To understand NPLs we need to understand what major asset classes contribute to NPLs. The key NPL components can be classified into three broad asset categories: (1) cyclical industries, (2) infrastructure and (3) single name exposures or concentration risk.
Cyclical industries (moderate to low risk): Loans to cyclical industries account for nearly 60% of the total bank loans in the system. With the economic boom that India experienced during 2004-10 and attendant euphoria, the industry saw new capacity build up, overseas acquisitions and, of course, hunger for testing new derivative instruments. Unfortunately, a lot of it came on the back of debt and not counter-weighed by promoter equity. There were underlying assumptions of GDP floor rates of 6% or so and lower growth rates were assumed to be unthinkable. Demand from demographic profile of the population was overestimated. All in all, the margin of error left was very limited. Come 2011, the eurozone crisis strikes, export demand begins to taper off and NPL saga begins. High leverage across most industries, rapidly falling free cash flow (FCF)—those industries that grew faster than system average have higher NPL—and stretched working capital cycle simply accelerated the NPL build up. The key was lower demand and inability to pass on increased costs.
Today, the eurozone is poised to come out of depression. Export demand is beginning to pick up, especially in the textile sector. Auto and auto ancillary are likely to see increased demand on the back of 4%-plus agriculture growth. Cement, steel, construction being laggards will take a couple of quarters more to react. Corporates have taken a cue and have ploughed back profits, sold some assets, reduced debt and stayed away from greenfield expansions. All good news from credit perspective.
Overall, cyclicals are expected to be at the bottom of their cycle, and slow and steady recovery may be expected post two quarters.
Infrastructure loans (moderate risk): This has to be analysed from a few different angles as the problem is self-brewed poison in home backyard. First, the policy-related issues—be these fuel linkages, land acquisition, environmental and forest permits—have thrown this sector into a negative spiral. Second, these assets were financed for average debt maturity of 9 years against the asset life of 15-30 years. They had to come up for refinancing or restructuring and they indeed have. So why are we surprised? For the sake of records, infrastructure lending started big time in 2004-05. Third, the greed of promoters has also not helped the cause. There is borrowing in holding company through loan against shares transaction. This is reflected as equity in project SPVs. Banks lend against this so-called ‘equity’. So, on a consolidated basis, it is all debt with little equity. Fourth, overoptimistic demand projections, especially traffic growth forecasts in the road sector.
These four culprits have brought the sector to its present plight.
But the million-dollar question is, what is the loss given default (LGD)?
If the asset physically does exist and if we are convinced that India is an infra-deficit country and, therefore, long-term demand is intact, then the LGD is expected to be minimal. One must remember that unlike normal bank loan documents, the financing agreements for infrastructure projects are very creditor-friendly giving several options to lenders/creditors for resolution.
Do we expect infra loan restructuring to stop? Probably not. Given that there is a policy angle which needs resolution too, the restructuring may continue well into the next 6-8 quarters. But to re-emphasise, the ultimate LGD could be negligible if timely and smart financial engineering coupled with regulatory support is deployed.
Single name exposures (high risk, most worrisome): India Ratings was the first to highlight this risk back in August 2011 in its report titled “Greater Resilience Post-2008, but Mounting Concentration Risk”.
Concentration risk is defined as the exposure of top 20 entities (and not groups) as a percentage of equity. If it exceeds 200%, it is a red flag. This report was almost prophetic. Come March 2012, some of the large borrowers with hefty debt from banks turned NPLs. This has dealt a blow to the capital adequacy ratios of some banks. The potential risk from this situation could be quite worrisome and bank managements really need to work on methods to alleviate this risk given its lethal power to sink the banks.
Given India’s imperfect regulatory regime, ascertaining LGD or the time to recovery is next to impossible. India ranks a poor 121 in the World Bank scoring model on insolvency resolution.
If we consider entire reported restructured loans as NPLs from FY2007 onwards, the banking system will still continue to report healthy pre-provision operating profit (PPOP). But this time around, the magnitude is high. However, if we look at the past track record, the restructured loans have shown a recovery of about 85% in the last decade or so. Even if the assumption is 70% recovery with 10% restructured loans, the NPL increment over 2-3 years is still 3%.
Given the above background on different classes of NPLs and restructured loans, it will be necessary to run a stress test to see whether what is said above does bear fruit.
India Ratings has carried out a second round of stress test with three critical assumptions.
l CET-1 (common equity tier-1) downward adjustment: Reported CET-1 of each bank was reduced for (a) upper tier II, (b) perpetuals, (c) fully providing for pension liability, and (d) increased specific loan loss coverage ratio to 60%. The impact has been quite astounding for some of the large banks. In case of one of the largest public sector banks, the CET-1 reported stands at 9.8% but under stress case assumptions for CET-1 it reduces to 7.5% even before imposing NPLs.
l Reduced PPOP: PPOP reduced between 5-20% for each bank depending on CASA level. Lower CASA meant a 20% reduction in PPOP from March 2013 level.
l NPLs at 15%: The NPLs included approximately 10% from cyclicals and 3% infra. Additionally, two of the top 20 borrowers were taken as NPLs. The top two large borrowers may default for idiosyncratic reasons and not necessarily a sectoral downturn reason.
Results of stress test
l Nationalised banks have shown excellent risk pricing performance for cyclicals. At 10% cyclical stress, most banks report profits and are above CET-1 of 6.125%.
l Superimposing stress for single name/infrastructure most nationalised and a few private sector banks report losses for the year.
l Only five mid- to small-sized banks fail to maintain CET-1 of 6.125%.
To bring these five banks to a CET-1 of 6.125%, the bailout package costs $1.9 billion. The rest of the banks continue to remain profitable or at least remain above the minimum CET-1 of 6.125% even under such a severe stress test. This extends a substantial stability to the system and must be taken cognisance of while talking about the bank NPL situation.
Given the paltry magnitude of a bailout package one feels that the concern has been overdone to make it a crisis. I am reminded of a Sanskrit shloka that says “Yad bhavam tad bhavati—What you believe, so shall you become.”
Finally, one only hopes that by broadcasting banking crisis all the time and all over we do not end up inviting one where there is none.
The author is MD & CEO, India Ratings