Digital Finance Analytics (DFA) Blog
"Intelligent Insight"
Using a speech from the Governor of the Bank of England as a starting point, we examine three erroneous assumptions people make about the financial system, and why they lead to disaster.
Please consider supporting our work via Patreon
Please share this post to help to spread the word about the state of things….
We discuss the implications of the ACCC Cartel case against ANZ and others.
Let’s talk about the bank bail-in conundrum. A couple of weeks back I discussed whether bank deposits in Australia would be safe in a crisis. The video received more than 1,400 views so far, and has prompted a number of important questions from viewers. So today I update the story, and addresses some of the questions raised. The bottom line though is I think we are being sold a pup, which by the way refers to a confidence trick originating in the Late Middle Ages!
Watch the video, or read the transcript.
First, a quick recap, for those who missed the first video. Officially, in Australia currently bank deposits are protected up to $250,000 per person by a Government Guarantee – The Financial Claims Scheme. For banks, building societies and credit unions incorporated in Australia, the FCS provides protection to depositors up to $250,000 per account-holder per ADI according to APRA. Only deposit products provided by ADIs supervised by APRA were eligible to be covered. Amounts between $250,000 and $1 million are not be covered under the Guarantee Scheme. Above $1m banks can elect to pay a fee to the Government for this for protection, but none do. However, as we will see there are even questions about the sub $250k. But note this, the FCS can only be activated by the Australian Government, whilst APRA is responsible for administering the Scheme.
The RBA says upon its activation, APRA aims to make payments to account-holders up to the level of the cap as quickly as possible – generally within seven days of the date on which the FCS is activated. The method of payout to depositors will depend on the circumstances of the failed ADI and APRA’s assessment of the cost-effectiveness of each option. Payment options include cheques drawn on the RBA, electronic transfer to a nominated account at another ADI, transfer of funds into a new account created by APRA at another ADI, and various modes of cash payments.
The amount paid out under the FCS, and expenses incurred by APRA in connection with the FCS, would then be recovered via a priority claim of the Government against the assets of the ADI in the liquidation process. If the amount realised is insufficient, the Government can recover the shortfall through a levy on the ADI industry. Ok, maybe in the case of a single failure
Now, since the Global Financial Crisis, regulators have been working on ways to avoid a tax payer based rescue in a crash, because for example, the UK’s Royal Bank of Scotland was nationalised in 2007. This cost tax payers dear, so regulators want measures put in place to try to manage a more orderly transition when a bank gets into difficulty.
The New Zealand the Open Bank Resolution (not to be confused with Open Banking) is the clearest example of a so called bail-in arrangement. Customer’s money, held as savings in a distressed bank can be grabbed to assist in a resolution in a time of crisis. The thinking behind it is simple. Banks need an exit strategy in case of a problem, and Government bail-outs should not be an option. So a manager can be appointed to manage through the crisis. They can use bank capital, other instruments, like hybrid bonds and deposits to create a bail-in. This approach to rescuing a financial institution on the brink of failure makes its creditors and depositors take a loss on their holdings. This is the opposite of a bail-out, which involves the rescue of a financial institution by external parties, typically governments using taxpayer’s money.
So what about Australia? Well, the Financial Sector Legislation Amendment (Crisis Resolution Powers and Other Measures) Bill 2017 is now law, having been through a Senate Inquiry. It all centers on the powers which were to be given to APRA to deal with a banking collapse.
In the bill, there is a phrase “any other instrument” in the list of bail-in items. Treasury said, “the use of the word ‘instrument’ is intended to be wide enough to capture any type of security or debt instrument that could be included within the capital framework in the future. It is not the intention that a bank deposit would be an ‘instrument’ for these purposes”. Yet, deposits were not expressly excluded.
In fact, when the Bill came back to Parliament it went through both houses with minimal discussion (and members on the floor, the chambers were all but empty). And despite a proposal being drafted and with Government Lawyers in parallel to exclude deposits, it was passed on 14th February without this change, leaving the door wide open under “any other instrument”. All the verbal assurances are meaningless.
