INDONESIA BANKS OVERWEIGHT. The going gets better. 22 Sep 2016

Transcription

1 INDONESIA BANKS OVERWEIGHT Sector Reinitiation Financials Sector 22 Sep 216 The going gets better Reinitiating with OVERWEIGHT as momentum turns bullish Our positive outlook on Indonesia banks is built on 1) firmer macro circumstances (a U-stretched recovery), 2) renewed loan growth momentum, 3) NPL volatility peak-out, 4) mitigated NIM stress, and 5) rising profitability metrics. Our economist team expects GDP growth at 5.2%/5.4% in 216F/17F, a continued uptrend vs 4.8% last year bottom. Domestic consumption will lead loan growth momentum to re-emerge with micro and consumer leading followed by SME and corporate. We expect 1) a better YoY loan growth at 13-14% for F (216F: 11.6%), 2) improving credit cost outlook, and 3) good cost control to more than offset F NIM stress, which ultimately should drive earnings growth upward to 14% YoY on average for F vs 216F s -1.8%. NPL cycle bottom to offset F NIM stress The bottom of asset-quality deterioration is encountered already in 1H16, NPL blow-outs are unlikely for banks prudent credit cost and sufficient NPL buffers (3.2% PPOP/loan, 18.2% Tier-1 CAR, and 4% excess NPL buffers). NPL ratio is to peak in 3Q16F with 3Q17F the prelude of consistent NPL downtrend. Albeit on structurally low-rate milieu and regulatory risk (to a less extent now) pressuring NIM, our calculation reveals a +9.2% of net impact to earnings on better loan growth (+1.9%), better credit cost (-7bp), and NIM loss (-32bp) movements in F relative to 216F. PRICE PERFORMANCE Source: Bloomberg, BCA Sekuritas Igor Nyoman Putra igor.putra@bcasekuritas.co.id ext 178 Liability spread matters more than asset spread Tightening LDR is not an immediate issue yet. On a longer-term perspective, structurally declining excess liquidity will force loan-deposit growth multiplier to narrow to 1.1.x (from now 1.5x), proving liability spread management to become increasingly imperative than asset spread. Banks with solid deposit franchise (high CASA and low LDR) would outperform. Large banks would extend their dominance vs small banks in this space. A fruitful tax amnesty is one chief catalyst serving as a game changer that we have not factored in. Improved earnings drivers lead to a sector re-rating With healthier operating environment leading to upgraded earnings visibility, we expect F ROE to stabilize at 15.8% for the first time since 21 s consistent drops. This would be trailed by P/B multiple and eventually promote a sector re-rating. By preference order, we prefer large banks (BMRI, BBRI, and BBNI) over small banks (BBTN, BNGA, BDMN, and BJBR) for their positive swing in bottom-line and stronger position to address structural long-term issues in NIM and LDR. In the large banks, we like BMRI for significant earnings recovery, retail loan mix focus, and best non-interest income mix. We like BBTN amid small banks for robust loan growth prospect, understated funding strength, and largely recovered NPL. The sector is trading at 1.9x/12.7x 217F PBV/PER, below mean/14% discount to earlier peak, which appears as an attractive entry point. BUY selective banks to catch up with the anticipated up-cycle. Downside risks include NPL relapse, NIM declines, and slower loan demand than estimates. Exhibit 1. Indonesia banks valuation table Ticker Rating TP CP Mkt cap. Net profit (IDRbn) EPS gr. (%) P/E (x) P/B (x) ROE (%) Yield (%) (IDR) (IDR) (USDm) 216F 217F 216F 217F 216F 217F 216F 217F 216F 217F 216F 217F BBCA NR na 15,175 28,152 19,718 21, BBRI BUY 14,2 12, 22,539 24,948 27,951 (1.8) BMRI BUY 13,45 11,375 2,28 14,51 17,976 (28.6) BBNI BUY 6,4 5,475 7,774 1,46 11, BDMN HOLD 4, 3,81 2,78 2,883 3, BBTN BUY 2,4 1,99 1,63 2,249 2, BNGA BUY 1, 815 1,559 1,485 1, BJBR HOLD 1,575 1,675 1,237 1,61 1, Sector Overweight 85,853 77,441 88,578 (1.8) Source: Bloomberg; BCA Sekuritas estimates * BBCA is not rated ** Closing prices as of 21 September 216 MARKET DATA Ytd 1M 3M 12M Absolute 16.7% -.3% 19.6% 29.3% JCI Return 16.3% -1.4% 9.5% 22.1% Relative.4% 1.1% 1.1% 7.2% Source: Bloomberg, BCA Sekuritas

2 Contents Sector section Overweight on multiple catalysts.. 3 Have we encountered the bottom of asset-quality weakening?.. 16 Could lower credit cost and better operating environment compensate for the NIM stress?. 24 Tapering excess liquidity not an immediate issue.. 26 Improved earnings drivers = sector re-rating?. 34 Investment idea and risks 42 Valuation charts and forecasts tables 46 Company section Bank Mandiri - A catch-up play.. 5 Bank Rakyat Indonesia - Sustaining value-accretive growth 55 Bank Negara Indonesia - Undervalued performer 6 Bank Tabungan Negara - Extending growth momentum 65 Bank CIMB Niaga - Moving forward. 7 Bank Danamon - Not yet there 75 Bank Jabar Banten - Profitable but pricey. 8 Appendix

3 Jan-5 Jul-5 Jan-6 Jul-6 Jan-7 Jul-7 Jan-8 Jul-8 Jan-9 Jul-9 Jan-1 Jul-1 Jan-11 Jul-11 Jan-12 Jul-12 Jan-13 Jul-13 Jan-14 Jul-14 Jan-15 Jul-15 Jan-16 Jul-16 Overweight on multiple catalysts We reinitiate coverage on Indonesia banks with OVERWEIGHT following multiple positive catalysts which we think have yet to be fully priced-in. Albeit recent share prices rally (JAKFIN was up 14.7% YTD) seems to be premised on exaggerated optimisms (tax amnesty finally being passed into the law and Indonesia s cabinet revamps i.e. the return of veteran economist Sri Mulyani), these hopes are not present for no reason. We believe tailwinds on recent relaxed monetary policy and administration reforms (i.e. BI rate cut, loosened LTV, accelerated gov t spending, gradual tax amnesty receipts etc) will gradually improve our economy. Since a fuller economic recovery is yet to be felt, we continue to see potential upsides from current level with the following drivers serving as positive catalysts: 1. Better macro circumstances providing firmer ground for business doings; 2. Renewed loan growth momentum on structurally declining interest rates and continued gov t s infra-related focus; 3. Peaking NPL volatility (and in turn lower credit cost) is to drive earnings; 4. Mitigated NIM stress, which is better than what the market is expecting; 5. Rising profitability metrics (ROE and ROA) warranting higher valuation multiples. Better macro circumstances The latest BI rate had stood at 6.5% (down from Nov-14 s peak of 7.75%) before the reference rate was changed to a lower rate of 5.25% of 7-days reverse repo rate (7-days RR rate) in Aug-16. This displays a softening policy rate trend which could be achieved since inflation and IDR volatility are well contained, coupled with concurrently easing global liquidity. We envisage a structurally lower interest rate going forward and view a more stable macro interest rates to result in stronger ground for business pick-ups. Banks would be better off since they are set to extend credits. Exhibit 2. Benign and stable macro-rate trajectory (%) BI Rate CPI Source: Bank Indonesia; BCA Sekuritas In the last five years, GDP has consistently trended down from 6.4% (in 21) to 4.8% (in 215) since macro volatility was intensifying monetary policy tightening, IDR instability, widening CAD, etc. On this backdrop, we expect aggregate banks (banks we cover) to generate negative earnings growth of 1.8% YoY this year (a negative reading the first time since 25), down from.4%/5.4% YoY in 215/14. However, since lately the macro setting appears to have recuperated, albeit gradually, this should work out the other way around leading 3

4 1Q9 2Q9 3Q9 4Q9 1Q1 2Q1 3Q1 4Q1 1Q11 2Q11 3Q11 4Q11 1Q12 2Q12 3Q12 4Q12 1Q13 2Q13 3Q13 4Q13 1Q14 2Q14 3Q14 4Q14 1Q15 2Q15 3Q15 4Q15 1Q16 2Q16 216F 217F 218F (.5) (1.5) (2.1) (2.7) (3.) (3.2) (3.6) (3.3) (3.2) (3.2) (3.) (3.1) (3.) (2.5) (2.2) (2.1) (2.1) (2.1) (2.2) (2.4) (2.4) Q1 2Q1 3Q1 4Q1 1Q11 2Q11 3Q11 4Q11 1Q12 2Q12 3Q12 4Q12 1Q13 2Q13 3Q13 4Q13 1Q14 2Q14 3Q14 4Q14 1Q15 2Q15 3Q15 4Q15 1Q16 2Q16 216F 217F 218F earnings growth profile to being better. As reference, our economist team forecasts a higher GDP growth of 5.4% YoY for F relative to a 5.2% YoY estimation this year. Exhibit 3. followed by an expected GDP rebound and (%) GDP growth Envisaging c.idr225tn of assets repatriation in the next 12M since the law being effective, boosting GDP potentially Source: Bank Indonesia; BCA Sekuritas estimates Meanwhile, the structural issue of blowing-out current account deficit (CAD) also seems to have abated, allowing us to anticipate a narrower trend onwards. Exhibit 4. narrowing CAD should present macro stability (%) CAD (% of GDP) (1) (2) (3) (4) (5) Bolstering the market confidence, less macro volatility and improving CAD Source: Bank Indonesia; BCA Sekuritas estimates Overall the Indonesia macro development is going to skew towards a more sustained and gradual improvement. This comforts us as the going will only get better from here. We could only see little evidences of major hindrances that could impede macro improvement. Accordingly, we are sanguine Indonesia banks, as one major economy driver, should be one of the main beneficiaries to capture the estimated up-cycle going forward. 4

