The EU’s implementation of Basel III: A deeply flawed compromise
Europe’s banks are in bad shape. Slowing growth and
rising capital adequacy ratios would stretch any bank. Doubts about
sovereign debt and the Eurozone’s future may push some EU banks over the
edge. Now the EU has decided how to implement the principles of the
latest round of globally coordinated banking regulations – Basel III.
This column argues that the EU has got it wrong.
By all accounts, EU
member countries have for months been debating how to implement the
minimum bank capital standards agreed under Basel III. Their arguments
have unfolded as the EU works to complete its fourth Capital
Requirements Directive and its Capital Requirements Regulation (see
Veron 2012).
Three issues have been contentious:
- Whether member countries should be permitted to enact minimum
capital ratios considerably tougher (higher) than those specified under
Basel III without approval of the EU;
- Whether the restrictions on what can be counted as high-quality
capital under Basel III should be scrupulously adhered to in EU
legislation; and
- Whether the Basel III deadlines for introducing an unweighted
leverage requirement for bank capital and two new quantitative liquidity
standards (the liquidity coverage ratio and the net stable funding
ratio) should be mirrored in EU legislation.
Unfortunately, the decision of the finance ministers announced on 15
May 2012 reflected a compromise that set back the cause of reform,
risking further instability for the banking system in Europe and the
global economy generally. The European Parliament should demand
significant changes before approving this very flawed measure.
It has been reported that in the debates, the UK, Sweden, and Spain,
with the support of the ECB, favoured a “yes” on all three questions.
For convenience, I call this position the Osborne View – named after
perhaps its most ardent proponent, George Osborne (UK Chancellor of the
Exchequer). Another group of EU countries, reportedly led by Germany and
France, with the support of the European Commission, opposed that
position. I call this position the Schäuble View in honour of Germany’s
Minister of Finance, Wolfgang Schäuble.
The 15 May compromise at a meeting of EU finance ministers was
endorsed unanimously by the Council. Not all the details have been
published, but the main features can be summarised as follows.
- Measured against banks’ total exposure, EU members will need EU
approval to implement in their national banking legislation minimum
(risk-weighted) bank capital ratios that exceed the Basel III minimum by
more than 300 basis points.
- For domestic and non-EU exposure, the threshold for EU approval will be higher, at 500 basis points.
These thresholds would allow the UK to implement the recommendations
of the Independent Commission on Banking (the Vickers Report) –
including the ringfencing retail banking operations and a minimum equity
capital ratio of 10%, without EU approval.
- The 15 May accord also permits EU banks to count as equity capital
several financial instruments with dubious loss-absorbency, including
the so-called “silent participations” of German banks and the minority
stakes of French banks in insurance companies.
Such a step weakens the Basel III guidelines on the quality of bank
capital. In one of the few concessions to the Osborne View, the
agreement adheres to the Basel III time schedules for the leverage ratio
and the two liquidity standards. Finally, to provide additional
macro-prudential tools against asset-price bubbles in real estate,
member countries will be allowed to modify (increase) the bank capital
risk-weights against exposures to residential and commercial property.
Over the next few weeks, the European Parliament will discuss these
steps and negotiate a final version with EU finance ministers. There are
reports that the European Parliament may demand restrictions on bonus
payments (in bank compensation policies) before approving the package,
however.
Evaluation of the 15 May decisions
From a narrow procedural perspective, the 15 May ECOFIN decision is a
step forward on financial sector reform. After all, the Basel III bank
capital standards cannot go into effect until they are embedded in
national banking legislation and the 15 May ECOFIN decision advances
their implementation.
But fundamentally the 15 May decision is a setback for reform.
Basel III’s main purpose was to enhance financial stability and to
limit the liability of taxpayers for bank losses by putting much more
high-quality bank capital into banking systems around the world. The 15
May decision makes that harder in two ways.
