For many people, understanding how large banks work is a mystery, with many considering large banks something of a black box. What most people do know is that many banks around the world were bailed out by their governments during the financial crisis and that taxpayers were often left to pick up the bill. Whilst it might seem that banking systems have now stabilised in developed economies, bank bailouts in Spain, Portugal, Belgium, France, Greece, the Netherlands and the US (via the FDIC) have occurred in the last 12 months. These serve as a reminder that not everything is as healthy as recent share market returns may imply. Rather, history implies we can expect bank bailouts to always be part of our financial systems. From a global perspective the threat of another major round of bank bailouts such as occurred in 2007-2009 can seem to be ever present. It can be argued that by the time one series of bank deficiencies has been identified and rectified another set of problems is developing.
The Basel 3 bank reforms were proposed as a response to the problems of 2007-2009, with the changes intended to substantially lessen the likelihood of another round of global bank bailouts. The main aim of this paper is to critique the major reforms, asking whether they will substantially lower the chance that bank bailouts will continue to be required. It argues that regulators should focus on capital reforms and micro credit assessment, with the liquidity reforms offering minimal benefit at substantial cost.
This paper begins with an explanation of the two types of bank bailouts, and a discussion of the factors that lead to a request for government assistance. This is followed by commentary on why governments (and by extension taxpayers) provide such support. A case study with two divergent examples of government intervention is then presented. Finally, the major reforms are analysed with alternative recommendations put forward.
What is a Bank Bailout?
A common misconception about bank bailouts is that they are all the same as they all involve governments giving money to banks. There are actually two types of bailouts, but as they often happen at around the same time it is easy to miss the difference. To explain the reasons why they occur and to help understand what can be done to prevent them, I will use the analogy of a fire in a theatre throughout the paper.
Imagine yourself at a theatre watching a movie. Suddenly from behind you, someone yells out “Fire, fire”. How do you react? If you are like me, the first thing I would do is to look around for the tell-tale signs of a fire; is there heat, smoke or flames apparent anywhere in the theatre? Was the person calling out playing a bad joke on everyone in the theatre? How are others reacting? Are they heading for the exits? Have any staff or emergency personnel arrived and what is their demeanour? Are there any alarms sounding or have any of the overhead sprinklers been activated?
If I see any evidence that there is a fire or if there are warning signs of a fire nearby then I won’t be staying. I can always come back to the movie another day. It is only a small inconvenience to go somewhere else compared to the substantial damage that could occur if I stay with a major fire underway nearby. It’s better to be safe than sorry. If I leave early then there’s little chance I’ll suffer any harm.
The analogy of the fire in a theatre helps illustrate the difference between a liquidity problem and a solvency problem, the two causes of bank bailouts. A liquidity problem occurs when a material percentage of the bank’s depositors all leave at the same time, because they fear that there might be an issue with getting all of their money back in the future. This is similar to a situation where patrons decide to leave the theatre en masse, in fear of a potential danger. Once a few start to leave, it doesn’t take much (there may be no evidence of a fire at all) for the rest of the patrons to follow. If people panic and rush for the exits there could be an ugly situation, but if done in an orderly fashion everything should be fine.
A solvency problem is when the value of the bank’s loan are worth less than or are at significant risk of being worth less than the deposits it is obliged to repay over time. In accounting terms, the bank’s assets are worth less than its liabilities, (or are likely to become so) and thus there is little equity or negative equity in the bank. This is similar to a situation where there is a major fire in the theatre, with the patrons at real risk of death or serious injury if they fail to leave.
Using the analogy, it is easy to see how the problems can be intertwined. If there is a fire then a rush for the exits is likely. If it looks like a bank is at risk of failure, then understandably many depositors will withdraw their funds. When a liquidity problem occurs it is often, but not always an indicator that a solvency problem may also exist.
It is important to differentiate between the two types of bank bailout as the solutions to each are very different. If a bank has a solvency problem the only solution to keep the bank alive is to inject new long term subordinated capital, via preference shares or equity. If done in adequate amounts, this lifts the amount of balance sheet equity back to a normal healthy level. For a liquidity problem the solution is to replace the depositors who have left with an equivalent amount of more patient senior capital. This is typically done by providing a government guarantee to senior capital/depositors or by government provided asset purchase/repurchase facilities.
