'Looking beyond risk management as a cost'

With risk management in prime global focus after the fall of Lehman Brothers, there is a concerted effort to ensure banks are ring fenced to ward off risks associated with their operations, more so to ensure they provide the much needed stability to the economies.

Mr Alok Tiwari, Chief Executive Officer, Aptivaa Consulting Solutions, a global risk management firm, outlines what the proposed Basel III norms aims to achieve and the safeguards that it intends to put in to ensure compliance.

Almost all public sector banks in India are looking for recapitalization. This is so, despite them paying huge dividends to the government on a regular basis. Would not retention of earnings serve the purpose?

This issue will be well addressed in the proposed Basel III norms. In Basel III they have introduced two concepts — one is the counter cyclical buffer and the other, capital conservation buffer. Based on ratio of capital, central banks will specify a policy of dividend payout. At the moment there is a direct conflict of the roles as the Government is both the owner and the manager of the banks (as management is chosen and appointed by the government). The roles are conflicting, as, finally, the management has to act independently.

What is the opportunity cost of the two buffers? Will there be a resistance to the setting for these buffers?

That may not be the right view. The right way is to look at it as a bank's stability to generate ROE (return on equity). What kind of return can the bank generate on capital? This is the big question in the Western markets as post Basel III the ROE is expected to be down 3-4 per cent. One is on account of higher capital requirement and another is that Western markets are ring fencing the banks under the Dodd Frank Act in US and Vickers rule in UK. They say that banks cannot do certain types of activities. They are trying to ring fence the casino and utility banks.

If I am a utility bank, my returns are limited, as I am restricted to deposit and pure lending. If you see the last 6-8 years in global banking, returns have come from investment banking.

In India, there is no plan to ring fence the banks as our markets are not so evolved as the Western markets. So, in their trading books, at the maximum they do government securities, some merchant trading and inter-bank forex. But, the higher capitalization requirements would definitely lead to one to two per cent decline in ROE.

What is your take on the concept of central counter party for OTC (over the counter) products?

In principle, it is very good. As of now , in OTC you have no way of assessing counter-party risks. That is what people discovered post-Lehman.

Also, in general, at the systemic level, I have no view of the way credit risk is distributed within the system as also the counter party risk. The question is, how this will be implemented as OTC products would continue to evolve. Settlement is a concern, as unlike an exchange-traded product where you have settlement prices and liquidity available, here there will be huge issue in terms of valuation and then settlement prices.

Are any checks factored in for off-balance sheet transactions?

Off-balance sheet transactions are being addressed in two ways. One, they are increasing the capital on the structure of the instrument — the arbitrage between on and off-balance sheet. Second is the ring fencing, whether the activity is permitted under regulation. In Basel III they are going to include off and on balance sheet transactions in leverage ratio.

Certain Indian banks have very low NPAs (non-performing assets). What do you think they have done better than their peers?

NPA has three components. One is origination. If I have a strong and prudent origination, half the problem is solved. If I am choosy about what I underwrite, which I believe is supported by the origination. Second is the credit risk management — the frequency of reviews to monitor transitioning credit rating of debtors. The last is transparency of provisioning. There are ways of ever-greening. Sometimes, I have an account which is heading for NPA. I create another advance which can pay off the first. That also goes on. Some creativity is involved. Globally, it is a big issue. Even in the sovereign debt crisis, there is no agreement on how to recognise and deal with it.

At the strategic level, every bank makes its own growth choices. There are a set of banks which prefer growth in assets over asset quality. They only work on the percentage of delinquency. The assumption is that I am going to be in an economic growth cycle. If you are in a prolonged period of economic growth, this holds good as my balance sheet grows ten times and I am containing my delinquency below two per cent. In percentage terms it is fine.

The second type of banks, focus on asset quality. There is a trade off between growth and asset quality. The third would be extremely conservative banks, which focus only on quality.

When you do data analysis on Indian banks, you can slot them in these segments. The problem is, if you are highly quality conscious, the ability to generate returns is limited. If today, if one wants to lend to AAA, every one wants to lend to them. Banks should start looking at risk adjusted return, as on a risk adjusted basis one can generate superior returns, by lower credit risk. I am not saying go to the extreme.

The best example is Wells Fargo Bank of US which is extremely focused on the mid-market. Even during the crisis, it was one bank that stood rock solid and continued to generate returns. That is why you see large global banks focusing on mid market now as it is more diversified. If one picks good mid-market accounts it will offer better returns at relatively less risk.

Indian banks adopt various quantitative models to manage risk. How robust are the models or should each bank develop its own risk model?

There are two major issues. One is data. Indian Banks' Association initiated an important credit risk data consortium in 2008, which is on hold. If you are moving to advanced approaches in credit risk, you need quality industry level data.

Second, the model risk in India is going to be extremely high.

This is based on my experience with banks here as most are not going through proper model validation.

They simply buy third party software and then for the compliance perspective get it validated. Most banks procure an inbuilt software model, built overseas, implement it and sign off. Then someone does a quick validation on it. He does not structurally go into the assumptions or whether the model is back tested. Many times there is no previous data to test it. They are not going through a development lifecycle.

Risk is still seen as a cost or a checklist. The outlook is to meet minimum regulatory requirements. Even globally the perception is the same. There is nothing wrong in profit taking precedence, but at what cost.