5 Jan-5 Jun-5 Nov-5 Apr-6 Sep-6 Feb-7 Jul-7 Dec-7 May-8 Oct-8 Mar-9 Aug-9 Jan-1 Jun-1 Nov-1 Apr-11 Sep-11 Feb-12 Jul-12 Dec-12 May-13 Oct-13 Mar-14 Aug-14 Jan-15 Jun-15 Nov-15 Apr-16 Sep-16 Exhibit 5. Improving trend (%) 1yrs govt bond rate have been reflective in improved government bond yield Source: Bank Indonesia; BCA Sekuritas Exhibit 6. Our base-case economics forecast Year-end 31-Dec F 217F GDP/capita (USD) - Nominal 3,659 3,523 3,377 3,525 3,75 Real GDP (% YoY) Private consumption (% YoY) Government consumption (% YoY) Gross fixed capital formation (% YoY) Exports (% YoY) (1.97) (1.25) 2.45 Imports (% YoY) (5.84) (.5) 3.5 Inflation rate (% YoY) - year-end BI rate (%) - year-end * 6.* 7-Week Repo rate (%) - year-end Rupiah/US Dollar - year-end 12,189 12,44 13,788 13,372 13, Current account as % of GDP 3.2 (2.9) (2.6) (2.21) (2.35) Fiscal balance as % of GDP 2.3 (2.4) (2.8) (2.4) (2.5) FX Reserve (USDbn) - year-end Source: Bureau National Statistics (BPS); Ministry of Finance (MoF); Bank Indonesia; Bloomberg; BCA Sekuritas estimates Renewed loan growth momentum Our optimism over restored loan growth is predicated on tailwinds from government spending (9.4% of nominal GDP) and loosened monetary policy initiatives/economic stimulus implemented in the last twelve months. Next year more growth drivers should originate from private sector first since 1) infra-related spending has been aggressive, 2) various economic packages are coming online, and 3) interest rates are getting benign. 5

6 Dec-2 Jun-3 Dec-3 Jun-4 Dec-4 Jun-5 Dec-5 Jun-6 Dec-6 Jun-7 Dec-7 Jun-8 Dec-8 Jun-9 Dec-9 Jun-1 Dec-1 Jun-11 Dec-11 Jun-12 Dec-12 Jun-13 Dec-13 Jun-14 Dec-14 Jun-15 Dec-15 Jun-16 Exhibit 7. Lower policy rate subsequently boosted loan growth (% YoY) Loan growth (LHS) BI rate (RHS) (%) Jan-4 Oct-5 Jul-7 Apr-9 Jan-11 Oct-12 Jul-14 Apr-16 Declining policy rate had been followed by higher loan growth in the following quarters, vice versa Source: Bank Indonesia; BCA Sekuritas By type of loan, consumptions (Jun-16: 47% of total loan) and working capital (28%) would be in the first leg to rebound (or micro, consumer, and SME by loan segment definition). In the second leg, corporate loan would follow the suit. While banks may still cautiously provide loan to SME (on high-npl characteristics) since our economy is gradually recuperating, banks with loan mix dominated by retail (consumer and micro) is better positioned to see accelerated loan growth in 2H16-1H17F. Afterwards corporate lending should enjoy loan momentum more in the later leg. Exhibit 8. System loan growth by type of loan (% YoY) Working capital Investments Consumptions (1) 7 Source: Bank Indonesia; BCA Sekuritas A visual inspection at Exhibit 9 hints that most banks are exposed with consumer-lending but only a few with high-yield micro-lending. Meanwhile, banks with more exposure to interest-sensitive (auto, personal, mortgage, etc) would enjoy strong loan demand since we anticipate a rebound in domestic consumption. 6

7 F 217F 218F F 217F 218F Exhibit 9. Indonesia banks loan mix (2Q16) (%) Corporate SME Micro Consumer (ex. mortgage) Mortgage Others BMRI BBRI BBNI BBCA BBTN BNGA BDMN BJBR BBRI, BDMN, BBTN, and BJBR stand out to enjoy the first leg of loan momentum given their retail loan exposure Source: Companies data; BCA Sekuritas We project Indonesia banks to register % YoY of loan growth for F, rising from estimated 11.6% YoY this year, with loan-tonominal-gdp multiplier at 1.4x for the subsequent two years. Our forecast could be deemed as conservative since we are using a lower loan multiplier against 1-years average of 1.7x. The reason of lower multiplier is due partly to the fact that we are already in the new normal (mid-teens growth vs +2% in the past). Exhibit 1. Aggregate loan growth to recuperate (% YoY) Loan growth (LHS) Deposit growth (LHS) LDR (RHS) (%) Loan growth is reaching the new normal of mid-teens mainly on high LDR Source: Companies data; BCA Sekuritas estimates Exhibit 11. albeit with loan multiplier lower vs historical mean (x) Loan-to-nominal GDP multiplier 1 years average Average Our loan-to-nominal-gdp multiplier over the next two years are lower than the 1-year historical means, partially implying our conservatism.5. Source: Companies data; Bank Indonesia; BCA Sekuritas estimates 7

8 Further, our hypothesis of stronger loan growth momentum in the private sector is corroborated by our internal analysts expectation: 1. Auto: automotive sales is expected to grow 1% YoY from this year anticipation of 3% YoY driven by low interest rates and recovered consumer confidence. 2. Consumer discretionary: consumption power is to rise better next year since macro stability is already attained coupled with multiple economic stimulus programs. 3. Property: pre-sales is to grow 12% YoY in 217 vs 216F s 8% YoY spurred by lower interest rates trajectory, relaxed LTV, property tax certainty, and tax amnesty inflows. 4. Cement: domestic cement sales is forecast to expand 7% YoY next year (216F: 4% YoY) driven by property up-cycle, tax amnesty, and infra-related projects. One may raise a question of why our projected loan volumes are only decent if macro condition is recuperating. Our reply is that the present chief concern of asset-quality risks is paramount and needs to be managed properly, suggesting banks become stricter on credit underwriting standards and are judiciously selective on lending. This eventually ought to lead to cautiously optimistic loan growth in the near term. Therefore, an aggressive loan practice (and, hence, faster loan growth) during this gradual economic recovery could lead to further assetquality deterioration, in our view. Thus our loan growth pick-up in F is fairly sensible and may pose upside risk in 218F if economic performance turns out better than what we project. Exhibit 12. Indonesia banks loan growth forecast (YoY %) Q16 216F 217F 218F BMRI BBRI BBNI BBCA Large BBTN BNGA (3) BDMN (5) (7) (8) BJBR Small Aggregate Source: Companies data; BCA Sekuritas estimates Then the next question is, if decent loan volumes mean only a little rebound in loan growth, would this lead to tapering top-line growth vs past years? We appreciate that slower loan volumes (vs past years with +2% YoY loan growth) imply slower top-line growth relative to a historical context; however, we argue that asset quality plays a quite pivotal role for banks earnings improvement than loan growth. Our sensitivity analysis reveals that a credit cost change of 1bp affects earnings by % vs % for every 1% change in loan growth. Plus, asset-quality deterioration also poses risk of further NPL slippage or restructured loan relapse. We already assume net interest income growth at single-digit rate in F (detailed analysis in the following subsection), as opposed to double-digit growth in the past one decade. Thus, finding the right growth would be the best approach rather than finding high growth. 8

9 M16 216F 217F 218F M16 216F 217F 218F Lower credit cost to drive earnings The year of 216F is going to be the poorest year of credit cost trend since we predict the credit cost to peak at 2.1%, the highest since the last credit down-cycle in 29, and we could only expect a more encouraging trend of declining credit cost going forward. We project credit cost to fall 45bp YoY in 217F and further by 25bp in 218F which in turn will likely drive earnings. With NPL ratio drifting lower, the concomitant NPL coverage would also increase to % from this year estimated 137% (6M16: 129%). Exhibit 13. Credit cost to peak this year and improve onwards (%) Credit cost (LHS) NPL coverage (RHS) 3. (%) 25 A quite conservative credit cost assumption we have with 218F forecast still higher than Source: Companies data; BCA Sekuritas estimates Loan-loss reserves would continue to climb up beyond this year, providing banks with more provisions during any bad times. Exhibit 14. with NPL ratio gradually trending lower (%) NPL ratio (LHS) Loan-loss reserves (RHS) (%) Banks are expected to stay prudent as reflected in our projected rising loans-loss reserves Source: Companies data; BCA Sekuritas estimates The main recipe (benign policy rate) for an asset-quality enhancement is no longer missing. Exhibit 15 intimates that NPL growth is fairly correlated with BI rate. Elevated policy rate in the past suggests higher NPL growth volatility in the subsequent period whereas lower BI rate was responded by lower NPL growth later. Therefore based solely on this reason, this may suggest to us that one main factor for restoring asset-quality, a benign policy rate trend, has made a comeback. Accordingly this would be followed by benign credit cost going forward. 9