The expected cost for each type of bailout is also different. Banks that have a liquidity problem, but not a solvency problem have sufficient equity to be able to afford to pay interest or fees for the assistance they receive. On the other hand, banks with solvency problems have no equity left and any assistance provided should be accompanied by the expectation that some of the funds will never be returned and taxpayers will bear the loss. The public outcry when bank management and shareholders kept the gains in the good times but socialised the losses in the financial crisis was entirely justified. Government and regulators therefore have a right to expect banks to take far stronger measures to lessen the possibility that taxpayers will again pay for the recklessness of some banks. The table below summarises the key differences between liquidity and solvency problems.
|Key Difference||Liquidity Problem||Solvency Problem|
|Reason for bailout||Speculation of a solvency problem leads to depositor withdrawals||The bank’s assets are worth less than the amount owed to depositors|
|Type of capital required||Replacement senior capital or provision of deposit guarantees (low risk)||New equity in exchange for partial or full ownership (nationalisation) (high risk)|
|Typical time frame for resolution||<2 years||>2 years|
|Exit mechanism||Capital is repaid or guarantees are no longer required||Sale of bank assets or sale of government ownership position|
|Typical government/ taxpayer outcome||Full return of capital with profit, providing the bank is solvent||Substantial loss of invested capital|
Why do Governments Become Involved in Liquidity and Solvency Problems?
In a theoretical world, banks would be able to sell their assets as depositors left and use the sale proceeds to payout the withdrawals. In reality, highly liquid assets usually form only 10-20% of total assets. The vast majority of assets are illiquid loans for purchasing properties or businesses. These loans are tricky to sell in good times and near impossible to sell in a time of panic. If banks were to attempt to sell illiquid loans when rumours are circulating about their solvency, potential buyers would see that as a “fire sale” and would substantially reduce their bids. If a bank was to sell its illiquid assets at those discounted levels, they are likely to end up with a solvency problem if they didn’t have one already.
Ideally, the way to solve a liquidity problem is to hold the assets and deal with the situation gradually. Existing loans will progressively be repaid and after a period of no negative news depositors are likely to return. Whilst this seems a rational approach, in a panic it is usually the government alone who is willing to take a long term view. If the problem is limited to liquidity, the government can provide replacement senior capital/guarantees for the short term (charging interest and/or fees) and allow the bank’s management time to take measured action.
As mentioned in the previous section it is easy for liquidity and solvency issues to be intertwined. Many bank bailouts begin as a solution to liquidity problems but then quickly move on to solvency issues. The primary hurdle in raising equity for a distressed bank is that there are few potential investors for such a high risk proposition. The government is left with a decision to either intervene or let the bank fail.
At the time of the decision, the government is likely to be pressured by a panicked public and the financial industry to provide a “solution”. The bank management is typically saying that the problems are small and will be quickly resolved. There is usually less than a week for an independent investigation to be conducted to determine the true asset and liability situation. Many commentators wrongly attributed some blame for the financial crisis to the US government’s inaction on Lehman Brothers. The same commentators might now be saying that a similar disaster is looming if nothing is done.
Given all the pressure, it is very unlikely that politicians will allow a distressed bank to fail. The threat of an imminent disruption to the economy trumps the unquantified long term costs. The short term issues like political popularity and being seen to “take action” rather than being cast as being “asleep at the wheel”, will far outweigh long term issues such as increased government debt, higher taxes and austerity measures. The saying that prevention is better than cure also applies to bank solvency. Regulators need to be vigilant throughout business cycles to steer banks well away from solvency issues, ensuring that banks and politicians are not given the opportunity to leave taxpayers with a bill that can take decades to overcome.
Two Examples of Bailouts During the Financial Crisis
In 2008, rumours began to circulate that Irish banks were heavily exposed to a widespread property bubble. Following the collapse of Lehman Brothers in September 2008, many offshore depositors and funders withdrew their deposits and sold their bonds, fearing that Irish Banks would also fail. On the 29th of September 2008, the shares prices of the major Irish banks fell between 15% and 45%. The Irish banks appealed to their government for an emergency injection of liquidity. Failure to provide liquidity was expected to result in the failure of all major Irish banks within a matter of days.
