Danareksa Bank Mandiri (BMRI IJ Rp5,250 HOLD) Turning positive, but more work is still needed

The turnaround continues; upgrade to Hold
We raise our FY10-11E EPS estimates by 14-16%, reducing the estimated provisioning charges to align our forecast with the company’s guidance. Progress has been made and Mandiri has done well in restructuring its legacy bad debt problems. Yet more still needs to be done. Looking ahead, additional reversal of provisioning shall come from two large debtors: Garuda and Domba Mas – but because the timing is still unknown we exclude it from our forecast. Despite the increased optimism, history suggests we remain cautious on the potential NPL increases - there have been net loan downgrades over the last four years! A relatively low NIM remains the bank’s weakness despite the slight improvement expected this year. Following our adjustments, our DDM-derived TP is raised to Rp5,250. This implies 2.7-2.4x FY10-11E PBV, a 42% premium to its 3-yr average or at par with its domestic peers ex-BBCA – too pricy, in our view, given the bank’s inferior NIM. Hence, among the SOE banks, we prefer the cheaper BBNI to Mandiri despite the latter’s turnaround story.

Garuda and Domba Mas: near-term catalysts
There are likely to be more provisioning write-backs going forward. So far, a total of Rp33trn has been written off – with Rp6.1trn in the last 3 years alone. Recovery is underway, yet it is doubtful the bank can fully recover all bad debts. Two imminent reversals shall come from Garuda and Domba Mas, together totaling around Rp2.4trn, albeit the timing is still unknown. The reversals are positive: 9% and 7% higher FY11-12E EPS (assuming a split over 2 years) and therefore ROE in excess of 24% - although concerns remain on the potential increase in NPLs with no significant improvement in operational indicators such as NIM. Indeed, last year’s lower NPLs were mainly due to write-offs - a strategy that could lead to further write-backs in later years. The higher amount of loans restructured also suggests that loan channeling needs to be significantly improved.

A focus on high yielding consumer loans
We expect Mandiri’s NIM to expand by a mere 20 bps this year. This improvement shall come from a lower COF rather than higher asset yields - albeit the latter should remain in excess of 9%. Nonetheless, Mandiri’s NIM will still be inferior relative to the industry benchmark. The lower COF will come from the re-pricing of maturing high cost deposits. This is thanks to the stable BI rate, with Mandiri likely to see an improvement in its low-cost CASA as fears of liquidity problems ease. The bank will also work hard to grow its high yielding consumer loans - a successful strategy in the past 4 quarters – and this should help keep the NIM above 5.3% we believe. As a result, the dominance of low yielding corporate loans will decline further. Note that the bank’s consumer and micro loans have grown 23% and 40% respectively over the past 3 years, or far exceeding the 16% growth for corporate loans. Over the next 2 years, we expect the proportion of retail loans to increase to 30% from 25% currently, with the asset yield stable at 9.0-9.2%.

In need of more capital!
With 12.4% tier-1 capital, sustaining 18-20% 3-yr CAGR loan growth will mean additional capital raising – be it either a sub-debt issuance or rights issue. The latter would lead to 2-5% higher CAR, depending on the structure of the rights issue. There are basically three possibilities – 1) government divestment; 2) a rights issue which dilutes the government’s stake or 3) a combination of a rights issue with government divestment. An added incentive for boosting its shareholders’ base is the lure of 5% tax incentives. EPS dilution would be minimized to 2-6% this year, with potential earnings upside of 1-5%. This assumes the rights issue is conducted in 4Q10 and the proceeds are placed in SBI in the first year and channeled into loans the next year. Like the BBNI rights issue, we think Mandiri stands to benefit from conducting a rights issue since it will facilitate brisker loans growth without the bank having to take on new debts. Our scenario analysis favors a combination of government divestment and a rights issue, limiting dilution to around 2-3% in the first year and with 5% expected higher EPS growth the next year, in our estimates.