- Requiring EU countries to win approval of the EU for minimum capital
ratios much in excess of the modest standards in Basel III will
discourage urgently needed strong capitalisation (and recapitalisation)
of banks in the Eurozone.
- Watering down the Basel III standards for the quality of bank
capital may encourage a race to the bottom on the loss-absorbency of
bank capital. Such a development would weaken limits on taxpayer
liability for bank losses.
I base my conclusion on five main points that I address in turn.
EU banks are undercapitalised
First, EU banks remain undercapitalised when evaluated by the
appropriate metrics. The most reliable metric of capital adequacy –
warts and all – is the simple unweighted leverage ratio (the ratio of
equity to total assets). All the risk-weighted measures of bank capital
have been distorted by political pressures, conflicts of interests, and
gaming of the regulations by banks (Helwig, 2010, Admati et al, 2011).
Exhibit A: with all the questions about debt sustainability in some
EU periphery countries, the sovereign debt of EU countries still
receives a zero risk weight for the purpose of calculating risk-weighted
assets, the denominator in all risk-weighted capital ratios. Experience
in the run-up to the global economic and financial crisis of 2007-2009
also demonstrated the poor quality of risk-weight calculations for
banks’ trading assets, securitised instruments, credit ratings of
complex financial instruments, and the estimation of risk-weighted
assets in banks’ internal models (Goldstein, 2012).
For this reason alone, the stress tests conducted by the European
Banking Authority (EBA), which focus on risk-weighted capital ratios,
ought to be viewed skeptically. Indeed, concern over deficiencies of
risk-weighted capital metrics led the authors of Basel III to include a
minimum (unweighted) leverage ratio over the objection of the banking
industry.
The April 2012 Global Financial Stability Review reports that the
leverage ratio (the ratio of Tier 1 common capital to adjusted tangible
assets) for EZ banks in 2011 was a little more than 4% – versus about 5%
for UK banks and roughly 6.5% for US banks (IMF, 2012a).[1]
Other measures of bank fragility point to the weakness of EU banks.
- The IMF (2011) estimated that between end-2009 and August 2011, EZ
banks had a €200 billion increase in credit risk associated with
holdings of the stressed sovereign debt of Greece, Ireland, Portugal,
Belgium, Italy, and Spain.
- Inter-bank exposures to these countries brought the increase in credit risk to €300 billion.
No wonder that IMF Chief Christine Lagarde, speaking in August 2011,
emphasised the urgent need for recapitalisation of EU banks (Lagarde
2011).[2]
EZ banks also rely more heavily on wholesale financing than US or UK
banks, with loan-to-deposit ratios hovering at 125 – versus 105 for UK
banks and less than 80 for US banks. The vulnerability of relying on
wholesale funding was of course dramatically underlined in the second
half of 2011, when wholesale funding strains for European banks
compelled the ECB to launch its three-year Long-Term Refinancing
Operation (LTRO) – a two-stage rescue effort that has risen to more than
a trillion euros.
While necessary to prevent large-scale liquidity problems from
generating massive deleveraging and exacerbating solvency concerns, the
LTRO has produced undesirable side effects.
- It has facilitated a carry trade on peripheral sovereign debt that
has led banks in those countries to load up even more sovereign debt,
aggravating the adverse feedback loop between bank debt and sovereign
debt.
- In addition, by tying up large amounts of collateral, the LTRO has
further imperiled the position of unsecured bank creditors in the medium
term. Spanish banks alone have enormous exposures to real estate.
- Despite a Spanish housing bubble bigger than the recent one in the
US, property prices in Spain have fallen from their peak by less than in
the US.
On top of all this, the IMF (2012a, b) has estimated that Eurozone
banks are likely to face pressures to deleverage more than $2 trillion
of bank assets by the end of 2013. The fund has also warned that such
deleveraging would slow EU economic activity and reduce the health of
its banks.
In the face of such obvious banking fragility in the EU, the 15 May
ECOFIN decision on bank capital seems divorced from realities.