On the 30th of September the Irish Government announced that it would guarantee all senior deposits of Irish Banks for two years. Within three months, banks were receiving billions of euros to meet capital shortfalls in exchange for preference shares. What was initially expected to be a short term liquidity guarantee has resulted in more than €70b of losses so far. This is equivalent to 32% of the current GDP of Ireland or to sixteen months of federal government expenditure. A decade of negligent regulatory oversight and low levels of bank capital combined to see Irish taxpayers picking up an enormous bill on behalf of their banks that will impact the country for decades to come.
Not unlike Ireland, Australia also offered to guarantee the senior deposits of its banks in October 2008. Financial institutions were to pay fees of 0.70%-1.50% per annum for guaranteed deposits over A$1m. The scheme was closed to new deposits in March 2010, with usage peaking at A$170b in February 2010. The Australian government had collected A$3.38b in premiums as of January 2012, with the final total likely to be in excess of A$4b. To date, no financial institutions have failed and no claims have been paid.
The extreme difference in outcomes between the two countries can largely be attributed to the different levels of risk tolerated by the banks and their regulators in the lead up to the financial crisis. Whilst both Irish and Australian banking regulators believed their banks were solvent and well capitalised when liquidity guarantees were granted, it is only after the fact that the truth could be known with certainty. The examples show that a bank with good quality loans and robust capital levels is very unlikely to strike solvency issues. However, the two examples also show that liquidity issues can impact both healthy and unhealthy banks alike.
What are the Main Targets of the Basel 3 Reforms?
As discussed earlier in the paper, one potential response of a bank to depositors leaving is to sell assets and use those proceeds to meet the withdrawals. The issue is that the vast majority of bank assets are illiquid and can’t be sold quickly for a fair price. One of the key Basel 3 reforms is to force banks to hold more liquid assets that are easily saleable in a time of panic. These liquid securities can be cash, central bank deposits or government bonds. These securities are considered to be very low risk and can be easily sold or redeemed at any time. More liquidity is obviously a positive, but it comes at a cost and only provides a thin layer of additional protection.
Banks have protested the need for more liquid assets arguing that it will result in higher interest rates on home loans and business loans. Low risk assets pay a much lower return than illiquid assets, often less than the return paid to depositors. Banks will seek to recoup the lost revenue, offsetting losses on liquid assets with higher pricing on illiquid assets.
Liquid assets also provide only limited protection in the event that many depositors seek to withdraw their money at the same time. Banks are now holding in the order of 10-20% of their total assets in liquid assets, a buffer that can be quickly exhausted if there is widespread fear of a bank becoming insolvent. Their existence simply allows for a slightly larger group of depositors to leave before the government will be pressured to intervene. For solvency issues, liquid assets do very little to prevent or limit damage, with the possibility that the resultant higher interest rates lead to adverse selection issues.
Another key Basel 3 reform is banks are now expected to source more of their deposits for longer periods. This reform strikes at the heart of one of the key imbalances that banks have always faced; depositors want to commit their funds for short periods whilst borrowers want long terms loans. Fixing this mismatch of maturities with longer term deposits theoretically seems easy, but the practicalities are quite difficult.
In a theoretical world, banks would be forced to match the maturities of their obligations to the maturities of the loans they make. This would create a situation where banks that had depositors leaving would be able to cease new lending and thus allow the amount of loans they had to fall in line with the reduction in their deposits. This theoretical point has practical out-workings in investments such as mortgage backed securities and covered bonds, where banks are able to get very long term funding to match the very long term commitments of residential property loans. These securities are helpful developments and should generally be encouraged by government and regulators.
The two primary downsides of shifting to a matched funding profile are that the cost is material and that sticking to the matched profile could accentuate cyclicality in credit markets. On the cost side, retail and institutional providers of deposits to banks both prefer shorter commitment periods and demand substantial premiums for longer term funding. Like the costs for holding more liquid assets, the increased costs for matched funding would be passed on to borrowers.
Alternatively, banks could reduce the term of loans to their customers, but in the event of an economic downturn many bank customers would find their loans maturing at the same time and may have no other option but to sell their assets to repay debts. This would likely result in a severe downturn in asset prices as there would be many sellers and few buyers. The situation would be accentuated further as potential buyers would struggle to be able to get financing and would therefore have to pay cash. This cash would likely come from deposits they withdrew from a bank, which might again lead to more borrowers selling assets. This scenario has the potential to make boom and bust cycles much more severe. It also has similarities to the depression era of the 1920’s and 1930’s when government, businesses and consumers all reduced spending at the same time.
