JPMorgan: India PSU banks Basel 3 capital raising - a challenge, not a crisis

Our estimate of PSU banks’ capital needs indicate a Rs1.8tr requirement
by FY18E. Even if the government has to fully fund this (in a worst-case
scenario), it is unlikely to lead to any systemic issues – the incremental
impact on the fisc is ~0.3% if spread out over FY14-FY18. Moreover, our
assumptions are quite conservative and any upturn in the cycle will bring
this requirement down quite sharply. The main risk is for equity investors
– the apparently attractive valuations are significantly negated by the
possibility of significant dilution. Our preference, thus, continues for
better capitalized PSU banks like BOB.
 Mind the gap - Rs1.8tr. We estimate the PSU banks will need Rs 1.8tr
in outside equity capital by FY18. We capture a reasonably gloomy
scenario for this - PPOP ROAs at last five years' average (including
periodic pension hits), gross NPAs rising 50% from here and the banks
reaching 70% provision coverage including the slippage from
restructured assets. We also assume 9.5% CET1 ratios and a premium
for the larger banks, in line with RBI proposals. We've detailed our
methodology inside the report.
 Economic recovery not captured. We have assumed the economic
cycle staying weak through to FY18 (this is a stress test of sorts). A
recovery would reduce the problem on two counts - higher levels of
internal profit generation and the ability to raise capital externally.
Conversely, a shock to the economy could see our conservative asset
quality assumptions tested, especially in infrastructure. In this case, the
government's ability to put up the capital may be tested, though lower
growth would largely address the gap.
 Can be financed. The Rs 1.8tr gap should be financed by the
government over five years with reasonable comfort (assuming markets
remain as unhelpful as it is now). We estimate this to be ~0.3% of GDP
through to FY18E. They key however, is that the government has to
make this a continuous process rather than leave it for the end – which it
has done with an Rs150bn allocation in FY14.
 Dilution risk severe, divergence among banks. A systemic crisis looks
unlikely, but equity investors still need to be very wary of dilution risks.
The extent of dilution is expected to be widely divergent across banks,
given the very different starting positions on both capital adequacy and
ROAs. We are wary of low-CAR, low-ROA banks which look
apparently cheap - the post-dilution valuations may not look so
attractive. The equation changes dramatically if the economy and the
market recovers in time for these banks to be able to access markets for
the capital, but that’s a double-edged sword we are wary of.

��

-->
Methodology and assumptions
 This is based on a bottom up analysis for the banks. We have estimated PPOP on
the basis of a PPOP ROA and a steady asset growth. NPL provisions have been
separately estimated based on a peak NPL and coverage assumptions. The capital
gap is a terminal requirement for FY18E, assuming a one-shot infusion.
 We have assumed that the NPLs increase by 50% on an absolute basis from FY13
levels – for simplicity’s sake we have assumed that this goes up in a linear
manner through to FY18. This does imply a decline in NPL ratios, but we think
that's reasonable given the elevated NPL levels already. This assumption captures
any possible slippages from restructured assets.
 We’ve estimated provisions on the basis of 70% provision coverage. This is the
outcome of various discussions about loss-given-default ratios we have had with
bank managements and rating agencies over the years. The assumption is
reinforced by the RBI diktat in FY10 forcing banks to raise provision coverage to
70%. We have ignored write-offs in this calculation, so it is quite a conservative
estimate.
 We have assumed loan growth of 15% for all the banks on a normalized basis till
FY18E. This is based on a credit multiplier of 1.1-1.3 and nominal GDP growth
of ~12-14% till FY18E. We have also assumed, for simplicity, that PSU banks do
not lose market share to private banks

 The risk weighted asset growth is assumed at 15%, implying a static RWA/loan
ratio over this period. This is, again, conservative as India mandates a higher risk
weight for most retail loans than most other countries in the region. Given the
structurally improving asset quality led by the credit bureau, there is a likelihood
that this may be revised in the medium term: our analysis ignores this possibility.
 We have assumed the entire 9.5% CET1. While this may appear conservative, we
see very little scope for banks to raise Basel 3 compliant tier 1 bonds, so we think
this is realistic.
 While arriving at the retained earnings we have assumed a tax rate of 30% and
dividend payout of 20% based on the historic trends. The dividend payouts,
technically, could be adjusted downwards if the capital problem gets acute, but
that makes little difference to the ultimate outcome and could send out negative
market signals.