10 Exhibit 15. NPL growth would decline amid falling policy rate (% YoY) System NPL growth (LHS) BI rate (RHS) (%) NPL growth is lagging policy rate trend with positive correlation (2) 1 Jan-4 Oct-5 Jul-7 Apr-9 Jan-11 Oct-12 Jul-14 Apr-16 Source: Companies data; BCA Sekuritas estimates A quick glance at below aggregate covered-banks balance sheet (Exhibit 16) implies that Indonesia banks have been superior on maintaining their balance sheet strength during several credit down-cycles with the last cycle interestingly lasting for 31 months. Comparing to the last credit cycles, present circumstance (2Q16) comforts us since 1) banks loan buffers (PPOP as of loans) and tier-1 CAR stay solid with 2) loan growth remaining strong at double-digit growth, 3) maintained +1% NPL coverage, 4) prudent FX liquidity management, and 5) manageable NPL ratio. The banks ability to continue to sustain balance sheet strength further validates our anticipation of positive credit cost prospect. Exhibit 16. Aggregate banks balance sheet comparison during three credit cycles (%) Chg Chg '13-15 chg 2Q16 GDP growth (1.1) (3.) (.6) 5.2 BI rate (1.6) IDR/ USD (IDR) 9,283 9, ,133 1, ,585 11,838 13, ,3 Loan growth (9.7) (16.1) (9.7) 12.3 NPL coverage (84.4) (25.5) 129 Tier-1 CAR (.8) PPOP/loans* (3.6) (.1) 6.5 NPL ratio FX LDR (27.) (18.6) (.3) 67. Cycles duration 18 months 9 months 31 months na Source: Bank Indonesia; Companies data; Bloomberg; BCA Sekuritas * Annualized figure for 2Q16 Mitigated NIM pressures Banks 2Q16 NIM in aggregate has been rising +6bp/32bp QoQ/YoY driven largely by small-four banks of +17bp/54bp (BBTN, BNGA, BDMN, and BJBR) and to a less extent large-four banks of +4bp/26bp (BMRI, BBRI, BBNI, and BBCA). The latter experienced less of NIM increases for their high CASA mix. Several factors affecting NIM performance YTD were that of government s push for a lower lending rate and easing policy rate. As for next year, NIM stress will likely mount since room for further time deposit (TD) rate reduction is narrowing while interest rates are going to continue trending lower. 1

11 1Q5 3Q5 1Q6 3Q6 1Q7 3Q7 1Q8 3Q8 1Q9 3Q9 1Q1 3Q1 1Q11 3Q11 1Q12 3Q12 1Q13 3Q13 1Q14 3Q14 1Q15 3Q15 1Q16 3Q16 2Q11 3Q11 4Q11 1Q12 2Q12 3Q12 4Q12 1Q13 2Q13 3Q13 4Q13 1Q14 2Q14 3Q14 4Q14 1Q15 2Q15 3Q15 4Q15 1Q16 2Q16 Exhibit 17. Large 4 vs small 4 NIM performance (%) Large 4 banks NIM Small 4 banks NIM Large-banks would have recorded a negative QoQ trend in 2Q16 had BBRI not accrued interest income from the late KUR payment Source: Companies data; BCA Sekuritas * Large-four banks include BMRI, BBRI, BBNI and BBCA; Small-four banks include BBTN, BNGA, BDMN and BJBR Longer term justifiably that the new normal loan growth and structurally lower interest rate would pose NIM risks, bringing NIM into the new low level this is expected and inevitable. We have reflected the structural NIM drops in our forecast over the next 1 years led by 1) declining average interest-earning assets (IEA) yield amid low-rate environment and 2) limited room for further deposit cost cuts. Despite of the NIM pressures, however, we believe the impact to bottom-line is still manageable on improving trend in other variables i.e. loan growth, credit cost, and etc. Further details can be found on section Could lower credit cost and better operating environment compensate for the NIM stress?. Exhibit 18. Aggregate NIM looks strongly correlated with JIBOR 3M (%) JIBOR 3M (LHS) Aggregate NIM (RHS) (%) Structurally lower interest rates trend is to be apparent in JIBOR Source: Bank Indonesia; Companies data; BCA Sekuritas * Aggregate NIM is only available until 2Q16 as for the latest 11

12 F 217F 218F 219F 22F 221F 1Q5 3Q5 1Q6 3Q6 1Q7 3Q7 1Q8 3Q8 1Q9 3Q9 1Q1 3Q1 1Q11 3Q11 1Q12 3Q12 1Q13 3Q13 1Q14 3Q14 1Q15 3Q15 1Q16 3Q16 Exhibit 19. as well as 1-year gov t bond yield trend (%) 1yr gov't bond yield (LHS) Aggregate NIM (RHS) (%) and bond yield that will further bring NIM down Source: Bank Indonesia; Companies data; BCA Sekuritas * Aggregate NIM is only available until 2Q16 as for the latest In the short-to-medium term (4Q16-217F) we are of the view that banks NIM risk should intensify on the followings: 1. Continued competition for high-quality customers. 2. Naturally decreasing interest rates via benign macro state (i.e. Jibor, government bond yield, etc). The two above-mentioned points are already reflected in our estimates as we assume NIM for aggregate banks to gradually abate from the peak in 215 of 6.1% to 5.3% by 221F (2Q16 aggregate NIM: 6.4%). We have not incorporated all loan segments being in the single-digit rate this year, but only for mortgage, corporate, and some low-tier SMEs. Next year SME segment largely is assumed to go single-digit rate. On a positive note, the issue of single-digit rate appears to have receded with our meeting with banks confirming of subsiding noises. From all of our discussion above, nevertheless, we think there is one factor that could be understated and thus is not factored in, which is the banks ability to mitigate the NIM pressures through: 1. Reducing expensive deposit rates continuously. 2. Improving IEA mix from loan. 3. Being growth-focused on high-yield loan mix. Exhibit 2. No longer a high-nim story (aggregate) (%) Aggregate NIM Large 4 NIM Small 4 NIM 7.5 Normalizing towards a new low level of NIM NIM peak * Source: Companies data; BCA Sekuritas estimates * NIM surge on accounting changes for micro lending interest income recognition 12

13 Jun-1 Sep-1 Dec-1 Mar-11 Jun-11 Sep-11 Dec-11 Mar-12 Jun-12 Sep-12 Dec-12 Mar-13 Jun-13 Sep-13 Dec-13 Mar-14 Jun-14 Sep-14 Dec-14 Mar-15 Jun-15 Sep-15 Dec-15 Mar-16 Jun-16 Jun-1 Sep-1 Dec-1 Mar-11 Jun-11 Sep-11 Dec-11 Mar-12 Jun-12 Sep-12 Dec-12 Mar-13 Jun-13 Sep-13 Dec-13 Mar-14 Jun-14 Sep-14 Dec-14 Mar-15 Jun-15 Sep-15 Dec-15 Mar-16 Jun-16 The first attempt banks could offset loan yield drops impact to NIM is by reducing TD mix (i.e. TD rates reversions). Under current condition largefour banks have room of bp to further cut time deposit (TD) rates reaching the bottom in mid-213 while small-four banks have bp. Small banks command more room since their deposit franchise is comparatively weaker than large banks. With this room left, we are expecting to see another 1bp and 175bp of deposit rate cuts within the next 12 months for large and small banks, respectively. We do not assume current TD rate to revert back to that of similar bottom level in 213 due to the new high LDR. Exhibit 21. Weighted TD rate comparison among large banks (%) BMRI BBRI BBCA BBRI, Jun-13, BMRI, Jun-13, BBCA, Mar-13, The bottom level of deposit rate in 1H13 was a period before the prelude of tightening monetary policy in the last credit cycle From TD rate bottom, ideally BMRI/BBRI has space left to slash deposit rate by another 186bp/173bp Source: Companies data; BCA Sekuritas Note: BBNI does not disclose TD rate data in particular Exhibit 22. Weighted TD rate comparison among small banks (%) BDMN BBTN BJBR BDMN, Jun-13, BBTN, Jun-13, BJBR, Jun-13, while BDMN/BBTN/BJBR has room to reduce TD rate by 63bp/166bp/343bp Source: Companies data; BCA Sekuritas Note: BNGA does not disclose TD rate data in particular Nonetheless, we are doubtful the TD-rate reduction could persist in the longer term as liquidity is getting scarce with banks already being in the same ground following TD-rate cap (1bp/75bp premium from BI rate for BUKU III/IV). To reflect the new liquidity level, we assume higher TD rate by 85-15bp than the past bottom. Further, following the changed policy rate to 7-days RR rate (formerly BI rate), this remains unclear whether the TD-rate cap will also refer to 7-days RR rate of 5.25% (BI rate: 6.5%) - one positive catalyst we have yet to reflect in our forecast. Second attempt to mitigate the NIM stress is by growing loans more aggressively than other IEA. As loan commands the highest yield among other interest-earning assets, the overall blended IEA yield ought to also expand. Over time this should be relevant since excess liquidity is drying up, prompting banks to extend loan by replacing other IEA with loan. Large-four banks appear to be standing out due to comparatively low loan participation in IEA mix as opposed to small-four banks, indicating ample room to grow more loan books more aggressively relative to other 13