The third key area of Basel 3 structural reform is requiring banks to increase their capital levels, and applying more stringent rules to measure how much capital banks have. Returning to the analogy, increasing bank capital is like requiring theatres to adopt a raft of fire safety measures such as sprinklers, hose reels and fire extinguishers that are easily accessible and are regularly tested. If capital levels are set at conservative (high) levels this is like a situation where if a fire starts the theatre can be flooded with water, to the extent that only a catastrophic fire is able to overwhelm the safety measures. In almost all circumstances patrons will be safe, that is taxpayers will not be left with a loss.
The Basel 3 reforms increase the minimum amount of subordinated capital that a bank must hold. The old rule of 8% of risk weighted assets is now increased to a minimum of 10.5%. What the reforms have only partly addressed is the gaming of the risk weighted assets system. This occurs when banks select assets that require little or no capital to be held against them but which pay high returns and are therefore deemed by the market to be higher risk. The clearest example of such gaming is sovereign debt, which many regulators require little or no subordinated capital (implying low risk) to support. Some of these governments should be considered as being medium to high risk, as reflected in current market prices. Many European banks were gaming the system with Greek government debt, and consequently suffered material losses when the Greek debt was eventually restructured.
The partial fix to gaming of the risk weighted assets system is the introduction of a leverage ratio. This requires that regardless of the risk weighted ratios, tier one capital (essentially common equity + preference shares) must be a minimum of 3% of total exposures. This still allows banks to have total exposures up to 33 times their tier one capital, if they are wily in their asset selection. To meaningfully reduce the risk of solvency problems a minimum of 8% for the leverage ratio should apply when measured on a total capital (including tier two/subordinated debt) basis.
The short comings of the risk weighted assets system was perhaps best exposed by the Dexia bailouts in 2011 and 2012. Following the release of the European Banking Authority stress tests in July 2011 Dexia boasted of its “strong capital base” and that it “will continue to ensure that appropriate capital levels are maintained.” In August 2011 Dexia announced a loss of €4.03b for the half year and reiterated that it had “strong tier one (11.4%) and core tier one (10.4%) ratios.” However, when calculated as percentage of total assets the tier one leverage ratio was an anaemic 2.79%.
In October 2011, Dexia announced that it would undergo a substantial restructure of its operations with Belgium and France providing €90b of liquidity guarantees. Public announcements in November 2011 emphasised the bailout was addressing liquidity issues, pointing to the tier one ratio of 9.9% to argue that Dexia was well capitalised. The share price had collapsed, other banks had withdrawn their funding, the leverage ratio was at extreme levels, but through gaming of the risk weighted asset rules Dexia still showed what seemed to be fairly healthy capital ratios. In November 2012, the smokescreen of a “strong capital base” was blown away with the announcement that Belgium and France would together provide €5.5b in exchange for preference shares, with existing equity to be wiped out.
What Needs to be Done to Reduce the Risk of Banks Requiring Future Assistance
Focus on Capital
Unlike liquidity problems, solvency problems are largely preventable. They are far more likely to occur and will be far more costly to fix if a bank starts with low levels of capital. Governments and regulators need to remember that adding more capital, lengthening the funding profile and holding more liquid assets will all impact the cost of credit. Only the capital reforms will have any meaningful impact on the prospect of a government and its taxpayers incurring a loss if there is a bailout. Rather than expending energy fighting with banks over improved liquidity profiles, regulators would be better served by pushing for total capital to risk weighted assets of at least 16%.
The intertwined natures of liquidity and solvency problems mean that higher capital levels are even more critical in our fast moving generation. Traditional media will harangue a bank with queues outside its branches and social media allows rumours and poorly researched opinions to flourish. Credit default swaps allow investors to quickly short a bank and push up its cost of funding. The increased availability of economic and financial data can make it easier to identify banks exposed to emerging risks. All of these developments point to liquidity problems becoming more commonplace. A bank with (1) a healthy amount of equity and other subordinated capital and (2) a modest leverage ratio can mount a strong argument of its safety in the face of a panic. It can also provide a much higher level of reassurance of repayment if it eventually does require liquidity support from its government.