With potentially large future EU bank losses on the horizon,
constraining the ability of regulators in some EU countries to set
minimum bank capital standards in excess of the Basel III minimums is
imprudent. Doing so would prevent banks in some EU countries from having
enough “self-insurance” to handle potential losses. Consequently, the
burden of financing those losses would again fall unfairly on taxpayers
and lead to greater dependence on the ECB for liquidity.
Cross-country differences in the size of too-big-to-fail banks call for cross-country differences in minimum capital standards
The 15 May decision also does not adequately address differences
among EU countries in the extent of the too-big-to-fail problem and the
implications for minimum bank capital ratios. The combined assets of the
five largest banks relative to GDP are three times as high in the UK
and the Netherlands (at about 450% of GDP) as in Spain and Italy
(Goldstein and Veron, 2011).
The greater the size of too-big-to-fail banks in an individual
country, the more pressing the need to provide adequate self-insurance
against losses in those institutions. Such self-insurance must come from
higher bank capital. It is thus no accident that two countries where
the combined assets of the few largest banks are particularly large
relative to home-country GDP – Switzerland and the UK – have already
moved to enact national minimum bank capital standards tougher than
those in Basel III (Goldstein, 2011).
The May 15 agreement constrains the ability of EU countries to set
such rational self-insurance requirements for their banks. Those EU
countries with large banks (relative to home-country GDP) will
henceforth face two unpalatable choices. Either they will have to
break-up their largest banks to minimise the government’s prospective
liability – something that they have so far (unfortunately) refused to
consider seriously – or they will have to accept (implicitly) the
reality that if these very large banks do come under acute distress, the
cost of saving or liquidating them will fall predominantly on
taxpayers.
European leaders have pledged repeatedly to spare taxpayers from that
burden. More generally, it makes no sense to constrain differences in
minimum bank capital ratios across EU countries when there is no pan-EU
deposit insurance and bank resolution regime in place and when there are
sizeable differences in banking risk across these economies. Such
differences reflect troubled legacy assets, sovereign debt burdens,
cyclical positions, and longer-term structural factors (like the size of
too-big-to-fail banks).
The minimum capital ratios in Basel III are way too low
Third, the minimum capital standards agreed under Basel III itself,
while better than those in Basel II, are still way too low. Thus,
individual countries – whether or not they are members of the EU – need
the latitude to exceed those Basel III minimums in their own national
legislation.
Such freedom supports the spirit of successive Basel bank capital agreements, which have always aimed to set minimum
standards to prevent a race to the bottom and to avoid maximum
standards that would discourage a race to the top. Costly banking crises
occur when banks have too little capital – not when they have ample
capital.
The literature suggests several approaches for estimating the minimum or optimal ratio of bank capital.
- One derives from the fact that banks typically hold bank capital in
excess of the Basel minimums even at the bottom of the business cycle.
Banks arguably behave that way because the markets – nervous about
the near-death experience of some banks in the previous crisis –
pressure them to do so. The following question is then posed: if we want
banks to remain solvent after suffering credit and trading losses in
severe crises and still have enough capital to meet the market imposed
minimum at the bottom of the cycle, how high would the minimum capital
ratio have to be at the top of the cycle?
A recent study by Hanson, Stein, and Kashyap (2010) noted that U.S.
banks lost roughly 7% of assets during the 2007-2010 period and that
very large U.S. banks had a common tier one capital ratio (relative to
risk-weighted assets) of about 8% in the first quarter of 2010 (near the
bottom of the cycle). From that, they concluded that the minimum
capital ratio needed to be about 15% at the top of the cycle – far above
Basel III’s minimum 7% ratio (for common core tier 1 capital) or its
9.5% ratio for the globally most systemically-important banks. Applying
this approach to other countries’ bank-loss experience and to banks’
capital holdings at different times generates estimates of minimum
(common equity) capital ratios in the 12-25% range. Again, these levels
are far higher than the Basel III minimums.