14 China Hong Kong Average Singapore Indonesia India Korea Thailand 214 2Q16 218F 214 2Q16 218F 214 2Q16 218F 214 2Q16 218F 214 2Q16 218F 214 2Q16 218F 214 2Q16 218F 214 2Q16 218F interest-earnings asset. Plus, this would allow banks to have more NIM resilience in the longer term. Higher loan portion in small banks IEA could be attributed to their weaker deposit franchise. Hence, improving asset mix by growing more loans and alleviating other IEA mix (i.e. liquid assets) are key drivers to help impede NIM decrease. Exhibit 23. Loan participation in interest-earning assets (%) Loan as of interest-earning assets Large banks are outperforming most of small banks in this metric BMRI BBRI BBNI BBCA BBTN BNGA BDMN BJBR Source: Companies data; BCA Sekuritas estimates If a rising loan as % of IEA should bode well for NIM, structurally how far could banks grow their loan book replacing other IEA? We make a comparison within regional banks to measure the potentials Indonesia banks could go. Exhibit 24 displays that Indonesia banks still have lots of rooms to improve their IEA mix. Since Thailand banks have presently the highest loan participation of 77% within ASEAN banks, we assume this being the maximum level of loan within IEA. In our model Indonesia banks would need to grow loan book by 13.3% CAGR F (higher vs other IEA CAGR at 6.9%) to reach the same level to those of Thailand banks. As reference, the loan portion of IEA for Indonesia banks stood at 7% as of 2Q16. Exhibit 24. Loan of IEA comparison within regional banks (%) Loan of IEA (215) More rooms to go means more NIM cushions Source: Companies data; Bloomberg; BCA Sekuritas * Note: Indonesia banks data uses our covered aggregate banks The third attempt to reduce NIM pressures is by growing high-yield loan and fixed-rate loan more aggressively i.e. consumption, micro, and pensioners loans. This effort should bolster the overall blended loan yield. In the short-to-medium term, banks with retail loan participation (microand auto-lending) are of more resilience to downward re-pricing for their fixed-rate characteristics (<3 years of loan tenor) while those exposed to mortgage (1-2 years of loan tenor) should be more resilient in the longer term. The retail loan is also high-yield relative to other loan segments. 14

15 Exhibit 25. High-yield loan mix focus (2Q16) (%) High-yield loan of total More aggressive in growing retail loan book means more NIM shelters BJBR BDMN BBRI BMRI BBNI BBCA BNGA BBTN Source: Companies data; BCA Sekuritas On the flipside one argues that high-yield loan could encounter more downside risks for its above-average high lending rate since the government has been pushing for a low-rate lending environment. We believe there exists low likelihood of this to transpire as this loan segment has different risk and overhead cost profiles. In our view, the government moreover is also fully aware of this practice which could jeopardize banks profitability and capital, whereas the government may be disinclined to inject capital given current austerity measurement. Therefore the government may want banks to maintain their profitability and eventually capital. Already we also witness implementation of government-backed credit People s Business Credit (KUR) and subsidized housing program (FLPP) introduced for the sake of public interest, thus the government will need banks to stay profitable in order to run these programs. Discussion over next year interest rate KUR subsidy (if it does still exist) as well as FLPP funding mix stay vague. 2 15

16 Mar-16 Jun-16 Have we encountered the bottom of asset-quality weakening? Yes, we have encountered the bottom of asset-quality deterioration in 2Q16. Volatility remains in 2H16-1Q17F, partly on seasonal reasons, but we will not witness any higher/sharper growth/trend that we have already experienced in 1H16. In terms of ratio, peaking NPL would be visible in 3Q16F. A consistency of softening NPL trend onward is expected to be encountered starting in 3Q17F. First of all the trends of low commodity prices, IDR fluctuation, mild GDP, high interest rates, and high CPI have all participated in growing NPL book. We should ask how these trends are going to behave in the future. While we do expect an improvement, however, recently the trends have reversed favorably. They will continue to trend better as we are sanguine of a healthier economic growth trajectory onwards (explained in the early section). System NPL en route to recovery Before delving deeper into banks-specific under coverage, we examine the existing NPL trend in the banking system to see where we stand now. As of Jun-16 system NPL grew at 3% YoY with YTD average growth (from monthly figures) slowing down to 28% YoY vs 215 average of 32% YoY. We argue this is a gradual recovery process and will continue to do so in the next 12 months. Interestingly, the system s NPL development is in fact already en route to recovery amid soft GDP attributable to improved credit underwriting and strong NPL buffers. Going forward the followings are why we are upbeat on the potential asset-quality improvement: 1. GDP uptrend and lower interest rate. 2. Banks carrying out prudent lending practice perennially. 3. Well-contained CPI (not deflation) trajectory. 4. Relaxed NPL definition (explained in the subsequent subsection). 5. Less IDR USD volatility. GDP uptrend and lower interest rate. In the past GDP appears to be inversely correlated with NPL growth. We expect GDP growth to continue the uptrend (2Q16/1Q16/4Q15: 5.2%/4.9%/4.8%) which should augur well for future NPL recovery. Exhibit 26. System s NPL growth stabilized despite softening GDP growth (% YoY) NPL growth (LHS) GDP growth (RHS, inverted) (% YoY) (1) (2) 4 (9) (1) (12) 14 (5) Having been inversely correlated, a mild GDP growth in the past three years has led asset-quality restoration to move forward slowly Source: Bank Indonesia; BCA Sekuritas 16

17 Jun-7 Dec-7 Jun-8 Dec-8 Jun-9 Dec-9 Jun-1 Dec-1 Jun-11 Dec-11 Jun-12 Dec-12 Jun-13 Dec-13 Jun-14 Dec-14 Jun-15 Dec-15 Jun Mar-16 Jun-16 Banks carrying out prudent lending practices. In the past two policy tightening cycles (25-6 and 28-9), loan growth has slowed down in part to cope with high NPL volatility. Slowing loan growth was eventually followed by decelerating NPL growth in the subsequent periods. Banks have carried out similar lending practice thus far during the recent tightening, and we believe we should see a similar trend of NPL growth deceleration soon. Exhibit 27. as loan growth well behaved in the past years (% YoY) NPL growth Loan growth (1) (2) Source: Bank Indonesia; BCA Sekuritas Well-contained CPI reading (but is not subject to deflation). To some extent easing CPI should prop up people s purchasing power which ultimately ought to underpin loan momentum, and help fix the new NPL formation. Below Exhibit expresses that CPI normally precedes NPL growth by 8-9 months, meaning higher CPI would result in higher NPL growth in the next 8-9 months, vice versa. We have a positive read-through for NPL growth in the next few months given the latest CPI data was soft at 2.8% in Aug-16 and is likely to continue to stay mild going into F ( % as per our economist team projection). Exhibit 28. with reduced inflationary pressures helping control NPL trajectory to some extent (adjusted chart *) (% YoY) NPL growth (LHS) CPI (RHS) (%) (1) (2) (3) Source: Bureau National Statistics (BPS); Bank Indonesia; BCA Sekuritas * The above NPL growth is 9 months ahead of the CPI As of Jun-16, the system non-performing loan (NPL) and special-mention loan (SML) growth have genuinely stabilized to the level off the peak. SML growth (one level before slipping into NPL), which is usually an indicator of NPL potentials, meanwhile has meaningfully dropped to 9.6% YoY equating to the lowest level since Oct-15 s 51.5% YoY. NPL growth has also been down to 29.8% YoY vs the earlier peak in Nov-14 of 4.3% YoY. This concludes that the system NPL has remained in check and on track to see further positive progress onward. 17

18 Jan-93 May-95 Sep-97 Jan- May-2 Sep-4 Jan-7 May-9 Sep-11 Jan-14 May-16 Jan-12 May-12 Sep-12 Jan-13 May-13 Sep-13 Jan-14 May-14 Sep-14 Jan-15 May-15 Sep-15 Jan-16 May-16 Jun-7 Dec-7 Jun-8 Dec-8 Jun-9 Dec-9 Jun-1 Dec-1 Jun-11 Dec-11 Jun-12 Dec-12 Jun-13 Dec-13 Jun-14 Dec-14 Jun-15 Dec-15 Jun-16 Exhibit 29. System s NPL and SML growth off the previous top (% YoY) SML growth NPL growth Oct-15, Nov-14, (1) (2) (3) Having SML being one leading NPL indicator, the exhibit on the left demonstrates a receding NPL outlook Source: Bank Indonesia; BCA Sekuritas Exhibit 3. System s NPL ratio by loan type (%) Working Capital Investments Consumption Total NPL Consumption NPL remains below to those of other segments Source: Bank Indonesia; BCA Sekuritas Reduced IDR USD volatility. We understand that in the last two years IDR has been volatile against USD on both global instability as well as Indonesia macro fluctuation. While some companies have IDR-USD mismatch between their revenue and cost, we think this has somewhat been instrumental in exacerbating the NPL trend. Despite that, we expect IDR to stabilize onward and that our system FX LDR (with USD the lion s share) is actually already improving. System s FX vulnerability is also mitigated now with 14.4% of FX loan from total loan in Jun-16. Exhibit 31. On top of firmer IDR ahead, banks FX exposure reduced (%) FX LDR (LHS) FX loan/total loan (RHS) (%) Learning from the past AFC 1998 and GFC 28, FX portion has continued to improve Source: Bank Indonesia; BCA Sekuritas 18

