The 1988 Basel I Accord created a level playing field for international banks in terms of a minimum recommended amount of capital.[1]It was probably the most successful financial standard ever conceived. More than 100 countries claimed to have implemented the agreement, and in many countries it was applied to all banks. Basel II was agreed upon in 2004. Many of its 216 pages and wealth of supporting documents[2]were on topics mostly relevant for larger, complex banks. It might be said that it was written largely thinking about 100 banks from 10 countries rather than 10 banks in each of 100 nations.
Towards the end of the Great Moderation[3]– the fall in the volatility of growth and other economic variables in advanced economies that started around the end of the 1980’s — Basel II’s more sophisticated elements (that permitted banks to use their own parameter estimations and models to measure risk), allowed some institutions to have very low capital levels just before the global financial crisis erupted. Capital could take alternative forms — some not particularly good at absorbing losses. Basel III came in 2010 as a response, tightening the definition of capital and boosting minimum required levels.[4]
Financial Innovation and Regulation in Emerging Economies
Arguably, these weaknesses in Basel II were less relevant for emerging markets (EM’s). In general, EM’s did not permit banks to fix capital levels using the more esoteric methodologies, and it’s often argued that EM capital markets tend to be less-developed, so banks didn’t dabble too much in new financial instruments, either investing in them (assets) or issuing them (liabilities or capital).
But this latter point puts the cart before the horse. It is banks that frequently spur financial innovation to arbitrage regulations, or make the most efficient use of their capital, depending on your perspective. When I made a life change and moved from being an academic in the UK to become a bank regulator in Latin America, I was surprised by just how intrusive banking regulation actually was — a reflection of a different legal system and the previous 1980’s financial crises. Banks need explicit authorization to use certain financial instruments. It was more the culture of bank regulation that limited the pace of capital market growth and less the other way around. Either way, many EM’s escaped much of the global financial crisis in part as their banks did not partake in the excesses of financial innovation.[5]
Basel III: Advanced and Emerging Economies (Home-Host) Issues
Still, aspects of Basel II were an uneasy fit for EMs and this carries through to Basel III. For example, Basel III continues the myth that bank regulation is about regulating a (single) bank. Pillar 2 of the Accord details what banks and supervisors should do, and it is written almost entirely from the standpoint of the supervisor of the overall or consolidated entity – sometimes called the lead supervisor.[6]Consolidated supervision is necessary, but it is not sufficient.[7]Most EM’s dictate that if an international bank is to operate in its territory it should be a subsidiary and that subsidiary should be regulated and supervised pretty much like a local bank.[8]While there are supporting documents that do discuss home-host issues, this important topic should be addressed explicitly in the Accord itself.
More specifically, Pillar 2 could outline some ground rules for supervisory cooperation. For example, if the subsidiary is large for the host country but small for the entire international bank, it makes sense to rely mostly on the host supervisor for onsite inspections and other elements of information gathering and supervisory oversight. But perhaps most importantly, there should be an agreement about early intervention if the bank (or one or more of its subsidiaries) runs into trouble and liquidity and capital ratios dip. In the worst scenarios, the incentives of different supervisors will be at odds and conflicts are almost inevitable as national authorities attempt to protect domestic depositors and deposit insurance agencies.[9]An agreement on early actions well before such scenarios emerge would be a useful addition to Pillar 2.
Turning to the actual quantitative requirements (Pillar 1), there are also interesting home-host issues. One is that what is a domestic currency for the international bank is not the same as for a subsidiary located in another country. There has been much discussion of late about the capital requirements on government bonds. Basel III has an exception for government bonds denominated and funded in domestic currency — domestic banks may be able to treat them essentially as risk free. It was argued by some EM regulators that as the international bank was forced by its lead supervisor to treat those bonds as foreign currency assets this resulted in an asymmetry with costs for EM’s. While this is true, it also raises the more general issue of the treatment of government bonds on bank balance sheets. In some cases, governments have used regulated banks to “gamble for resurrection,” and large quantities of government debt have built up on bank balance sheets complicating any subsequent debt restructuring if that proves necessary.[10]A rule implying symmetrical treatment between EM and home supervisors, but combined with a recommended limit of the amount of government assets on bank balance sheets, might be the way to go.
A further issue regarding home-host issues is on Pillar 3 of Basel III, which is about what banks should disclose: transparency. Consider the case of a domestic EM bank that is listed on the local stock market and has issued bonds domestically. Now suppose an international bank buys the local bank and it becomes an integral part of the international conglomerate. This process may well result in a significant loss of information. The subsidiary may be delisted from the local market and may obtain liquidity from the group rather than through domestic issuance. In effect, a not particularly transparent guarantee (depending largely on reputation) replaces transparent market information. Material subsidiaries do have to disclose information, but material to whom? For any subsidiary of an international bank that is important to a host country, the host supervisor should decide.
Basel III: The Quantitative Requirements
Turning to the central quantitative requirements (Pillar 1) of Basel III, EM’s may find themselves between two stools. On the one hand, the approaches that allow banks to determine firm risk and other parameters imply a considerable degree of autonomy for banks, in sharp contrast to some regulatory cultures and a challenge in terms of supervisory capacity for many EM’s. On the other hand, as there are not that many firms or assets that have credit ratings, the approach that allows banks to use rating agency opinions to gauge risk may not be very useful. Giovanni Majnoni and I recommended a middle-way.[11]This would be to ask banks to rate clients based on a standardized scale, much as they already do for extensive public credit registries found in several countries, including in Latin America. Moreover, this data can be harnessed to consider the calibration of the actual quantitative requirements. This data is already used to monitor loan quality and assess provisioning requirements — expected losses. It can also be used to assess capital requirements — normally thought of as the buffer against unexpected losses.
