Implementing the regulatory reform agenda - the pitfall of myopia
Speech by Mr Stefan Ingves, Chairman, Basel Committee on Banking Supervision and Governor, Sveriges Riksbank at the Federal Reserve Bank of Chicago, 6 November 2014.
As prepared for delivery
The subject of this conference is indeed very timely - six years after the outbreak of the financial crisis, there has been substantial progress in the post-crisis regulatory reform agenda, with a number of important milestones reached. Therefore, now is a good time to take a step back and ask how the different bits and pieces of the regulatory framework fit together.
And, more specifically - have the vulnerabilities revealed in the crisis been adequately addressed? Are additional adjustments still necessary?
Or, conversely, have we gone too far and created a regulatory Frankenstein's monster that no-one has full control over and that stifles lending and economic growth?
This latter view is one that I sometimes hear when meeting representatives of the banking industry. The feeling seems to be that we are overwhelming the financial system with a regulation tsunami with too many reforms being implemented too soon. This will lead to unacceptable consequences in the form of higher funding costs, reductions in market liquidity with market-makers pulling out of markets, collateral shortages; and many banking activities simply disappearing, or moving to the so called shadow banking sector.
And indeed - the financial crisis has led to a comprehensive response from regulators and policymakers across the world. Compared to the pre-crisis era, international banks will face:
- substantially higher capital requirements,
- higher demands on the quality of capital,
- a leverage ratio,
- an international liquidity framework, with both short-term and structural liquidity requirements (I am proud to note that the Basel Committee, less than a week ago, published the final standard for the net stable funding ratio, NSFR), and
- a regulatory framework for global systemically important banks (G-SIBs).
When you add to this ongoing work related to reducing RWA variability and disclosure, you end up with a pretty impressive list - a list that represents an unprecedented leap forward in terms of global banking regulation.
So then, how do I see this?
Do I claim to know how all these new rules will play out together? Am I confident that there will be no inconsistencies and contradictions? No, definitely not. We have every reason to be humble in this respect. Monitoring and assessing the effects of reforms will therefore be imperative.
Will the reforms be costly for banks in the short term? Yes, they will.
Will banks have to adjust their activities? Yes, a return to pre-crisis banking behaviour is neither appropriate nor viable.
Do I therefore think that regulation has gone too far and that parts should be undone? No, not at all.
In this presentation I will try to explain why I think this is so. I will also speak about what is still lacking and the regulatory challenges we face ahead.
Why we shouldn't back-track on regulation
There are several reasons why I don't think the regulatory agenda has gone too far. First of all, my experience is that important regulatory and structural reforms are all too often hindered by myopia. People tend to focus on costs and pains in the short run, leaving aside the longer term gains that reforms aim to achieve.
The perceived short-term costs are simply much easier to sell politically, compared to the abstract benefits of lowering the risk of crises. This is especially so, since the benefits may accrue only to future generations - a group that has difficulties making its voice heard in today's policy debate.
This time has been no exception: for years, people shied away from necessary actions to strengthen the financial system. When the crisis hit, perceptions changed, providing a window of opportunity for regulatory reforms that were long overdue. However, we must not begin to close this window and lose sight of why we are undertaking these reforms.
Let me start with a reminder of the regulatory framework before the crisis. Both Basel I and II included a risk-weighted capital adequacy framework. However, for the last 20 years banks' balance sheets ballooned, while their equity failed to take off. For example, from 1993 to 2008 the total assets of a sample of what we call global systemically important banks saw a twelve-fold increase (increasing from $2.6 trillion to just over $30 trillion). But the capital funding these assets only increased seven-fold, (from $125 billion to $890 billion). Put differently, the average risk weight declined from 70% to below 40%. The problem was that this reduction did not represent a genuine reduction in risk in the banking system.
To take an even more concrete example from my own country: during the past twenty years or so, the risk weights for retail mortgages in the major Swedish banks have decreased from 50% to 35% with the adoption of Basel II (from Basel I) and further, to about 6% when banks themselves were allowed to model risk weights. In equity terms, this means that instead of SEK 17,000 of their own equity to fund a mortgage of 1 million, banks' models implied that SEK 1,200 was enough.1 In retrospect, it is clear that the decrease in risk weights did not reflect actual risks and banks therefore needed more capital.
Furthermore, although it is a historical fact that banks' problems often start in the form of liquidity constraints, there were no global liquidity regulations for banks prior to the crisis. This meant that banks could rely heavily on very short-term market funding to finance highly illiquid and long-term assets. This worked fine during the Great Moderation, but unfortunately with the collapse of Lehman Brothers another old truism suddenly came to life: "markets function the worst when you need them the most".
Against this background, it is quite embarrassing that so few could see the crisis coming. From a regulatory point of view, all the ingredients were there, or rather they were lacking. And this is the first point I want to make - the regulatory framework was unsatisfactory and becoming more so the more complex the financial system became.