19 3Q8 4Q8 1Q9 2Q9 3Q9 4Q9 1Q1 2Q1 3Q1 4Q1 1Q11 2Q11 3Q11 4Q11 1Q12 2Q12 3Q12 4Q12 1Q13 2Q13 3Q13 4Q13 1Q14 2Q14 3Q14 4Q14 1Q15 2Q15 3Q15 4Q15 1Q16 2Q16 Aggregate banks risky loan has stabilized In this section we will be using the term of classified loan and risky loan more frequently as we feel this is our more conservative approach of how we analyze the current and outlook NPL development. We define classified loan as NPL and SML, while risky loan includes classified loan and currentcategory restructured loan. Classified loan growth has been on the decline since 2Q14 s peak. Nevertheless, following the introduction of a relaxed NPL definition by Indonesia Financial Services Authority (OJK), the classified loan improvement has in fact been simultaneously followed by risky loan staying elevated, resulting in a gap. This suggests to us that the elevated current-category restructured loan has kept the risky loan remains elevated. Exhibit 32. Aggregate banks classified loan improving while risky loan stabilizing (% YoY) Classified loan growth YoY Risky loan growth YoY (1) (2) 2Q14, Q15, Q15, The gap had widened starting in 3Q14 and further expanded in 3Q15 when the OJK introduced the relaxed NPL definition policy Source: Companies data; BCA Sekuritas As far as the loosened NPL definition is concerned, the three pillars fulfillment required for loan-quality definition (financial condition soundness, repayment ability, and business prospect) was loosened into one pillar only, which is the repayment ability. As such, the NPL definition is eased concurrently with easiness of doing loan restructuring, allowing banks to smooth earnings by delaying NPL potentials and in turn resulting in smooth credit cost. The widening gap starting in 3Q14, as earlier highlighted, signals the current restructured loan may be rising quite rapidly. It is true for current restructured loan being at an advanced stage, but not true for the sharp increase given risky loan growth is now at a lower stage off the 3Q15 peak driven by declining current restructured loan growth. A further confirmation is visualized best in Exhibit 33. We view the expected restored macro situation to speed up the asset-quality healing process. 19

20 4Q1 1Q11 2Q11 3Q11 4Q11 1Q12 2Q12 3Q12 4Q12 1Q13 2Q13 3Q13 4Q13 1Q14 2Q14 3Q14 4Q14 1Q15 2Q15 3Q15 4Q15 1Q16 2Q16 Sep-8 Dec-8 Mar-9 Jun-9 Sep-9 Dec-9 Mar-1 Jun-1 Sep-1 Dec-1 Mar-11 Jun-11 Sep-11 Dec-11 Mar-12 Jun-12 Sep-12 Dec-12 Mar-13 Jun-13 Sep-13 Dec-13 Mar-14 Jun-14 Sep-14 Dec-14 Mar-15 Jun-15 Sep-15 Dec-15 Mar-16 Jun-16 Exhibit 33. Restructured loan decomposition intimates current restructured loan growth has peaked out (% YoY) Total restructured loan growth Restructured SML growth Restructured NPL growth Current restructured growth The lion s share of restructured loan is under current-category (49% as of 2Q16) (3) (8) Source: Companies data; BCA Sekuritas Moving into a narrower NPL definition of classified loan (vs risky loan); it is encouraging to see that SML (decreasing consistently) and NPL (off the peak) growth are showing a reassuring trend. Exhibit 34. NPL growth stabilizing with SML trending down (% YoY) NPL growth SML growth 4 2Q15, (1) 3Q15, (2) Source: Companies data; BCA Sekuritas * There were kitchen-sinking in 2Q15 (BBNI) and 2Q16 (BMRI), inflating the trend Based on these studies accordingly we conclude banks risky loan is already at an advanced stage and starts showing an encouraging trend. We are upbeat this would continue to translate into good NPL ratio in the following years. We expect NPL ratio to peak in 3Q16F (2Q16: 2.7%) before further waning off to 2.8%/2.7%/2.5% in 216F/17F/18F with the peak of NPL growth already past. 2

21 1Q12 2Q12 3Q12 4Q12 1Q13 2Q13 3Q13 4Q13 1Q14 2Q14 3Q14 4Q14 1Q15 2Q15 3Q15 4Q15 1Q16 2Q16 2Q8 3Q8 4Q8 1Q9 2Q9 3Q9 4Q9 1Q1 2Q1 3Q1 4Q1 1Q11 2Q11 3Q11 4Q11 1Q12 2Q12 3Q12 4Q12 1Q13 2Q13 3Q13 4Q13 1Q14 2Q14 3Q14 4Q14 1Q15 2Q15 3Q15 4Q15 1Q16 2Q16 216F 217F 218F Exhibit 35. likely to lead NPL ratio to gradually peter out by 218F (%) NPL ratio Source: Companies data; BCA Sekuritas estimates Banks command adequate NPL buffers We appreciate the positive reading from both system s and aggregate banks NPL development. However, such a risk of NPL relapse remains since asset-quality movement is profoundly volatile and unpredictable with risky loan at an advanced stage. Though we argue the scenario is less likely to occur, in the event of worsening asset-quality scenario, banks equipped with strong NPL buffers should stand out. Overall we see that banks NPL coverage is improving coupled with ample NPL buffers (2Q16 aggregate: 2.7% NPL ratio vs 3.5% loans-loss reserves, 3.2% PPOP of loan, and 18.2% Tier-1 capital adequacy ratio), which are a positive indication. Seasonality suggests that banks are normally under-provisioned in 1H and vice versa in 2H (Exhibit 36). Hence, our covered Indonesia banks do not seem to us as being under-provisioned during 1H16. The loans-loss reserves uptrend also signifies banks continuous prudent credit cost policy, which is a positive move amid current challenging setting. Exhibit 36. Proactively being prudent = rising loans-loss reserves + high credit cost (%) NPL formation Credit cost Loans-loss reserves Credit cost has been attempted to match new NPL formation every quarter, exhibiting banks proactive stance towards risk management Source: Companies data; BCA Sekuritas As a comparison to small banks, large banks have larger NPL coverage while credit cost strategy continues to stay prudent. This is sensible concerning their larger exposure to the whole economy, requiring them to build up a more solid coverage. Meanwhile, small banks asset quality had been hit ahead of large banks back in 214, and so they became less aggressive on provisioning given nearly done asset-quality recovery. Both large and small banks have registered escalating risky loan coverage QoQ and YoY in 2Q16, which is encouraging. 21

22 Going into 217F, we forecast large banks and small banks to generate lower credit cost by 49bp YoY and 3bp YoY, respectively, bringing NPL coverage to climb up to 169% and 88%. We are comfortable and deem their coverage level as sufficient. Exhibit 37. NPL coverage trend (large 4 banks) Risky loan coverage (LHS) Classified loan coverage (LHS) (%) (%) Credit cost (RHS) Q15 2Q15 3Q15 4Q15 1Q16 2Q16 Source: Companies data; BCA Sekuritas * Note: Large 4 banks covered include BMRI, BBNI, BBRI and BBCA Q16 classified coverage is stable QoQ Risky coverage is slightly decreasing QoQ on aggressive restructuring and NPL recognition led by BMRI in 2Q16 Exhibit 38. NPL coverage trend (small 4 banks) Risky loan coverage (LHS) Classified loan coverage (LHS) (%) (%) Credit cost (RHS) Q15 2Q15 3Q15 4Q15 1Q16 2Q16 Source: Companies data; BCA Sekuritas * Note: Small 4 banks covered include BDMN, BNGA, BBTN and BJBR Small banks have smaller coverage compared to large banks, despite of on the rise Large banks are at an advanced stage of asset-quality drop vs small banks which had already suffered during the first leg of the last tightening policy. Exhibit 39. Aggregate banks credit quality coverage (Jun-16) (%) Risky loan coverage Classified loan coverage NPL coverage BMRI BBRI BBNI BBCA BBTN BNGA BDMN BJBR In the following quarters we are estimating to encounter a higher risky loan coverage and that this would serve as a positive catalyst for stock prices Source: Companies data; BCA Sekuritas 22

23 Based on 2Q16 position, our calculation of how much additional NPL ratio banks could absorb before their earnings turn zero (or interrupting capital base) reveals that large banks are mostly in a better position to withstand any asset-quality deterioration risks. In this study we assume all of 2Q16 NPL turns sour, leaving banks to use the excess NPL buffers to absorb additional NPL ratio. We will examine how much banks could write off before their earnings turn negative (and in turn disrupting capital). The excess NPL buffers indicates the remaining buffers left from loans-loss reserves and PPOP as % of loans after writing off NPL. Not only may this study understate the banks ability to further absorb NPL shock, since we do not include the capital base, it may also understate the NPL shock magnitude stemming from SML slippage which is not incorporated in this study. Large banks could still write off additional bp of NPL ratio vs small banks of bp. This study rules out BBTN and BJBR which are unable to cover any NPL slippage without interrupting their capital or suffering a negative earnings. BDMN and BBRI stand out with the above-aggregate excess NPL buffers for its high-yield loan mix (retail). However, as the nature of retail loan, these two micro lenders are also faced with high SML ratio (2Q16: %) vs large non-micro lenders ( %). In the case of BBTN, it is less alarming since the bank is heavily exposed to high-quality collateral asset (2Q16 mortgage loan accounts for 26% of total loan) whose asset recovery process is fast. Moreover, its subsidized mortgage loan (2Q16: 33% of total loan) is covered by an insurance agency. As for BJBR, most of its loans are low risk (pensioners, civil servant), insured (KUR), and partly secured (4% of commercial loan book). Exhibit 4. The maximum additional NPL ratio banks could absorb before the entire earnings being wiped out (from Jun-16 position) (bp) Max additional NPL ratio banks could endure from excess buffers It is comforting to see that banks could still absorb additional NPL ratio by bp, thus validating our positive earnings visibility outlook BDMN BBRI BBCA BBNI BMRI BNGA BJBR BBTN Source: Companies data; BCA Sekuritas estimates 23

