level of solvency required to address future issues but the immediate problems are resolved without the consumption of capital compromising solvency
Liquid assets, in contrast, buy time to resolve problems but they do not in themselves solve any underlying issues that may be the root cause of the liquidity stress.
The relationship between liquidity and solvency is not symmetrical; liquidity is ultimately contingent on a bank being solvent, but a solvent bank can be illiquid.
While more liquid assets are not a substitute for holding capital, a more strongly capitalised bank is less likely to be subject to the kinds of liquidity stress events that draw on liquid assets so holding more capital relative to peer banks can reduce liquidity risk
Being relatively strong matters in scenarios where uncertainty is high and people resort to simple rules (e.g. withdraw funding from the weakest banks; even if that is not true the risk is that other people express that view and it becomes self-fulfilling)
It is important to recognise that the focus of relationship between capital and liquidity risk described above is the capital position relative to peer banks and market expectations, not the absolute stand-alone position
The “Unquestionably Strong” benchmark used in the Australian banking system to calibrate the overall target operating range for capital in the ICAAP anchors the bank’s Liquidity Risk appetite setting.
Expressed another way, the capital requirements of Liquidity Risk are embedded holistically in the capital buffer the target operating range maintains over prudential minimum capital requirements.

It is entirely possible that I am missing something here – I hope not but let me know if you see an error in my logic

Using machine learning to predict bank distress

Interesting post on the Bank Underground blog by Bank of England staff Joel Suss and Henry Treitel.

This extract summarises their findings

“Our paper makes important contributions, not least of which is practical: bank supervisors can utilise our findings to anticipate firm weaknesses and take appropriate mitigating action ahead of time.

However, the job is not done. For one, we are missing important data which is relevant for anticipating distress. For example, we haven’t included anything that speaks directly to the quality of a firm’s management and governance, nor have we included any information on organisational culture.

Moreover, our period of study only covers 2006 to 2012 – a notoriously rocky time in the banking sector. A wider swathe of data, including both good times and bad, would help us be more confident that our models will perform well in the future.

So while prediction, especially about the future, remains tough, our research demonstrates the ability and improved clarity of machine learning methodologies. Bank supervisors, armed with high-performing and transparent predictive models, are likely to be better prepared to step-in and take action to ensure the safety and soundness of the financial system.”

Possible pitfalls of a 1-in-X approach to financial stability – Bank Underground

Bank Underground is a blog for Bank of England staff to share views that challenge – or support – prevailing policy orthodoxies. The views expressed are those of the authors, and are not necessarily those of the Bank of England, or its policy committees. Posting on this blog, Adam Brinley Codd and Andrew Gimber argue that false confidence in people’s ability to calculate probabilities of rare events might end up worsening the crises regulators are trying to prevent.

The post concludes with their personal observations about how best to deal with this meta-uncertainty.

Policymakers could avoid talking about probabilities altogether. Instead of a 1-in-X event, the Bank of England’s Annual Cyclical Scenario is described as a “coherent ‘tail risk’ scenario”.

Policymakers could avoid some of the cognitive biases that afflict people’s thinking about low-probability events, by rephrasing low-probability events in terms of less extreme numbers. A “100-year” flood has a 1% chance of happening in any given year, but anyone who lives into their 70s is more likely than not to see one in their lifetime.

Policymakers could be vocal about the fact that there are worse outcomes beyond the 1-in-X point of the distribution.

— Read on bankunderground.co.uk/2020/02/06/possible-pitfalls-of-a-1-in-x-approach-to-financial-stability/

Capital geeks take note – the IRB scaling factor strikes again

Capital

Another reminder on the importance of paying attention to the detail courtesy of a story that I picked up reading Matt Levine’s “Money Stuff” column on Bloomberg.