Then, turning to my second point, which is: The costs of financial crises are huge. This is true in general, but especially so for the recent one. For example, according to a recent study by IMF economists, in a sample of countries representing just over 50% of world GDP, the total amount of government recapitalisation, asset purchases and guarantees during the period 2007-2011 amounted to nearly $5 trillion. This is equivalent to 16% of the GDP of these economies, or nearly $5,000 per citizen.2
But, this is only a lower bound of the cost of the crisis. If we also include the impact on GDP and the loss of production relative to its pre-crisis trend, the costs rise. This has been showed by several studies, including the one just mentioned by IMF economists, which estimates that banking crises that occurred between 1970 and 2000 are resulting in output losses of more than 20% on average if we look at all countries, and more than 30% of GDP in advanced economies.3 These results are in line with the BIS finding that the median discounted cumulative loss of output over the course of a crisis in the same period was about 19% of pre-crisis GDP.4
Now, the question of exactly how much regulation leads to the optimal outcome in terms of long-term growth is, of course, debatable. But let me underline that ambitious attempts have been made by the BIS, but also the OECD and others, to assess the net effect of recent regulatory reform measures, and the results generally point in one direction: that the net effect of reforms is positive.
In addition, let me also underline that the Basel Committee has not been blind and deaf to the worries expressed by the industry about excessive regulation. Many adjustments have been made, not least when it comes to the new liquidity regulation. It is also standard procedure that new regulations are subject to industry consultation and in many cases additional discussions also take place with the industry itself, as well as with investors, to avoid unintended consequences.
In this context, however, let me remind us all that the reactions we get from the banking industry are sometimes slightly biased, if I dare say so.
A telling example is the lobbying effort during the design of the Basel II framework. As part of that work, in 2003 the Committee consulted on a new securitisation framework, which, with the benefit of hindsight, turned out to be very weak. Yet the comments from the industry on the proposed securitisation framework were in general quite alarming.
Allow me to quote just a couple of the replies to the consultation proposal that the Committee received (all of which are publicly available):
- One bank wrote: "The prescribed risk weightings for securitisation exposure(s)-result in excessive risk weights compared to the economic risks of securitisation tranches, particularly for retail and mortgage portfolios." - This particular bank happened to incur $24.7 billion in losses from CDOs during the crisis.5
- Another bank wrote: "If adopted, the current proposal for securitisation will materially impair the ability of banks to distribute risk from their own balance sheets into the capital markets." - This bank incurred USD 13 billion losses in Q1 2008 and USD19 billion in writedowns on real estate and related structured credit positions.6
Let me emphasise that there is nothing special with these two examples. I can assure you that there are many more similar examples to quote - the message being that the proposed reforms were overly restrictive, would damage the market and reduce activity. This illustrates that we need perspective when assessing the feasibility of reforms.
To sum up so far: yes, there has been a strong regulatory reaction to the crisis, but as I see it, this is appropriate, given
- the pre-crisis regulatory framework,
- the costs crises give rise to, and
- the efforts that the Basel Committee has made to mitigate risks of unintended consequences,
The problem is that myopic observers tend to forget these aspects.
Are we there yet? What are the remaining challenges?
I would now like to change perspective slightly and ask, are we there yet? Have our efforts done the trick, or are there still challenges to be tackled?
Well, from a Basel Committee perspective I am pleased to be able to say that the Basel III framework is now agreed - in principle. This is a major achievement that all participating parties should be proud of. If I widen the scope, beyond the Basel III framework, and look at other parts of the reform agenda, it is obvious that the work on ending the "too big to fail" problem has been difficult, and that some work still remains to be done.
However, the reason we have not yet reached our goal is not lack of effort, but simply that the resolution of very large, cross-border banks is not easy. The main remaining issue here concerns how to ensure that global systemically important banks have sufficient capacity to absorb losses in resolution, without having to ask tax-payers to foot the bill. This work goes under the name of T-LAC, or total loss absorbing capacity. I find it reasonable to believe that there will be an agreement on a consultative document to be published in the context of the G20 summit in Brisbane.
So, viewed against the broad regulatory reform agenda put in place as a reaction to the crisis, it is fair to say that we are indeed seeing some light at the end of the tunnel. The main pieces are starting to come in place.
Unfortunately, concluding the post crisis reform agenda does not mean that we can lie down, relax and declare "mission accomplished". We need to look closely at the regulatory framework, remind ourselves of the reasons we put these measures in place, and ask whether they are delivering the right outcomes. And here I would like to focus on the interlinked issues of implementation and calibration. Let me start with some reflections on implementation.
For some time now, the Basel Committee has engaged in the process of monitoring and assessing how members implement what has been agreed by the Committee. The assessment work is carried out on a jurisdictional as well as on a thematic basis.