24 Could lower credit cost and better operating environment compensate for the NIM stress? Yes, the NIM stress next year could be compensated by our expectation of better credit cost consistency and improved operating environment ahead. Though earnings changes are most sensitive to NIM movement since it participates the bulk of banks total income (75% as of 2Q16), we believe below-the-line posts (i.e. opex, non-interest income, etc) could also cushion the NIM impact to bottom-line. Exhibit 41. Sensitivity analysis for aggregate banks EPS chg Loan gr ±1% CIR ±1% CoC ±1bp NIM ±1bp (%) 16F 17F 16F 17F 16F 17F 16F 17F BMRI BBRI BBNI BBCA BBTN BNGA BDMN BJBR Average Source: Companies data; BCA Sekuritas estimates * Note: CoC is credit cost; CIR is cost-to-income ratio For aggregate banks under coverage we calculate the net impact to earnings for the next two years (on average) on the following variables movement: We forecast NIM drops by 32bp for F vs the expected 22bp YoY fall this year to affect earnings by -17.1% for F. Cost-to-income ratio (CIR) is to slightly rise (on picking-up business activity and expansions while income is under pressure) to 48% for F vs the expected 46.6% this year, affecting earnings by 4.7% on average for the next two years. Loan growth would recover to % YoY in F vs the expected 11.6% YoY this year, supporting earnings on average by +5.9% for the next two years. Credit cost (CoC) is projected to narrow in the next two years to 142bp vs this year s peak of 212bp, helping earnings changes +25% for F. The average net impact for earnings from the above-mentioned variables movements are +9.2%. Exhibit 42. Net impact to earnings from key variables movement (%) Net impact to earnings on average F (5) (1) (15) (2) (4.7) (17.1) NIM CIR Loan growth CoC Net changes Source: Companies data; BCA Sekuritas estimates 24

25 Among the key variables, there exists likelihood of lower NIM (earnings downside risks) and lower credit cost (earnings upside risks) than we estimate. For the former our assumptions could be too aggressive on potentially rising regulatory risk, but possibly too conservative for the latter on speedier asset-quality recovery risk. Further, we may also understate the net impact to earnings since we do not incorporate other essential variables movement i.e. non-interest income and opex, which are below-the-line posts. 25

26 F 217F 218F 219F 22F 221F 222F 223F 224F 225F 226F 227F Tapering excess liquidity not an immediate issue yet In this section we would like to highlight a structural trend of declining excess liquidity in the banking system and how important liability spread is. We conclude that 1) liquidity is not yet an issue next year but would start to be visible in 218F and 2) liability spread matters more than asset spread. A fruitful tax amnesty will likely be a game changer for easing liquidity competition in the banking sector. One is pondering if liquidity tightness is an immediate issue for Indonesia banks in F? We do not think so as attributable to 1) ongoing gradual loan demand recovery (rather than super-strong loan demand) and 2) comfortable bank-only LDR level at 87% as of Jun-16 for our aggregate banks (vs System s 92%). We envisage liquidity issue to be more visible in 218F rather than now when the loan demand is projected to rebound a bit stronger, concurrently with picking-up business activity. Clearly this scenario could transpire either quicker or later were banks to grow their loan book more aggressively than their deposit base or were loan demand to pick up profoundly. We assign a low probability for the latter to materialize. As reference, loan-to-funding ratio (LFR) is becoming more prevalent nowadays as a new liquidity measure with banks frequently accessing to wholesale funding. As of Jun-16, system s LFR stood at 85% vs 82% a year earlier. Exhibit 43. LDR scenarios (%) Base-case 75bp in excess of loan vs deposit growths 1bp in excess of loan vs deposit growths 2bp in excess of loan vs deposit growths 227F, 1 % 22F, 1 % 224F, 1 % In the longer term we expect loan-growth-to-deposit-growth multiplier at 1x, and thus providing a longer duration for LDR reaching 1% F, 93 % 7 6 Source: Bank Indonesia; Companies data; BCA Sekuritas estimates One question frequently debated is what would happen when large banks have difficulty on retaining/growing deposit base, as usually their deposit rates are lower than smaller banks. We answer that high deposit rates offered by smaller banks may not sustain in the longer term as their NIM would be squeezed by more competitive lending rates and ongoing structurally lower interest rates. Hence, we take a view of unsustainably high deposit rates provided by the smaller banks to ultimately decline and that deposit flows will return to the large banks. In the long run we argue that large banks should be able to increase their domination in deposit share within the system. 26

27 Jan-93 May-95 Sep-97 Jan- May-2 Sep-4 Jan-7 May-9 Sep-11 Jan-14 May-16 In the following sub-sections we will also examine how important liability spread is given increasingly scarce liquidity in the long future as opposed to asset spread, which investors look like to be more focused on presently. We conclude that liability spread matters more than asset spread on driving NIM and loan growth given the following three reasons: 1. Declining excess liquidity (defined as surplus deposit in excess of loan) pointing to loan growth risks. 2. Tighter funding competition suggesting eventually loan growth risks. 3. NIM has largely been driven by liability spread more than asset spread in the past 15 years despite of lower loan yield. The driver is emanating from improved liability spread from rising CASA. All of the above-mentioned points accentuates the loan growth risks stemming from declining excess liquidity. This highlights the importance of banks ability to properly manage their liability spread. Shrinking excess liquidity to a new normal We conduct an analysis from banking system data points and find that excess liquidity as of Jun-16 has actually shrunk to 8.9%/8.2% of loan/deposit vs 11.9%/1.6% a year earlier and 24.3%/19.5% five years ago. This finding suggests to us that banks liquidity level is at a new normal (which is low), which we believe to persist for the next few years. Exhibit 44 demonstrates an interesting point; albeit in absolute terms excess deposit balance have been steady and averaging IDR446tn since the 1998 Asian Financial Crises/ AFC (Jun-16: IDR374bn), they have trended down consistently as % of balance sheets (loan and deposit). The excess deposit gap between loan and deposit have narrowed rapidly since 213 when banks started concentrating on managing liability spread (i.e. slowing down deposit gathering) in light of weak loan demand. This initiative resulted in decelerating deposit growth quite drastically. As a result of this falling excess deposit, we think the new normal of loan growth at mid double digits is going to linger in the foreseeable future. The past buoyant loan growth of +2% y-y is unlikely to repeat unless banks could grow their deposit base quite fast to sustain the loan expansion funding. Exhibit 44. Excess deposit balance sustained but dwindling as % of loan and deposit (%) Excess deposit as of loan (LHS) Excess deposit as of deposit (LHS) Excess deposit balance (RHS) (5) Narrowing (IDR tn) (1) The narrower gap in excess deposit between as % of loan and as % of deposit will likely be a game changer in the longer term Source: Bank Indonesia; BCA Sekuritas 27

28 Jan-1 Oct-1 Jul-2 Apr-3 Jan-4 Nov-4 Aug-5 May-6 Feb-7 Nov-7 Sep-8 Jun-9 Mar-1 Dec-1 Sep-11 Jul-12 Apr-13 Jan-14 Oct-14 Jul-15 May-16 The level of excess deposits narrows whenever additional loan expansion is higher than deposits gathered, implying the loan expansion sustainability is highly reliant on deposit growth momentum (Exhibit 45). This in turn suggests to us that loan growth momentum relies on each bank s deposit growth profile. Exhibit 45. implying loan-growth-to-deposit-growth gap should narrow to preserve excess deposit (IDR tn) Excess deposit Incremental loan YoY 59 Incremental deposit YoY but a narrower gap intimates 1) loan growth risk and 2) higher cost of funds (1) Jan-1 Dec-2 Nov-4 Oct-6 Sep-8 Aug-1 Jul-12 Jun-14 May-16 Source: Bank Indonesia; BCA Sekuritas Relative to historical context, current aggregate banks LDR of 87% in Jun- 16 (LFR: 8%) is tight but still at a comfortable level, in our opinion. The high-ldr issue should be more apparent in 218F and beyond as we anticipate a faster loan growth (vs deposit growth) to continue to keep up with improved economic growth momentum while excess liquidity would become more limited. As reference, existing system LDR is tight at 92% (though still off Dec-15 s peak at 92.7%) while existing loan-deposit growth multiplier is still high at 1.5x (above the 1.3x average). If this multiplier continues to be elevated, we anticipate this to result in LDR reaching near to 1% sooner rather than later (see earlier Exhibit 43). Exhibit 46. System data - when LDR nears all-time high, loan-deposit multiplier should narrow to ±1.1x (%) LDR (LHS) Loan-deposit growth multiplier (RHS) Multiplier average 1.3x 91.8 (x) (1) A narrower multiplier suggests that either loan needs to be slowed or deposit needs to accelerate, which either way does not augur well for NIM Source: Bank Indonesia; BCA Sekuritas 28