So, the result appears to be APRA has wider powers now to handle a bank in crisis, and deposits are potentially accessible. They are not expressly excluded, and in a time of crisis, could be bailed-in.
But this is not the end of the story. Treasurer Morrison issued a letter to Liberal government members with some talking points to justify this actions, in response to a wave of protests. But in so doing, he raises more questions.
The first point is that the Deposit Guarantee scheme (the one up to $250k) is not currently active. The Government would need to activate it, and can only do so when an institution fails. This is important because it means that in theory at least, APRA could mount a deposit bail-in before the Government activates the deposit protection scheme. Consider what would happen if many banks all got into difficulty at the same time, as could be the case in a wider banking crisis – after all, they all have similar banking models.
The second point is that the Treasurer makes reference to the 1959 Banking Act, and says that depositors have a claim above other creditors in a bank failure. But in fact the 1959 Act says depositors do indeed rank ahead of other unsecured creditors, but that means the secured creditors come first. So would anything be left in case of a bank failure given the massive exposure to property?
Next, the letter says APRA has now enhanced powers to protect the interests of depositors – not deposits. And looking at the New Zealand situation the bail-in provisions there are framed to do just this, by utilising deposits to help keep the bank afloat, thus protecting depositors. The Reserve Bank of New Zealand says this is IN THE INTERESTS OF DEPOSITORS.
Oh, and finally, Morrison says the way the Bill went into Law was quite normal by being listed in the Senate Order of Business, meaning members had the opportunity to debate the bill if they had wanted to. In fact, only seven Senators were there despite really needing a quorum of 19, but there is a get out in that a quorum is only needed if a division was called, and in this case it was simply nodded through. Democracy in action.
So there you have it. No Deposit Protection currently exists. Its limited to $250k per person if activated by the Government, at their discretion, and the legalisation leaves the door wide open for a New Zealand style of Bail-In. Not a good look.
So what should Savers do? Well, this is not financial advice, but the New Zealand view is that savers should make a risk assessment of banks and select where to deposit funds accordingly. But I am not sure how you do that, given the current low level of disclosures. APRA releases mainly aggregate data and protects the confidentially of individual banks as they are required to do under the APRA act.
Next, do not assume deposits are risk free, they are not. This means lenders should be offering rates of return more reflective of the risks we are taking, currently they are not (in fact deposit rates are sliding, as banks seek to repair margins). You might consider spreading the risks across multiple institutions
Consider alternative savings options (which are limited). Clearly, property, stocks and shares and even crypto currencies are all risky – there are no safe harbours. I guess there is always the mattress.
One other point to make. Several people are calling a bill to bring a Glass-Steagall split between core banking operations and the speculative aspects of banking. Glass-Steagall was enacted in the US in 1933 after the great crash, separating commercial and investment banking and preventing securities firms and investment banks from taking deposits. But in 1999 the US Congress passed the Gramm–Leach–Bliley Act, also known as the Financial Services Modernization Act, to repeal them. Eight days later, President Bill Clinton signed it into law. Following the financial crisis of 2007-2008, legislators unsuccessfully tried to reinstate Glass–Steagall Sections 20 and 32 as part of the Dodd–Frank Wall Street Reform and Consumer Protection Act. Both in the United States and elsewhere, banking reforms have been proposed that refer to Glass–Steagall principles. These proposals include issues of “ring fencing” commercial banking operations and narrow banking proposals that would sharply reduce the permitted activities of commercial banks.
The point of the bill was to isolate the risky bank behaviour, relating to derivatives and trading from core banking activities. In the case of a banking crisis, triggered by a collapse in the financial markets such an arrangement would protect the operations of the core banking. We got a glimpse of that a month ago when US trading volatility shot through the roof.
But, in Australia, the bulk of the risks in the banking system comes not from the derivatives side of the business, but the massive exposure to household debt and the property sector, and the risky loans they have made. We discussed this on the ABC yesterday. More than 60% of all banking assets are aligned with home lending, plus more relating to commercial property. Thus I do not believe a Glass-Stegall type separation would help to mitigate risks to the banking sector here much at all.