In the jurisdictional assessment we look at how Committee members have implemented the Basel standard - determining whether or not it is a fair reflection of the Basel III requirements. After an assessment has been thoroughly debated in the Committee, the final assessment becomes public.7
The assessments, and the publication of the results, have proved to be a powerful tool. To date, more than 200 adjustments have been made by member jurisdictions in response to findings raised by the assessment teams. In addition, the process has also generated a positive feedback loop, meaning that the lessons learnt from assessments are used to improve and clarify the standards. So far, the assessments have concentrated on the capital framework, but from 2015 onwards the scope of this work will widen further to include the implementation of the liquidity coverage ratio and the SIB-requirements.
However, for the new, stricter requirements to bring the benefits we are aiming for, it is important that they be properly reflected, not only in national legislation, but also at the level of individual banks. To use an analogy of car safety, if we are now providing banks with air bags, in the form of higher capital requirements, it is important that those airbags are actually activated in case of an accident. For this to happen, the sensors need to be functioning and well-calibrated. For banks, this means that risk weights need to signal appropriately the risks that individual banks actually face.
This aspect is captured in the Committee's thematic assessments. To put it simply, in these assessments we examined whether the banks' risk-weighted assets could be trusted. The results showed that banks' risk-weighted assets differ to an extent that goes well beyond what can be explained by business models and historical experiences. If we just take the banking-book results, two banks with exactly the same assets could report capital ratios that differ by as much as 4 percentage points.
The potential for differences this wide, particularly as they are derived from only a part of a bank's business, weakens confidence in the measurement of bank capital. Of course, this was not a total surprise. It was a reflection of what I mentioned earlier: that internally-modelled risk weights lead to capital not keeping pace with asset expansion. This has undermined the confidence in banks and the credibility of the concept of banks' internally-modelled risk weights. Ensuring consistency in the implementation of risk-based capital standards will therefore be a key factor in restoring confidence in banks.
The Committee is thus assessing bank capital ratios with a view to ensuring that they appropriately reflect the risks that banks face. There should be "truth in advertising" for the regulatory ratios that banks present. To achieve this, the regulatory framework needs to deliver readily comprehensible and comparable outcomes. In my view, these assessments, both the jurisdictional and the thematic that compares risk-weighted assets, are absolutely vital for achieving our goals. This will be an important focus for the Committee in the coming years.
I would now like to take a step further and focus on the link between implementation and how the system should be calibrated. Because my view is that there are a number of trade-offs at play here, which need to be taken into account.
For instance, if we don't implement the necessary changes and succeed in properly restoring the credibility of risk-weighted capital ratios, a more important role will have to be played by other parts of the regulatory system, such as the leverage ratio. For now, our working hypothesis is a regulatory minimum leverage ratio of 3%, but to me this is more of a place-holder. What the final outcome should be will depend on the calibration of the whole regulatory framework, in which the risk weights and leverage ratio are important pieces.
An important element in this calibration will be transparency - the more transparent banksare with methods and models to calculate risk weights, the better it will be for the credibility of the system as such.
If we widen the perspective further, I think there is also an interesting issue of calibration linked to the concept of going-concern capital requirements on the one hand, and gone-concern capital requirements on the other. When we discuss appropriate levels of TLAC we should keep in mind that the less we strengthen the credibility of the system for going concern capital requirements, the higher banks' gone-concern capacity to absorb losses will have to be.
So, to wrap up: I see no reason to pull the brake on regulatory reforms. We must not lose sight of the long-term benefits of limiting the costs to society that financial crises cause.
And, although a lot has been achieved, challenges still remain - especially when it comes to implementation, implementation monitoring and calibration of the whole framework. As I said earlier, I do not know with full certainty how all the different parts of the reforms will play out together. This further underlines the necessity to constantly monitor what is happening, very much in line with what the organisers of this conference are doing. And as financial systems have an amazing ability to reinvent themselves, regulatory reform is a never-ending task. Therefore, we need forums such as this conference to evaluate where we are, and where we should be going - hopefully, then, we won't have to make regulatory leaps quite as far as we were forced to this time.
1The figure SEK 17,000 =1,000,000*50%*3.4%, where 3.4% is the implied minimum core Tier 1 capital requirement under Basel I. The risk-weight for mortgages was 50% during this period. Under Basel II the implied minimum core Tier 1 capital requirement fell from 3.4% to 2% of risk-weighted assets, and was 2% in 2011-2012 when banks' internal average risk weights for mortgages was 6%. SEK 1,200 SEK = 1,000,000*6%*2%. For more information see Sveriges Riksbank (2013) Financial Stability Report 2013:2, November, pp 21-24.
3L Laeven and F Valencia (2013) "Systemic Banking Crises Database", IMF Economic Review 61(2), 225-270, IMF. Output losses are computed as the cumulative sum of the differences between actual and trend real GDP over the crisis period in relation to pre-crisis GDP.
4Basel Committee on Banking Supervision (BCBS) (2010) An assessment of the long-term economic impact of the new regulatory framework, August. The loss of output referred to represents the median cumulative discounted output loss reported by a number of academic studies assessed in the study, measured over the period from the peak to the end of the crisis. The output loss does not include permanent losses in GDP, ie where the GDP trend does not recover to the pre-crisis level.