One of the lessons from the global financial crisis that started in 2008 was that the prevailing Basel II rules were pro-cyclical, and losses were recognised as they showed up on a loan or a portfolio of loans. This implied that during downturns, as losses increased due to defaults and credit impairments, banks took increasingly larger hits on their balance sheets, while boom times were associated with low loss-levels.

As an alternative, leading international regulators agreed to implement counter-cyclical provisions through dynamic provisioning, which relies on historical data for the various asset classes to statistically calculate how much additional provisioning would be needed on a regular, quarterly basis. Dynamic provisioning was first introduced in Spain in 2000, and is credited with protecting, to a large extent, the major Spanish banks from the impact of the financial crisis at its severest.

This change implied the introduction of a buffer, which is not subject to influence and business pressures, and seeks to smooth provisions along the business cycle and thereby avoid pro-cyclical effects. The intention is primarily to create a loan loss buffer in good times, to be used during bad times, and additionally help smooth credit growth.

The Reserve Bank of India has released a discussion paper on dynamic provisioning (DP) in the Indian context, and sought views from industry participants by May 15. It has suggested credit costs of 1.37 per cent based on a weighted average estimated loss of nine individual banks, based on a model portfolio of corporate, retail, housing and other loans (49 per cent, 17 per cent, 6 per cent and 28 per cent weights, respectively). The proforma credit cost estimate of 1.37 per cent would vary from bank to bank, with the estimated credit costs of 62 basis points on corporate loans, 267 basis points on retail loans, 27 basis points on housing loans, and 226 basis points on other loans.

If the actual specific provisions (SP) are lower than 1.37 per cent, the excess provisions would be transferred to DP, and vice versa, subject to certain conditions.

The initial DP would be made up of outstanding provisions made on standard assets and floating provisions. An assumption is that the banks have reached a 70 per cent provision coverage ratio (PCR).

The suggested framework for Indian banks is conservative and the DP framework will include an element of general and specific provisions. RBI has proposed that up to the level of normal loss given default or LGD (0.84 per cent), DP provisions should be considered as specific provisions and used for arriving at net NPAs (non-performing assets). Above-normal LGD to actual levels (1.37-0.84 per cent), DP provisions should be considered as general provisions, and therefore as tier-II capital.

RBI has further suggested keeping two DP accounts in the balance sheet:

DP account based on normal LGD — at the end of every quarter, the balance in normal LGD should be treated as SP; and

DP account based on downturn LGD — which could be treated as general provision and as capital.

By RBI's admission, the current estimation involved substantial approximations due to non-availability of required data, even though these are based on expert judgment of banks. Going forward, banks will be increasingly required to put in place mechanisms and systems to collect the data required for calibrating dynamic provisioning.

Dynamic provisioning will clearly lead to stronger bank balance sheets and an ability to absorb losses in the medium term, though the norms proposed by RBI could negatively impact earnings in the near term. Banks that have implemented, or are in the process of implementing an internal rating based approach would come up with estimated loss assumptions based on their portfolio mix that are different from the 1.37 per cent benchmark.

The author is Partner, Head of Financial Risk Management, KPMG in India.

comment COMMENT NOW