The complexity of Basel II and III has reached China. In a revolutionary turn within seven years, the Chinese bank regulator has introduced capital adequacy as the tool of choice for supervision and ensured that banks in the process remain focused on implementing all the pillars of the internationally developed Basel Accords. Will it really make Chinese banks more resilient?
In the past, Chinese banks were famously undercapitalised and their loan portfolios were of rather dubious quality. For example, in 2003, on average, the banking system showed an overall equity to asset ratio of just 3.25%. Rural credit cooperatives had produced a negative ratio of – 0.52%. Since then, the banks have had a lot of homework to do: from recapitalisation exercises to further improvements in internal controls and loan cleaning. In order to achieve capital adequacy ratios (CAR) for commercial banks of 10.2% by the end of 2011 (and even 12.7% for the total capital adequacy). Now almost all commercial banks are compliant with the Bank for International Settlements’ (BIS) required 8%.
To reach such levels, the regulators went out of their way to revolutionise the way banking and banking supervision is done in China.
Before, the central goal in the banking sector was gathering deposits. Therefore the loan-to-deposit ratio was the single most important performance indicator: it was the basis for paying bonuses, for developing business targets and for judging branch effectiveness. All relied heavily on this single figure because, in the absence of an efficient money market, asset growth could only be achieved through deposit growth. Furthermore, each branch had to be self-sufficient in terms of funding because deposit transfers between branches across provinces were forbidden. Regulation and compliance were all based on the loan-to-deposit ratio (set at 75%).
As a consequence, capital and capital adequacy were not on the mind of either bank managers nor bank regulators and capital constraints were unheard of. Such strong deposit growth disregarding asset quality and capital adequacy also favoured the building up of non-performing loans (NPLs).
Thus when the China Banking Regulatory Commission (CBRC) issued a regulation on capital adequacy for commercial banks (Regulation governing capital adequacy of commercial banks) in February 2004 it was seen as revolutionary by many observers. The central bank, People’s Bank of China (PBOC), had previously published a minimum CAR of 8% (prescribed in the earlier Commercial Banking Law) but did not give any detailed calculation methods or definitions of its components, and adherence was not enforced. Furthermore, the new regulations took into account Basel I and Basel II rules as well as the prospects of Chinese banks soon facing foreign competition (through the entry to the World Trade Organisation (WTO) in 2007).
The transition from quantitative growth (based on attracting deposits or on a funds constraint system) to a qualitative growth path (reflecting the quality of the assets held or a capital constraint system) formally took place until the end of 2006. At the same time the banks were required to increase levels of provisioning.
Even though the rules issued in 2004 took into account only some of the new developments in Basel II, the CBRC continued straight on its trajectory of still stricter requirements. Over time it has in fact managed to become even more stringent than Basel III (Graph 1). These efforts have pushed Chinese banks in a new direction and led their risk management to higher quality – albeit starting from a low base.
With its pillars 2 and 3 in addition to highly complex risk calculations, the capital accord dubbed Basel II was always going to have a strong impact on Chinese banks and their environment. This is mainly due to the fact that Basel II and the whole risk management framework are at a stark contrast to the Chinese banking reality. The challenges for China with Basel II range from capital and risk management to data and disclosure, as well as organisational structures, incentive compatibility (between banks and regulators), market-oriented supervision and the fostering of financial innovation.
Before the financial crisis erupted in 2008, the CBRC had clearly stated that it would first concentrate on implementing Basel I requirements and only use risk governance aspects of Basel II rules. With the crisis unfolding and the NPL build up following the stimulus package, CBRC took a very different stance. In fact, that “external pressure” was used as an excuse to write ever more strict rules. Regulators did not want to lose control of the country’s banks as in other G20 countries. Furthermore, it was a good opportunity to show the world what China is capable of – an important asset when Chinese banks need to convince foreign regulators that they are fit enough to open branches outside their turfs. For regulators, the implementation of Basel II can potentially increase information and bank-level data availability for a better and more accurate view and understanding of banks’ risks and potential losses. This in turn will enable them to react in a timelier manner.
The banks had already started to prepare in 2006. For banks, implementation will certainly bring higher costs at first, but pressure to comply comes from the regulators, the competitors and the investors. A risk-sensitive approach to business can help draw a competitive advantage, and smooth entry in other foreign markets. Better risk management can also help increase investors’ confidence. (Table 2)
Although Basel II is complex, costly, requires a high amount of historical data, gives much autonomy to banks and is calibrated to G-10 countries, implementing only the standardised approach (SA) across the Chinese banking industry makes little difference to the (relatively) risk-insensitive Basel I. Most conditions required for the full implementation of the SA in China are not yet fully realised: credit rating agencies are woefully under-developed, externally rated borrowers are few and unlikely to turn to banks for financing, corporate bonds data is poor, and financial disclosure remains scant if not fraudulent.