29 Jan-93 Jun-93 Nov-93 Apr-94 Sep-94 Feb-95 Jul-95 Dec-95 May-96 Oct-96 Mar-97 Aug-97 Jan-98 Jun-98 Nov-98 Apr-99 Sep-99 Feb- Jul- Dec- Looking ahead we estimate the loan-deposit growth multiplier to skew towards ±1.1x in the next few years (as in during pre-afc 1998), in turn hindering interest income growth. With a gradual 217F loan demand pickups and better-than-system liquidity, this is of no concern and not an alarming issue for our covered banks. Exhibit 47. similar to the past (pre-afc 1998) when LDR had been hovering just above 1% (%) LDR (LHS) Loan-deposit growth multiplier (RHS) (x) Multiplier average 1.5x 4 2 (2) (4) (6) (8) (1) (12) Source: Bank Indonesia; BCA Sekuritas Tighter funding competition than expected In this sub-section, we outline another case study which highlights the importance of banks management over the liability spread. Lending rates (to a lesser extent) and TD rates (to a greater extent) have decreased by 64bp and 99bp up to Jun-16, respectively, less than the BI rate fall by 1bp YTD. This highlights 1) the banks need to proactively manage NIM better (since room for further TD rates reversion is getting limited while lending rates will continue to contract) and 2) tighter liquidity than we envisage. Exhibit 48. Interest rates have fallen with TD (more to some extent) and WC (less) declining albeit elevated BI rate (%) BI rate Time deposit Working capital Jun-15, 12.7 Dec-15, Jun-15, 8.3 Dec-15, Apr-4 Aug-5 Dec-6 Apr-8 Aug-9 Dec-1 Apr-12 Aug-13 Dec-14 Apr-16 Source: Bank Indonesia; BCA Sekuritas A more detailed study over the loan/deposit interest rates spread to BI rate reveals stickier (or longer time) time deposit rate spread to return to the normal level. We describe the normal level as the lowest spread level before the peak of a tightening cycle. We apply working capital rate and TD rate as a representative of system s lending rate and deposit rate, respectively, since the lion s share of system loan is working capital (47% as of Jun-16) while TD rate is not as sticky as CASA rate. Our study is best pictured in Exhibit

30 The 28-9 tightening cycle lasted 45 months for lending rate spread to revert back to a normal level in Mar-11, since the bottom in Aug-7. Deposit rates spread, however, took longer by 68 months to normalize, highlighting that competition in funding was tighter than lending. In the last tightening cycle of lending rates spread has not normalized to 4.6% as the normal level with 77bp of gap left as of Jun-16. Interestingly, deposit rate spread still has higher gap of 134bp left to the normal level, implying funding competition is tight than in lending. There is a prospect of intensifying deposit competition, since we envisage history to repeat where deposit rate spread will require longer time to normalize. This long-term structural issue has been well recognized in our financial modeling (Exhibits 58-61). Exhibit 49. but the last two tightening episodes show longer time needed for deposit rate spread* to normalize (%) Lending rate spread Lending rate spread bottom (1) (2) Lending rate spread bottom, Aug Deposit rate spread bottom, Dec-7 Deposit rate spread (.6) Lending rate spread bottom, Mar-11 Lending rate spread bottom, Aug-13 Deposit rate spread bottom, Aug (.8) Mar-7 Sep-8 Mar-1 Sep-11 Mar-13 Sep-14 Mar-16 Source: Bank Indonesia; BCA Sekuritas * Spread to BI rate Deposit rate spread bottom Longer time needed for deposit rate spread to normalize is reflected in higher current gap left vs lending rate spread gap, suggesting NIM risk ahead Conclusion: our study of the past two tightening cycles concludes that the time needed for deposit rates spread to normalize is longer vs lending rate spread, implying actually a greater funding competition out there. This means that managing the liability spread is harder as deposit rate spreads are stickier than lending rate spread. For the longer term (218F and beyond) we believe this study intimates that liquidity will be of most valuable for Indonesia banks. Liability spread matters more than asset spread The earlier study proves that liability spread management is going to be essential going forward. In this sub-section we present two additional case studies that will corroborate our thesis of liability spread matters more than asset spread. The case studies are better visualized in Exhibits Case study 1: structurally higher NIM over the past 15 years on improved liability spread We examine over the last fifteen years of aggregate banks quarterly NIM trend during 3Q1-2Q16. Our NIM analysis indicates that banks NIM has actually expanded nicely to 6.4% as of June in spite of declining IEA yield. The major driver resides at cost of funds (CoF) improvement led by better CASA profile. This is an evidence of liability spread serving as a more important role on driving up NIM. We note banks with good deposit franchise to look better positioned and to stand out in the longer term. 3

31 4Q96 3Q97 2Q98 1Q99 4Q99 3Q 2Q1 1Q2 4Q2 3Q3 2Q4 1Q5 4Q5 3Q6 2Q7 1Q8 4Q8 3Q9 2Q1 1Q11 4Q11 3Q12 2Q13 1Q14 4Q14 3Q15 2Q16 4Q96 3Q97 2Q98 1Q99 4Q99 3Q 2Q1 1Q2 4Q2 3Q3 2Q4 1Q5 4Q5 3Q6 2Q7 1Q8 4Q8 3Q9 2Q1 1Q11 4Q11 3Q12 2Q13 1Q14 4Q14 3Q15 2Q16 3Q1 1Q2 3Q2 1Q3 3Q3 1Q4 3Q4 1Q5 3Q5 1Q6 3Q6 1Q7 3Q7 1Q8 3Q8 1Q9 3Q9 1Q1 3Q1 1Q11 3Q11 1Q12 3Q12 1Q13 3Q13 1Q14 3Q14 1Q15 3Q15 1Q16 Exhibit 5. Aggregate banks NIM rising vs the past 1.5 decade (%) NIM Linear (NIM) * Source: Companies data; BCA Sekuritas * Accounting changes (one-event) on micro-lending interest income recognition inflated NIM Exhibit 51. in spite of lower interest-earning asset yield (%) Loan/ interest-earning asset (LHS) (%) Interest-earning asset yield (RHS) Source: Companies data; BCA Sekuritas Exhibit 52. as cost of funds have improved significantly (%) LDR (LHS) CoF (RHS) (%) (12.7) (1.8) (3.5) (5) (1) (15) (2) (25) (3) Source: Companies data; BCA Sekuritas * CoF is denoted as a negative number 31

32 Exhibit 53. on the back of nice CASA development (%) CASA ratio Linear (CASA ratio) Jan-93 May-95 Sep-97 Jan- May-2 Sep-4 Jan-7 May-9 Sep-11 Jan-14 May-16 1Q11 2Q11 3Q11 4Q11 1Q12 2Q12 3Q12 4Q12 1Q13 2Q13 3Q13 4Q13 1Q14 2Q14 3Q14 4Q14 1Q15 2Q15 3Q15 4Q15 1Q11 2Q11 3Q11 4Q11 1Q12 2Q12 3Q12 4Q12 1Q13 2Q13 3Q13 4Q13 1Q14 2Q14 3Q14 4Q14 1Q15 2Q15 3Q15 4Q15 Source: Bank Indonesia; BCA Sekuritas Case study 2: higher NIM in the last tightening cycle on improved liability spread Our NIM analysis during the recent credit tightening cycle further validates our thesis that liability spread is more critical than asset spread. Average NIM was higher at 6.2% during the tightening period compared to 6% in the pre-tightening period, even when average IEA yield spread fell. The key driver resided at CoF improvement. Exhibit 54. Average NIM higher in the recent credit cycle (%) Aggregate NIM Avg. pre tightening (%) Avg. post tightening BI rate (RHS) Source: Bank Indonesia; Companies data; BCA Sekuritas * Spread to policy rate Exhibit 55. though average asset yield spread fell Aggregate IEA yield spread (LHS) Avg. pre tightening (LHS) (%) Avg. post tightening (LHS) (%) BI rate (RHS) Source: Bank Indonesia; Companies data; BCA Sekuritas * Spread to policy rate 32

33 1Q11 2Q11 3Q11 4Q11 1Q12 2Q12 3Q12 4Q12 1Q13 2Q13 3Q13 4Q13 1Q14 2Q14 3Q14 4Q14 1Q15 2Q15 3Q15 4Q15 Exhibit 56. as CoF spread improved on average Aggregate CoF spread (LHS) Avg. pre tightening (LHS) (%) Avg. post tightening (LHS) BI rate (RHS) (%). (.5) (1.) (1.5) (2.) (2.5) (3.) (3.5) (4.) (4.5) (2.6) (3.6) The case study 2 has a similar conclusion to case study 1, which is a positive liability spread movement drives NIM more Source: Bank Indonesia; Companies data; BCA Sekuritas * CoF is denoted as a negative number ** Spread to policy rate Best picks to play out the new liquidity level Going into the new liquidity level (new low of excess liquidity), and new normal loan growth (mid double-digit growth rate), we continue to recommend investors be back to the original formula: banks with highquality CASA ratio, low LDR, and firm deposit growth profiles should be relatively sheltered from the intense funding competition going forward. Exhibit 57. LDR vs CASA comparison (2Q16) (%) CASA ratio (LHS) Time deposit (LHS) LDR (RHS) BBCA BMRI BBRI BBNI BJBR BNGA BBTN BDMN Source: Companies data; BCA Sekuritas (%) Low LDR and high CASA: large banks dominate and will continue to do so in the foreseeable future 33