Better to push for a definitive change to the APRA Bill and get deposits excluded from the risk of bail-in. Or place a levy on all banks to directly protect depositors as has been put in place in Germany, where a dedicated government entity has been created for just this purpose.
What I find remarkable is that following loose banking regulation for years, during which the banks have returned massive profits to shareholders, and ramped up their risks, depositors are being lined up by the Government to bail out a failing bank. This is simply wrong.
Interesting speech by Fed Vice Chairman Fischer on supervisory stress testing of large systemic financial institutions.
Stress testing has become a cornerstone of a new approach to regulation and supervision of the largest financial institutions in the United States. The Federal Reserve’s first supervisory stress test was the Supervisory Capital Assessment Program, known as the SCAP. Conducted in 2009 during the depths of the financial crisis, the SCAP marked the first time the U.S. bank regulatory agencies had conducted a supervisory stress test simultaneously across the largest banking firms. The results clearly demonstrated the value of simultaneous, forward-looking supervisory assessments of capital adequacy under stressed conditions. The SCAP was also a key contributor to the relatively rapid restoration of the financial health of the U.S. banking system.
The Fed’s approach to stress testing of the largest and most systemic financial institutions has evolved since the SCAP, but several key elements persist to this day. These elements include, first, supervisory stress scenarios applicable to all firms; second, defined consequences for firms deemed to be insufficiently capitalized; and third, public disclosure of the results.
The Fed has subsequently conducted five stress test exercises that built on the success of SCAP, while making some important improvements to the stress test processes. The first key innovation was the development of supervisory models and processes that allow the Fed to evaluate independently whether banks are sufficiently resilient to continue to lend to consumers and to businesses under adverse economic and financial conditions. This innovation took place over the course of several exercises and was made possible by the extensive collection of data from the banks. These data have allowed supervisors to build models that are more sensitive to stress scenarios and better define the riskiness of the firms’ different businesses and exposures.
The second innovation since the SCAP was the use of the supervisory stress test as a key input into the annual supervisory evaluation of capital adequacy at the largest bank holding companies. The crisis demonstrated the importance of forward-looking supervision that accounted for the possibility of negative outcomes. By focusing on forward-looking post-stress capital ratios, stress testing provides an assessment of a firm’s capital adequacy that is complementary to regulatory capital ratios, which reflect the firm’s performance to date. Although we view this new approach to capital assessment as a significant improvement over previous practices, we are aware that the true test of this new regime will come only if another period of significant financial or economic stress were to materialize–which is to say that we will not have a strong test of the effectiveness of stress testing until the stress tests undergo a real world stress test. The same comment, mutatis mutandis, applies to the overall changes in methods of bank regulation and supervision made since September 15, 2008.
Third, supervisory stress testing has been on the leading edge of a movement toward greater supervisory transparency. Since the SCAP, the Fed has steadily increased the transparency around its stress testing processes, methodologies, and results. Before the crisis, releasing unfavorable supervisory information about particular firms was unthinkable–for fear of setting off runs on banks. However, the release of the SCAP results helped to calm markets during the crisis by reducing uncertainty about firm solvency. Indeed, only one of the 10 firms deemed to have a capital shortfall was unable to close the identified gap on the private markets. Our experience to date has been that transparency around the stress testing exercise improves the credibility of the exercise and creates accountability both for firms and supervisors. That said, too much transparency can also have potentially negative consequences, an idea to which I will turn shortly.
With the benefit of five years of experience, the Fed is continuing to assess its stress testing program, and to make appropriate changes. Examples of such changes to date include the assumption of default by each firm’s largest counterparty and the assumption that firms would not curtail lending to consumers and businesses, even under severely adverse conditions. As part of that assessment process, we are also currently seeking feedback from the industry, market analysts, and academics about the program.