- A second approach is to employ a cost-benefit framework to the estimation of optimal bank capital ratios.
On the benefit side, higher capital reduces the probability of a
systemic bank crisis that would depress economic growth and undermine
government fiscal positions. On the cost side, higher capital is assumed
to increase bank funding costs, lower loan volumes, and impede economic
growth. A recent study done at the Bank of England (Miles, 2011)
applied such a cost-benefit approach and concluded that the optimal bank
capital ratio (as a percentage of risk-weighted assets) was about 20% –
more than double the Basel III minimum for even the most
systemically-important banks.
Also employing a cost-benefit approach, a group of twenty
distinguished professors of finance (Admati et al, 2010a) concluded that
the minimum unweighted leverage ratio for banks ought to be at least
15% – five times the minimum leverage ratio (3 percent) included in
Basel III. The main reason why the cost-benefit approach leads to
(optimal) minimum capital ratios higher than those in Basel III is that
while the social benefits of higher bank capital are substantial, the
social costs of higher bank capital turn out to be modest.
The yawning gap between the minimum standards in Basel III and the
minimum standards derived from the best empirical evidence makes it
counterproductive to prevent countries from imposing standards higher
than those in Basel III.
The quality, not just quantity, of bank capital matters
Basel III was not just about increasing the quantity of bank capital.
It was also about improving the quality of that capital so that it
would be truly loss-absorbing.
In this regard, the Basel Committee on Banking Supervision (BCBS) was
responding to one of the main lessons of the global economic and
financial crisis: when banks are permitted to count financial
instruments as regulatory capital that either are not fully
loss-absorbing or are loss-absorbing only if and when the bank is being
liquidated, de facto bank capital can be much less than regulatory
capital and the government winds-up having to inject funds as common
equity into the rescue.
To avoid repeating that mistake and to limit the future public-sector
liability, Basel III emphasised the highest quality bank capital –
namely, equity capital. It also restricted (to 15% of the common equity
component) the combined weight of lower-quality financial instruments
that had been previously allowed. These components included deferred tax
assets, mortgage servicing rights, and significant investments in
common shares of unconsolidated financial institutions (including
insurance companies).
The parts of Basel III dealing with the quality of bank capital
represent a compromise in which the US, Japan, and the EU each agreed to
restrict a low-quality component of bank capital that was viewed as
attractive by their own banks. For example, Japanese banks were
reluctant to disqualify tax deferred assets, U.S. banks wanted to hold
on to their mortgage servicing rights, and EU banks – particularly those
with large insurance subsidiaries – wanted to retain their minority
stakes in unconsolidated subsidiaries.
The 15 May EU agreement threatens to unravel the Basel III compromise
on the quality of bank capital – largely to appease some German and
some French banks, which do not want to go to the market to raise
high-quality bank capital in order meet higher capital requirements
under both Basel III and the EBA stress tests. Instead, these banks
pressured their governments to convince other EU members that various
low-quality capital components should satisfy such capital requirements.
In that way, these EU banks can continue to pay dividends and dole out
excessive compensation packages to employees.
If the 15 May concessions had not been forthcoming, these banks would likely be judged as under-capitalised.[3]
The Schäuble View on the quality of bank capital should be seen for
what it is: an effort to use smoke and mirrors to weaken the Basel III
restrictions on the definition of high-quality bank capital. Some EU
ministries of finance may also see short-term advantage in carrying the
banks water on this issue. Recapitalizing the banks via government
injections of capital, after all, would increase already high sovereign
debt levels and would anger voters fed up with government bailouts of
the financial sector.
The EU retreat on the quality of bank capital is likely to have two costly consequences.
- First, French and German banks will have less real, high-quality
bank capital than advertised, putting French and German taxpayers on the
hook for bank losses in excess of available, truly loss-absorbing
capital.