Lima 1: a Basel III compliant standard for Latin America and the Caribbean
At the same time, it should be noted that Basel III has gone a considerable way to accommodate different countries’ needs. There are many different options and sub-options available within the new Accord. On top of that, there has been active discussion of late regarding proportionality and its close cousin materiality.[12]On this note, EM’s should be careful for what they wish for. Basel I was an extremely successful standard that created a level playing field for banks across the world. But with all of Basel III’s options and sub-options, there is a danger of the very notion of a standard being lost. It is notable in Latin America that different countries are choosing quite different options but all say they will be compliant with Basel III. Several years ago, in a meeting of Latin American supervisors in Lima, I suggested a standard for LAC — I called it Lima 1. The idea of Lima 1 was not a competing standard but rather an agreement between regulators, in a region with increasing financial integration, on how to comply with these complex recommended minimum rules, choosing similar approaches while taking advantage of the excellent data and human capacity within the region, and at the same time ensuring that an actual standard was maintained.[13]As the region contemplates how to implement Basel III, this remains as relevant today as it was then.
[1]This blog reflects my views and draws on several papers written over the years including when I was an academic and a banking regulator. Some of the arguments expressed align with the recent report Making Basel III Work for Emerging Markets and Developing Economies. I am indebted to Thorsten Beck, Liliana Rojas-Suarez and the other members of the task force that produced that report for a set of interesting discussions on these and other related topics.
[2]The website listed 103 documents at the time, many of which supported Basel II — see Basel II and Developing Countries: Sailing through the Sea of Standards for a discussion. An updated comprehensive version of Basel II was published in June 2006.
[3]This phrase was coined by James Stock and Mark Watson in 2002 and refers to the fall in volatility of growth and other economic variables in advanced economies around the end of the 1980’s, which ended with the 2007 global financial crisis.
[4]I abstract here from other significant reforms such as the counter-cyclical buffer and the liquidity and leverage ratios. While Basel III was first agreed in 2010, it was revised in 2011 and then it took a further 6 years to agree on a set of further issues. The Basel Committee for Banking Supervision finally published Basel III: Finalising post-crisis reforms in December 2017.
[5]See Prudent Banks and Creative Mimics for a model of financial innovation and a discussion of the trade-offs.
[6]This point carries through despite the Pillar 2 enhancements made in Basel III. While these do refer to the importance of good firm-wide governance procedures, they do not talk about the challenges posed by complex group structures, nor do they discuss challenges of supervisory cooperation. In the Basel Committee’s Corporate Governance Principles for Banks there is some discussion of group structures. Tellingly para. 99 states that “In the case of a significant regulated subsidiary (due to its risk profile or systemic importance or due to its size relative to the parent company), the board of the significant subsidiary should take such further steps as are necessary to help the subsidiary meet its own corporate governance responsibilities and the legal and regulatory requirements that apply to it.” One would imagine that all subsidiary boards should take steps to meet relevant responsibilities, but again it’s telling that materiality references the importance to the group and not to the host country where the subsidiary may be located.
[7]To make the point, Lehman Brothers had hundreds of legal entities operating in over 50 jurisdictions. Winding up this bank took more than 75 bankruptcy procedures — see Lehman Brothers’ Bankruptcy Lessons learned for the survivors.
[8]Many EMs take this view due to the holes in the international financial architecture, such as inconsistencies in bankruptcy regimes.
[9]Bank failures create incentives for local authorities to ring-fence assets for the benefit of local liability holders. While “living wills” may assist, assuming they were agreed in advance between different supervisors, an agreement on early warning would be better still. See “International banks, cross border guarantees and regulation” with Giovanni Majnoni in International Financial Instability: Global Banking and National Regulation, eds: D Evanoff, G. Kaufman and J. LaBrosse, World Scientific Publishing, 2007 — from a Chicago Federal Reserve Bank Conference, for discussion.
[10]Argentina 2002 is a case in point — see Argentina’s Avoidable Crisis: Bad Luck, Bad Economics, Bad Politics, Bad Advice.
[11]The Internal Rating Based approaches imply a lot of autonomy for banks, while the Standardized Approach that employs credit ratings may not yield much in terms of linking capital to risk — due to a lack of rated claims. We named our middle way, the Centralized Rating Approach, which would build on public credit registries See the paper “Reforming Bank Capital Requirements: Implications of Basel II for Emerging Countries”,with Giovanni Majnoni. Economia. (Journal of Lacea) 2005.
[12]See for example Proportionality in bank regulation and supervision – a survey on current practices
[13]The fact that Basel III is a recommended minimum standard is important. For example in Reforming Bank Capital Requirements: Implications of Basel II for Emerging Countries, we found that attempting to recalibrate Basel II’s Internal Rating Based approach formula to some Latin American countries resulted in very high requirements due to greater systemic risks (covariances) and so we questioned the assumed 99.9% tolerance value. We suggested some EM’s may wish to lower this value resulting in moderately higher requirements. Further work on this employing the excellent public credit registry information in the region would be valuable.
Leave a Reply