While the SA does not seem feasible, challenges with International Ratings Based (IRB) approaches definitely exist. Apart from the availability of data, which is a challenge to all banks, the fact that banks have yet to experience a full economic cycle adds a layer of difficulty. This in turn makes stress testing and calibration difficult.
In October 2008, the CBRC issued the first notice concerning Basel II implementation in China (Notice on supervisory directive concerning the first batch of new capital accord implementation). The notice considers five parts regarding the measurement of regulatory capital and the regulatory and technical requirements for classification of risk exposures, internal ratings systems, specialised lending ratings, credit risk mitigation and operational risk management. The notice was followed, two months later, by a further pack of eight notices concerning market risk measurement with the advanced approach, interest rate risk management on the banking book, liquidity risk management, information disclosure on the CAR, validation of the approach for operational risk, calculation of the CAR, securitisation exposures, and supervisory review of the CAR.
The notices in effect introduced the Basel II framework. Most of the requirements and content from the original Basel II Accord are found in Chinese regulations. In some aspects the regulators have adapted the regulations to fit more closely with the Chinese situation and environment (for example reducing the number of risk exposures classes in the banking book). Its regulations are more detailed insofar as they require more build up of structures and processes to achieve Basel II standards (which should come as no surprise since Chinese banks have more to catch up and CBRC has a more hands-on approach).
After lengthy discussions, the BIS proposed additional indicators and measures for regulators to manage other risks which featured prominently during the crisis (namely liquidity risks and capital quality). The BIS has proposed a new liquidity coverage ratio as well as a stable funding ratio. A further document highlights the quality of capital, calls for strengthened capital requirements, adds leverage ratios to the supervisory tools and advocates a counter-cyclical approach.
At first, CBRC acknowledged the BIS publications and published a Chinese version, but did not publicly comment in detail on the proposals. Then in the second half of 2011, the banks were flooded with large proposals for new regulations (among which are the Trial Management Rule regarding Liquidity Risk at Commercial Banks, and the Management Rule regarding the Capital of Commercial Banks which is for now in the form of an exposure draft). In effect, the requirements set forth are more stringent and implementation should be swifter than that proposed under the international Basel III document (Table 3).
Within four years, CBRC has completely changed its approach (Table 4) from a cherry picking model to a full and stricter implementation of Basel III, extended to all commercial banks. Will that push the banks to the edge of their capabilities?
At the end of 2011, the CBRC statistics for commercial banks, including city level and rural entities, and foreign banks, show that overall, the banks produced a healthy capital adequacy ratio of 12.7% (and 10.2% for tier 1 capital). Furthermore their current ratio reached 43% – well above the required 25%. Finally, their loan loss provisions covered over 278% of all non-performing loans.
As far as the above ratios suggest, the impact of more strict requirements will have more effect on governance. In fact much of the discussed asset securitisations and complex capital instruments never found their feet in China, or at least only in very limited volumes.
Chinese banks only started to establish risk management structures and rating systems at the beginning of the new century. This late start was the consequence of years of policy lending, capped interest rates, historical burdens and poor incentives to create sound banks. The newly established risk management units are now separated from sales departments and banks have centralised risk management and lending decisions against the resistance of previously fiercely independent bank managers. Banks have changed the incentive structures of relationship managers, started off-loading NPLs and finally have been able to share data through the PBOC credit registry. Despite the high hurdles that Chinese banks face, more and more are moving into risk management.
But all is not well, because the structures are not fully centralised: while the systems are common to all entities and levels of decisions within one group, the branches still retain a say in decisions through their local risk management units. These units sit awkwardly between two lines of responsibility, the first to their risk management counterparts in head quarters and the other to their local branch manager. Influence by local managers is still a reality and there is still no reporting lines separation between those managing risks and those doing business – thus the incentivisation of credit officers is challenged.
Other banks have chosen to centralise credit decisions in a few separate centres: for example Industrial Bank has centres in Beijing, Shanghai, Guangzhou and Fujian. In those cases, the branches have to submit credit applications to these centres. To ensure that its officers are made responsible for their decisions (often as or within a committee), Industrial Bank has also established a special committee investigating responsibilities. Its credit policies describe among others which industries should be focused on (along the lines of government policies). Other large banks such as China Minsheng Banking Corp still have to implement risk management systems to cover all of their activities, products, borrowers and risk types. China Minsheng Banking Corp has also drawn three lines of defence: business department, risk management and audit department.
