The stress in key sectors like power and steel is too acute for credit costs to be delayed to FY18
ICICI’s near-term challenges persist. We cut FY17e/FY18e earnings due mainly to the frontloading of credit costs into FY17 and the cutting of PPOP due to lower growth. We see value in ICICI at a 0.9x P/BV (price-to-book ratio), given the strength of its retail franchise, but the asset quality headwinds may not dissipate soon and could drag on the stock in the next one to two quarters. Long-term investors should stay invested, in our view.
Asset quality frontloading. We now frontload our credit cost assumptions to FY17. The stress in key sectors like power and steel is too acute for credit costs to be delayed to FY18, in our view. We also see the risk of further regulatory pressure on classification; this may not be through another asset quality review (AQR), but through measures such as the harmonisation of classification across banks. Our FY16e, FY17e and FY18e credit costs are 185bp, 189bp and 74bp, respectively.
Positive news flow not enough: There has been some positive news flow around the stressed sectors – a rally in global steel prices, UDAY, MIP. However, we believe the stress is too deep for this to solve the problems, even if they are alleviated. A core problem is that growth has remained weak for a few years, and high real rates do not help.
Cutting growth estimates: We cut our pre-provision operating profit (PPOP) estimates for three reasons: (i) the accelerated NPL classification hurts both loan growth and margins; (ii) the outlook for corporate loan growth remains uncertain; and (iii) fee growth remains muted as the corporate segment continues to be weak. We fear that the adjustment could be long.
FY18e bounceback: We forecast a sharp bounceback in FY18 to an ROE (return on equity) of 18% – this underpins our value argument. Our forecast assumes that FY17 credit costs absorb most of the legacy issues, with some tail risk from provision coverage spilling into FY18. We assume the bank will continue to be able to isolate its legacy issues from the front book growth; this remains the challenge for the stock over the longer term.
Key estimate changes
We now build in higher credit costs of 189bp vs. 106bp earlier for FY17, given the heightened stress in sectors like power and steel and the accelerated classification of NPLs. Credit costs for FY18e are now lower, at 74bp vs. 101bp earlier, as we believe the bank will absorb most of the pain in FY17. Our methodology for calculating credit costs remains unchanged, at 30% stressed assets in infra and steel and LGD of 35%. The key changes are: (i) we frontload the credit costs in FY17 by apportioning 95% of losses in FY17 vs. 40% earlier, given the accelerated recognition of NPLs; (ii) we ignore the stress recognised in FY16; and (iii) we add some tail risk from provision coverage spilling into FY18.
Corporate loan growth is expected to remain weak, given the subdued environment and high real rates. The accelerated classification of NPLs (non-performing loans) will also negatively impact loan growth and margins. We expect retail loans to be the key growth driver, with strong growth in home and auto loans. We now factor in lower loan growth in FY17/FY18, at 17%/18% vs. 19%/21% earlier.
Corporate fees remain weak, given the lack of activity. The expected acceleration in fee growth has not crystalised because of weak corporate growth. In the absence of a material recovery in the investment cycle for FY17, we expect fee growth to remain muted in FY17 and hence cut our growth estimate to 13% vs. 17% earlier.
Gains from stake sale
We were expecting FIPB (Foreign Investment Promotion Board) approval for the general/life insurance stake sale to come through in 4Q16, which would have helped in absorbing higher credit costs. Given that the stake sale has been deferred, we now crystalise it in FY17. Part of the stake sale—4% to Premji Investments—was approved in Q3FY16. The balance—2% to a Temasek sub and the 9% stake sale in general insurance to Fairfax Holdings—is pending approval. Given the delay in approval, we now build in gains of R21 bn in FY17.
Earnings bounceback in FY18e
We expect a sharp improvement in ROE in FY18, to 18% from 15% in FY17. Given the accelerated NPL recognition in FY17e, we expect credit costs to normalise in FY18. We now build in lower credit costs of 75bp vs. 101bp earlier. We are conservative and build in some tail risk of provisions spilling into FY18. The sharp ROE improvement underpins our value argument in FY18.
—J P Morgan