This extract from Matt’s column captures the essential facts:

Here, from Johannes Borgen, is a great little story about bank capital. Yesterday Coventry Building Society, a U.K. bank, announced “a correction to its calculation of risk weighted assets” that will lower its common equity Tier 1 capital ratio from 34.2% to 32.6%. That’s still well over regulatory requirements, so this is not a big deal. But the way Coventry messed up is funny:
“The Society uses Internal Ratings Based (“IRB”) models to calculate its Risk Weighted Assets (“RWAs”) and is seeking to update these models to ensure compliance with upcoming Basel III reforms. During the process of transitioning models, the Society has identified an omission in connection with its historic calculation of its RWAs. Specifically, the necessary 6% scalar was not applied to the core IRB model outputs. The core IRB models themselves are not impacted.”
For banks that use Internal Ratings Based models, the way the Basel capital rules work is that you apply a complicated formula to calculate the risk weights of your assets, and then at the end of the formula you multiply everything by 1.06. That’s kind of weird. (The Basel capital regime for banks using IRB models “applies a scaling factor in order to broadly maintain the aggregate level of minimum capital requirements, while also providing incentives to adopt the more advanced risk-sensitive approaches.”) It’s weird enough that in the “upcoming Basel III reforms” regulators plan to get rid of it: The 1.06 multiplier is a kludge, and if you measure your risk-weighted assets a bit more accurately and conservatively, you shouldn’t have to multiply them by 1.06 at the end.
Matt Levine, “Money Stuff”, Bloomberg

For anyone new to this game who wants to dig a bit deeper into how the advanced capital requirements are calculated, the Explanatory Note published by the BCBS in July 2005 is still a good place to start. I published a note on that paper on my blog here. The RBNZ also produced a useful note on how they used the IRB function in the portfolio modelling work they used to support their recent changes to NZ capital requirements.

It should be noted however that none of these documents discuss the 6% scaling factor. I open to alternative perspectives on this but my recollection is that the 6% scaling factor was introduced post July 2005 in one of the multiple recalibration exercises the BCBS employed to ensure that the IRB function did not reduce capital requirements too much relative to the status quo operating under Basel I. It is effectively a “fudge” factor designed to produce a number the BCBS was comfortable with (at that time).

Tony

APRA announces that it will consider a non-zero default level for the counter cyclical capital buffer

The Australian Prudential Regulation Authority (APRA) announced today that it had decided to keep the countercyclical capital buffer (CCyB) for authorised deposit-taking institutions (ADIs) on hold at zero per cent. What was really interesting however is that the information paper also flagged the likelihood of a non-zero default level in the future.

Here is the relevant extract from the APRA media release:

…. the information paper notes that APRA is also giving consideration to introducing a non-zero default level for the CCyB as part of its broader reforms to the ADI capital framework.

APRA Chair Wayne Byres said: “Given current conditions, and the financial strength built up within the banking sector, a zero counter-cyclical buffer remains appropriate.

“However, setting the countercyclical capital buffer’s default position at a non-zero level as part of the ‘unquestionably strong’ framework would not only preserve the resilience of the banking sector, but also provide more flexibility to adjust the buffer in response to material changes in financial stability risks. This is something APRA will consult on as part of the next stage of the capital reforms currently underway.

“Importantly, this would be considered within the capital targets previously announced – it does not reflect any intention to further raise minimum capital requirements.”
“APRA flags setting the countercyclical capital buffer at non-zero level”, APRA media announcement, 11 December 2019

I have argued the case for a non-zero default setting on this buffer in a long form note I published on my blog here, and published some shorter posts on the countercyclical capital buffer here, here and here). One important caveat is is that incorporating a non-zero default for the CCyB does not necessarily means that a bank needs to hold more capital. It is likely to be sufficient to simply partition a set amount of the existing capital surplus. In this regard, it is interesting that APRA has explicitly linked this potential change to the review it it initiated in the August 2018 Discussion Paper on “Improving the transparency, comparability and flexibility of the ADI capital framework”.

I covered that discussion paper in some depth here but one of the options discussed in this paper (“Capital ratio adjustments”) involves APRA modifying the calculation of regulatory capital ratios to utilise more internationally harmonised definitions of capital and Risk Weighted Assets.

Summing up, I would rate this as a positive development but we need to watch how the policy development process plays out.

Tony