Markets will see through this tactic, so there will be little if any advantage in funding costs.
- Second, giving French and German banks a pass from an important part
of Basel III will almost surely encourage U.S. and Japanese banks,
among others, to press for similar concessions from their national
regulators, invoking other low-quality components of bank capital
relevant to their balance sheets.
If successful, such efforts will weaken the effectiveness of the
overall Basel III agreement and thwart one of its main objectives –
forcing banks to become responsible for financing their own losses,
without inflicting costs on innocent bystanders.
Unpersuasive arguments about the effect of higher capital requirements on economic growth and the single market
In campaigning against higher bank capital standards, the banking
industry argues that a higher capital ratio will be a calamity in terms
of increased bank funding costs, lower loan volumes, and slower economic
growth. Unfortunately, most EU finance ministers seem to have bought
into that fallacy. In addition, the Schäuble camp trotted out the
equally weak argument that permitting EU countries to implement much
higher minimum capital ratios than the Basel III minimums would
jeopardise Europe’s Single Market.
Perhaps the best rebuttal of this contention has been made by Anat
Admati and her colleagues at Stanford Business School and at the Max
Planck Institute (Admati et al, 2010a, 2010b, 2011). A summary of their
reasoning, which I strongly support, goes as follows.
- The cost of capital depends on the risk to which the capital is put.
- Relying more heavily on equity unambiguously reduces the volatility
of the return on equity and lowers the risk premium on equity.
- The required return on equity must therefore decline with an increase in equity.
- The end result is that the bank’s overall cost of capital is likely to be little affected by increasing its capital ratio.
Nor does a higher capital requirement necessarily imply a reduction
in loan growth unless banks are allowed to meet that requirement by
shedding assets.
Unlike liquidity requirements over bank assets, capital requirements
deal with the liability side of bank balance sheets and with funding
choices. Banks do not hold the securities they issue; investors do.
Capital is therefore not put in a strongbox. Many successful
non-financial corporations fund themselves more heavily from equity than
banks; yet these companies are not constrained by that funding choice
in expanding their investments. Historically, the most severe credit
crunches – like the 2007-2009 global crisis – occur when bank capital is
very low, not when it is high. When leverage is low, such as the
bursting of the internet bubble, the consequences for the macro-economy
are much more benign that when leverage is high.
The empirical evidence on the effect of bank capital on bank funding
costs and economic growth also supports the Admati View. For the US and
UK economies it is possible to get time-series data on bank capital than
go back to the 1840s. Over this long history, leverage ratios for bank
capital varied widely. At times the ratio was more than four times
higher than they are today. Yet statistical tests show no link between
higher bank capital ratios and higher bank funding costs or weaker
macroeconomic performance (Hanson, Kashyap, and Stein, 2010, and Miles,
2011). The cross-section evidence goes in the same direction. Small
banks in the US routinely have much higher capital ratios than large
banks, yet this difference has hardly led to their demise. The average
capital ratio for non-financial companies in the US is 70 percent. Yet
there no reason to believe that such a funding mix has hurt their
performance.
The Schäuble View on the impact of bank capital requirements on bank
lending also appears to run counter to that of the sitting Chairman of
the European Banking Authority (EBA), Andrea Enria. In a recent (April
2012) speech, Enria (2012b, p. 10) said:
“On this, I want to be blunt. I do not believe that high levels of
capital are a deterrent to new lending. On the contrary, banks with low
capital levels—or perceived by the market as being so – are those that
have had problems in increased lending. They either face major funding
difficulties – which in turn do not allow them to grant loans – or focus
primarily on preserving their meager capital. Banks with large capital
positions, by contrast, are less sensitive to cyclical shocks and more
likely to pursue lending growth strategies.”