Moreover, the bank’s boards of directors (BoDs) are now to be responsible for designing and implementing a risk management strategy. To this end they can use a number of committees, among which is a risk management committee. It is interesting to note however, that in a number of banks, there is not only one risk management committee, but one for the BoD and chairman, another under the president, and possibly another one at the headquarters. Observers might rightly question if such arrangements are efficient and can effectively increase the level of barriers to ensure good and independent risk management. Additionally, no bank has until now implemented a separation between business and credit reporting lines (all report finally to the president of the bank).
As an outside observer it is difficult to assess to what extent these credit rating systems are being used, how adequate they are and if they are being circumvented more than integrated into daily decisions. Furthermore while the professionalism of risk management departments will increase over time, it remains to be seen if the same happens with their independence and their responsibilities. Further to these, they also see more challenges. Credit and loss data are insufficient because a full economic cycle has yet to be experienced. A methodology measuring credit risk to support decision-making is lacking (still often based on collateral availability) and needs to be fully validated.
In line with the regulatory requirements, the banks are – while perfecting and strengthening their internal controls, credit monitoring and credit assessment systems – now establishing stress testing capabilities and taking steps to actively manage their capital (using economic capital, RORAC and EVA to allocate capital to different industries, borrowers, sub-portfolios, geographic areas and so on).
On the quantitative front, CBRC has conducted preliminary assessments similar to quantitative impact assessments of the BIS exercises abroad to gauge the potential impact of Basel II implementation on capital adequacy levels. Prior to the assessments, CBRC thought that capital adequacy levels would rise, but it appeared to be the contrary. Results for ICBC for example showed that the bank could actually lend more than under Basel I because it was more than adequately capitalized and its risk weights were better differentiated across exposures.
Because most banks show CARs which are already above the required 10.5% or 11.5% for smaller and larger banks respectively, the need for capital replenishment in the short term is rather limited. However, the new capital requirements are not the only ones that will slow down growth at Chinese banks (Table 5). With mounting fears over local government debts, real estate bubbles, trust lending and economic slowdown, the regulators have implemented further restrictions. Local government platforms will need to be adequately accounted for with appropriate risk weights and management controls, real estate exposures are welcome only for first homes, and trust products need to be moved on balance sheets. Facing a harsher environment, both liquidity and growth will be dampened. Moreover the banks could face a wave of fresh NPLs related in one way or another to the stimulus of 2008.
In addition to recapitalisation costs to defray the costs of higher NPLs, the banks will also have larger risk weighted assets to take into account (trust loans are required to be taken on-balance), which would probably mean a further capital hole of CNY1.2trn to cover these trust loans with sufficient capital. For the largest 17 banks (together covering 63% of the banking assets in China), rough calculations would imply an overall capital need of almost CNY2trn. But these back-of-the-envelope calculations fail to include liquidity, market and operational risks as well as take into account the profitability of banks (their profits are largely protected by the central bank’s base rates differentials).
Not only is the capital available going to increase, but costs are also on the increase: the costs of implementation are certainly high (at least CNY50m per small bank), but the costs of refinancing for banks will also increase, especially for those with poor standing and ratings. But more importantly, the question is less quantitative and should be more qualitative: banks should refrain from exchanging credit risk against model risk.
Apart from the quantitative impact as analysed above, the implementation of the Basel II accord will also have a qualitative impact on Chinese banks. Implementation will certainly refocus the rewards and incentives of officers and managers and delimit their responsibilities more clearly. Internal organisational structures are likely to be remodelled to comply towards a separation of reporting lines between risk management and operational departments. Information disclosure and transparency will encourage more stakeholders to review the banks’ activities and publications. The banks and the regulators will hold a wealth of data from which they can not only gauge risks but increase the level of financial intermediation. All these are likely to lead to a stronger credit culture and more proactive and dynamic risk management.
About the author
Violaine Cousin, fluent in Chinese, is a long-time observer and researcher of the Chinese banking system and financial sector. She has worked in various financial institutions as a credit analyst and consultant for Chinese banks. She recently published the second edition of “Banking in China” with Palgrave, where you can find more details on the above subject, and she contributes to the China Hand publication of the Economist Intelligence Unit every year. She is available to answer your queries or to provide research on issues of banking, ranging from products to stress testing and Basel III, and to write rating and credit analysis on selected Chinese banks. You can reach her at: email@example.com.