Encourage Greater use of Additional Tier One and Tier Two Securities
In order to build capital ratios without substantially increasing the cost of lending, regulators should encourage banks to issue greater amounts of preference shares and subordinated debt. The Basel 3 reforms have hit the nail on the head in terms of the appropriate rules for subordinated capital. The changes ensure that new subordinated capital is (1) not able to be redeemed without regulatory approval and (2) is undisputedly available to absorb losses before senior creditors or taxpayers are impacted. Many banks already have common equity ratios above 8% of risk weighted assets and may not need additional common equity. By adding further layers of 4% preference shares and 4% subordinated debt on a risk weighted assets basis, banks would be materially stronger at minimal additional cost.
Measure Capital by Risk Weighted Assets and Leverage
The introduction of a leverage ratio as part of the Basel 3 reforms is a positive development. However, the minimum tier one leverage ratio of 3% of total exposures is far too low and should at least be doubled. With the inclusion of tier two capital an overall minimum of 8% total capital to total exposures is recommended. By strengthening the risk weighted assets and leverage ratios solvency problems will be greatly reduced. With most banks benefitting from explicit and/or assumed sovereign support, governments and taxpayers are more than entitled to expect banks to raise their capital ratios.
Up-skill Regulators in Micro Credit Assessment
Bank regulators are too often strong on macro policing to the detriment of micro policing. Historically, the regulatory emphasis has been on top down ratios and macro stress tests. Regulators have largely avoided detailed analysis of lending books that includes judgements on whether a bank has conservatively analysed loan risks and made appropriate provisions. To make this change, regulators will need to employ more staff with a background in the day to day process of lending, and fewer staff from government, academic and consulting backgrounds.
One great example of how regulators can implement this concept is the US Shared National Credit (SNC) annual review process. Loans over $20m that are shared amongst three or more banks are reviewed annually, with the typical two page SNC report detailing the regulator’s risk grade, the amount of capital to be set aside and the rationale for the regulator’s decisions. All banks with exposure to the reviewed borrower are required to adopt a capital position at least as conservative as the SNC report. Whilst the SNC program is limited to institutional multi-bank exposures, these are usually the riskiest loans that banks make. The insights gained from the process are invaluable for a regulator who is attempting to assess whether banks are being conservative in their lending practices. These insights will guide the views on the macro health of banks, including setting of countercyclical buffers.
Securitisation Should be Encouraged
Of all the liquidity measures available securitisation is arguably the most effective, with added benefit of reducing solvency issues. By taking the longest dated loans (20-30 year residential loans) off balance sheet the liquidity mismatch is substantially lessened, improving the liquidity position. Any losses on those loans are ring fenced from the bank balance sheet, improving the solvency position. The pressure on governments to bailout unsophisticated depositors in banks doesn’t apply to the sophisticated buyers of mortgage backed securities. If there is a need for either a liquidity or solvency bailout, a smaller balance sheet means less assistance will be required.
The sheer size and number of liquidity and solvency bailouts during the financial crisis stunned governments and regulators. Four years on, bailouts are still commonplace. The outrage that taxpayers were left to foot the bill for excessive risk taking in banks was channelled into creating the Basel 3 reforms. These reforms aimed to increase liquidity and solvency, so that the probability and severity of future bailouts would be substantially lower. The noble aims of the reforms are now hitting substantial push back from banks, who argue that the reforms will increase the cost of lending and do little to lessen the risk that future bailouts will be required.
This paper has shown that healthy and unhealthy banks alike can strike liquidity issues, but that solvency issues occur when banks make poor quality loans and have low capital ratios. If a pragmatic trade-off must be made, taxpayers would be better protected by increasing the minimum capital and leverage ratios rather than having more liquid assets and term funding. This can be cheaply done through greater use of preference shares and subordinated debt. Finally, regulators should up-skill in individual loan assessment. Regulators must go beyond macro stress tests in assessing bank health, checking that a bank’s loan book is conservative and that countercyclical risks are fully understood.
Comments and criticisms are welcomed and can be sent to [email protected].
This article has been prepared for educational purposes and is in no way meant to be a substitute for professional and tailored financial advice. It contains information derived and sourced from a broad list of third parties, and has been prepared on the basis that this third party information is accurate. This article expresses the views of the author at a point in time, and such views may change in the future with no obligation on Narrow Road Capital or the author to publicly update these views. Narrow Road Capital advises on and invests in a wide range of securities, including securities linked to the performance of various banks.