The way in which higher capital requirements are implemented also
matters. As argued in Goldstein (2012), when revealing the results of
its stress tests and designing bank recapitalisation guidelines, the EBA
should have included a firm policy on bank dividends and executive
compensation. Any bank that did not meet the minimum capital ratio
should have been directed to suspend dividend payments until it reached
the target and was no longer in danger of falling below the target over
the next year. The EBA should also have included an adverse
macroeconomic scenario for the Eurozone to strengthen confidence in the
stress test results. Last but not least, the target capital ratio should
have been translated into a target for increases in bank capital alone.
That step would have avoided giving banks an incentive to meet the
target by decreasing the denominator (asset shedding) rather increasing
the numerator (raising new capital).
The way forward
To sum up, denying EU countries the scope to raise minimum capital
levels above the Basel III minimums, ostensibly to sustain lending and
economic growth in the region, flies in the face of both economic theory
and the evidence. Here too, the Schäuble View is merely another sop to
the banks.
The argument that large differences in minimum bank capital ratios
among EU countries would risk splintering the Single Market likewise has
little foundation. As maintained by Veron (2012), the largest
distortion to the EU market for financial services is that the framework
for supervision and resolution of financial institutions remains
predominantly national. For this reason, the financial health of banks
is tied to the health of their home-country sovereigns. If EU finance
ministers want to strengthen the single market for financial services,
they should implement an EU-wide framework for deposit insurance and
bank resolution, going if necessary beyond the European Financial
Stability Facility (EFSF) and European Stability Mechanism (ESM)
(Roubini, 2012), along the lines of the Federal Deposit Insurance
Corporation (FDIC) in the US. Tinkering with cross-country differences
in minimum bank capital ratios will do little to advance the Single
Market. And if the concern is that individual EU countries with high
minimal bank capital ratios would use those minimums to shed assets
outside their home market, this should be addressed by negotiating an
EU-wide coordination agreement on cross-border bank lending, like the
Vienna Initiative of 2009. A “convoy approach” to bank capital – in
which EU countries seeking safer banking systems are barred from doing
so out of fear of pressure on other EU countries with weaker regimes –
will not promote the Single Market. As in the case of Japan’s banking
crisis (Posen, 2000), such a convoy approach will delay addressing the
problem of under-capitalised banks and weaken economic recovery in the
euro zone.
The decision by European finance ministers on EU implementation of
Basel III on 15 May is a setback for financial regulatory reform not
only in Europe but globally. The priority ought to have been focused on
increasing the quantity and quality of bank capital to deal with the
under-capitalisation of EU banks and reducing the potential for an
adverse feedback loop between bank debt and sovereign debt. Instead, EU
finance ministers went in the opposite direction by constraining the
ability of EU countries to do more than Basel III on the quantity of
bank capital and by weakening Basel III on the quality of that capital.
Their action will further undermine confidence in the solvency of EU
banks and make resolution of the EU debt crisis more difficult. In
addition, it will impede efforts to implement capital reforms of Basel
III in the US and Japan, as banks in those countries try to engage in a
regulatory race to the bottom.
Fortunately, EU Finance Ministers still must negotiate and agree on a
final text of the EU bank capital rules with the European Parliament.
The Parliament should take seriously its responsibilities in this area.
It should demand significant changes to the 15 May ECOFIN draft
agreement. Specifically, the Parliament should press ECOFIN: (i) to
raise the (no-EU approval-required) threshold on national bank capital
minimum above the Basel III minimum to at least 700 basis points (for
total exposure); (ii) to reinstate the Basel III restrictions on what
can be counted as equity capital; and (iii) to introduce the amendment
that any EU bank failing an EBA stress test with regard either to the
common Tier 1 capital ratio or to the leverage ratio should be
prohibited from either paying dividends or employee bonuses until it
reaches the bank capital target and is (in the EBA’s judgment ) not in
danger of falling below the capital target over the next year. If these
negotiations between the Parliament and EU Finance Ministers take a
month or two, so be it: better to take the time to hammer out an
agreement that promotes reform on bank capital than to rush into a
compromise that undermines it.