Supervisory stress testing is not a static exercise and must adapt to a changing economic and financial environment and must incorporate innovations in modeling technology. Work is currently underway on adapting the stress testing framework to accommodate firms that have not traditionally been subject to these tests. The Dodd-Frank Act requires the Fed to conduct stress tests on non-bank financial institutions that have been designated as systemically important by the FSOC–the Financial Stability Oversight Council. Three of the currently designated financial institutions are global insurance companies. While distress at these firms poses risks to financial stability, particularly during a stressful period, certain sources of risk to these firms are distinct from the risks banking organizations face. A key aspect of this ongoing work includes adapting our current stress testing framework and scenarios to ensure that the tests for non-bank SIFIs–systemically important financial institutions–are appropriate.
Another area where work continues–and will likely always continue–is the Fed’s ongoing research aimed at improving our ability to estimate losses and revenues under stress. Supervisors have both to develop new approaches that push the state of the art in stress testing and to respond as new modeling techniques are developed or as firm activities and risk concentrations evolve over time. For example, forecasting how a particular bank’s revenue may respond to a severe macroeconomic recession can be challenging, and we continue to seek ways to enhance our ability to do so.
Supervisory stress testing models and methodologies have to evolve over time in order to better capture salient emerging risks to financial firms and the system as a whole. However, the framework cannot simply be expanded to include more and more aspects of reality. For example, incorporating feedback from financial system distress to the real economy is a complex and difficult modelling challenge. Whether we recognize it or not, the standard solution to a complex modeling challenge is to simplify–typically to the minimum extent possible–aspects of the overall modelling framework. However, incorporating feedback into the stress test framework may require simplifying aspects of the framework to a point where it is less able to capture the risks to individual institutions. Even so, one can imagine substantial gains from continued research on stress testing’s role in macroprudential supervision and our understanding of risks to the financial system, such as knock-on effects, contagion, fire sales, and the interaction between capital and liquidity during a crisis.
Finally, let me close by addressing a question that often arises about the use of a supervisory stress test, such as those conducted by the Fed, with common scenarios and models. Such a test may create the possibility of, in former Chairman Bernanke’s words, a “model monoculture,” in which all models are similar and all miss the same key risks. Such a culture could possibly create vulnerabilities in the financial system. At the Fed we try to address this issue, in part, through appropriate disclosure about the supervisory stress test. We have published information about the overall framework employed in various aspects of the supervisory stress test, but not the full details that banks could use to manage to the test. This–making it easier to game the test–is the potential negative consequence of transparency that I alluded to earlier.
We also value different approaches for designing scenarios and conducting stress tests. In the United States, in addition to supervisory stress testing, large financial firms are required to conduct their own stress tests, using their own models and stress scenarios that capture their unique risks. In evaluating each bank’s capital planning process, supervisors focus on how well banks’ internal scenarios and models capture their unique risks and business models. We expect firms to determine the risks inherent to their businesses, their risk-appetite, and to make business decisions on that basis.
The UK Prudential Regulation Authority (PRA) and Financial Conduct Authority (FCA) have today published a review of the changes introduced last year which were put in place to reduce the barriers to entry for new financial institutions. The purpose of the measures was to enable increased competition in the banking industry, to the benefit of customers. The changes focussed on two key areas: reforms to and a simplification of the authorisation process for new banks; and a major shift in the prudential regulation, such as capital requirements, for new entrants.
The two areas of focus were:
In the twelve months following the changes, the PRA authorised five new banks and there has been a substantial increase in the number of firms discussing the possibility of becoming a bank with the regulators. In the twelve months to 31 March 2014 the regulators held pre-application meetings with over 25 potential applicants. These firms have a range of different business models from retail and wholesale banking to FCA-regulated Payment Services firms who are looking to enter the banking market and offer deposits and lending to their current client base (including small SMEs) and others who are proposing to offer a mixture of SME or mortgage lending funded by retail and SME deposits.
The minimum amount of initial capital required by a new entrant bank is £1m compared to £5m under the previous regime. The on-ramp strategies have been helpful for applicant firms that may previously have faced challenges in raising capital or investing in expensive IT systems without the certainty of being authorised.
The PRA intends to publish statistics regarding banking authorisation annually.
In the Australian context, we know from recent client work that potential new entrants face a stiff climb to gain access to the local market. Consideration should be given to given to emulate the UK approach, because we need greater competitive tension in Australian banking.