34 Improved earnings drivers = sector re-rating? We expect to see better operating environment in F to benefit Indonesia banks while we have yet to incorporate the tax amnesty benefits given difficulty to translate into our forecast quantitatively. Therefore successful tax amnesty could be a positive surprise. With overall better operating trend amid gradual macro recovery, a sector re-rating is highly possible to transpire. Better operating environment 1. Recovering net interest income growth NIM is banks key driver for top-line growth following 67-92% of total revenue is contributed by net interest income. We are of the view that banks structurally are entering a new normal of NIM level with stress risk likely intensifying beyond 216F, on account of limited room for more TDrate cuts and persisting low-rate environment. This is incorporated in our assumptions in Exhibit 58 which displays higher loan yield falls (and eventually asset yield) vs deposit cost (in turn cost of funding) in F. This asset-liability gap should narrow going forward. A contrast between large and small banks can be witnessed with the former expected to experience steeper NIM declines largely triggered by smaller room for further deposit cost cuts. Exhibit 58. Our assumptions on net interest income composition Yearly changes (bp YoY) F 217F 218F Aggregate banks (8 banks) NIM (29) (3) 2 (3) 7 (22) (21) (11) Avg interest asset yield (14) 39 (64) (38) (6) (5) (59) (37) (23) Avg cost of funding (35) 34 (13) (7) (62) (3) 83 (13) (45) (22) (15) Avg loan yield (12) (54) (95) (37) 47 (6) (79) (58) (36) Avg cost of deposit (18) 128 (27) (8) (73) (14) 11 (12) (6) (2) (15) Large banks (4 banks) NIM (19) (7) (19) (23) (13) Avg interest asset yield (1) 42 (47) (39) (59) (5) (51) (38) (22) Avg cost of funding (38) 42 (91) (2) (57) (6) 77 (12) (39) (19) (13) Avg loan yield (16) 55 (32) (83) (8) (14) 82 (4) (62) (67) (39) Avg cost of deposit (45) 53 (97) (22) (59) (1) 8 (19) (29) (17) (13) Small banks (4 banks) NIM (63) 39 7 (71) 8 (36) (52) 6 (35) (14) (4) Avg interest asset yield (12) 38 (136) (42) (69) (33) 47 4 (86) (35) (24) Avg cost of funding (43) 14 (152) 33 (86) (5) (68) (32) (25) Avg loan yield (8) (25) (11) (59) 11 (8) (97) (49) (33) Avg cost of deposit (86) (18) 14 (4) (92) (23) (18) Source: Companies data; BCA Sekuritas estimates While those with micro lending exposure, a 1% lending rate reduction p.a. in micro (both KUR and non-kur) are factored in before reaching the breakeven level by 222F. The breakeven level for micro lending rate before the operation is running at loss is at 13.3%, based on our calculation. As for single-digit rate, we do not expect the government to apply a single-digit rate to consumption loans without giving any incentives (i.e. funding support like in FLPP, interest subsidy in KUR and SSB, etc) given operational cost and risk profile differential. All loan segment being charged at single-digit rate next year is still unlikely, in our view. We assume most of SME lending rate to be at single-digit rate next year with other credit fee charges also raised (i.e. CASA placement with the bank, credit provisions, administration fee, etc) to compensate for the NIM loss. 34

35 Exhibit 59. and micro-lending yield already conservative (%) Overall micro Non-KUR KUR F 217F 218F 219F 22F 221F 222F 223F 224F 225F 226F 227F F 217F 218F Source: Companies data; BCA Sekuritas estimates After reflecting all of the NIM risks into our model, we expect net interest income growth to rebound after bottoming out this year, led 1) largely by rebounding loan volume, 2) to a less extent TD rate reversal, and 3) shifting asset mix/loan mix towards high-yield loan. To recall, earlier we discussed banks should be able to cope with the NIM pressures and that the net impact to bottom-line is manageable. Though F net interest income growth may look less exciting, however, these assumptions are built on very conservative inputs (highlighted in the two earlier exhibits), limiting downside risk to our forecast while leaving potential surprise to the upside. Exhibit 6. Net interest income growth set to rebound (% YoY) Aggregate net interest income growth Source: Companies data; BCA Sekuritas estimates 2. Higher fee-based income growth Amid structurally NIM declines, banks will need to explore other avenue to bolster revenue other than net interest income. In the long run, fee-based income is to be the key focus that would drive banks ROE/ROA. 217F non-interest income is projected to grow 9% YoY compared to 11.6% YoY this year, before rebounding to 12.2% YoY in 218F, to be driven by pick-ups in business activity. We believe a lower non-interest income growth rate next year is a function of this year s higher base: 1) FX gain (on volatile IDR in 1H16), 2) trading incomes (on policy rates fluctuations in 1H16), and 3) asset recoveries booking (on intensifying asset-quality deterioration). These trends should reverse in 217F given more stability in macro circumstances. It is worth pointing out that they are non-core items. 35

36 F 217F 218F Exhibit 61. Aggregate non-lending business trend (%) Non-interest income of revenue (LHS) (% YoY) Non-interest income growth (RHS) Fee-based income growth (RHS) (6) (1) 5 26 (18) (2) (3) Despite subdued 217F noninterest income growth on higher base this year from noncore items, core item fee-based income growth rebounded actually Source: Companies data; BCA Sekuritas estimates Delving deeper, the F core non-interest income (or fee-based income) is to grow better at 15% YoY (see Exhibit 61), implying improved business flows. Those with higher participation from fee-based income would see their bottom-line better sheltered from declines in NIM. Exhibit 62. Aggregate s non-lending business (ranked by 2Q16) Non-interest income of revenue (%) 214 2Q16 218F BMRI BBNI BDMN Aggregate BBCA BBRI BNGA BBTN BJBR Large banks stands out with BMRI leading, implying their readiness to better weather the long-term structural NIM issue Fee-based income of revenue (%) 214 2Q16 218F BDMN BMRI BBNI BBCA BNGA Aggregate BBRI BBTN BJBR Source: Companies data; BCA Sekuritas estimates BMRI and BDMN are the two-highest fee-based income generators for their various income sources. BMRI is solid in its insurance and securities businesses while BDMN with auto finance and insurance businesses. Looking into 218F we expect large banks to outperform small banks in fee-income generation for their wider business diversification, extensive networks, and great transactional banking franchise. Also, banks with larger capital base would be able to enlarge their business faster via inorganic growth. We expect more M&A in the next 12 months on the back of this setting. 3. Maintained cost efficiency amid up-cycle We are impressed with Indonesia banks superior ability on managing costto-income ratio (CIR) during various credit cycles. Multiples time banks have managed to keep a lid on cost growth, less than revenue growth, even during the challenging times. This cost control is also one crucial alternative (other than fee-based income) to offset slower top-line growth. 36

37 M16 216F 217F 218F Exhibit 63. Aggregate s cost growth vs revenue growth (%) Total cost growth YoY Total revenue growth YoY Source: Companies data; BCA Sekuritas estimates Cost trend has been consistently on the downtrend since 212 in-line with the GDP downtrend, and thus the trend should work on the other way around when GDP expands. We attribute this to partly deliberate initiatives taken by the banks to support bottom-line by minimizing costs. In the next few quarters, cost growth would not continue to be soft since business activity is anticipated to pick up. Banks capex should escalate to catch up the uptrend. As such, we envisage cost growth to outpace revenue growth, though still at a manageable rate. We tend to look at cost-to-asset ratio as a cost efficiency measurement as cost-to-income ratio is distorted by falling NIM as its denominator. Banks will still be sustaining cost-to-asset ratio at % going forward, which we deem as efficient. One positive catalyst we have yet to factor in is the planned consolidation of networks (i.e. ATM, EDC, etc) within state-owned banks. No clear details as of now, but the pilot project is ongoing presently. The chairman of Himbara (State-owned Banks Association), Mr. Gatot Suwondo former BBNI CEO, stipulated that banks would be able to save more costs by as much as 3%. Were this attempt to be carried out successfully, there would be upside risks to our earnings forecast (see our prior sensitivity table). BMRI, BBRI, BBNI, and BBTN are the state banks which are likely to reap the benefits and among these banks, BBTN should relish the most for its least networks and infrastructures. Exhibit 64. Cost efficiency outlook (aggregate banks) Cost efficiency (%) Q16* 216F 217F 218F Cost-to-income ratio Aggregate Large Small Cost-to-average asset Aggregate Large Small Source: Companies data; BCA Sekuritas estimates * 2Q16 cost-to-average asset surged on seasonal factor from several banks (i.e. pension benefits, retirement benefits, and bonus) Note: Large 4 includes BMRI, BBRI, BBNI, and BBCA; Small 4 includes BDMN, BNGA, BBTN, BJBR; Aggregate includes all banks under coverage 37

38 4. Positive credit cost outlook With 1) ample excess NPL buffers being able to absorb up to additional 493bp without disrupting capital base, 2) Tier-1 CAR at 18.2%, and 3) better macro stability; we are upbeat of positive credit cost outlook to help drive the banks earnings. The 216F is to be the worst in terms of both credit cost and NPL ratio, in our estimates. Were the asset-quality deterioration to continue, we believe banks would still have ample NPL buffers to better weather the NPL volatility. Exhibit 65. Credit cost outlook (aggregate banks) Credit cost (%) Q16 2Q16 216F 217F 218F BMRI BBRI BBNI BBCA (.1) Large BBTN BNGA BDMN BJBR Small Aggregate Source: Companies data; BCA Sekuritas estimates Receding credit cost signifies the end of NPL cycle is nigh. We argue earlier that the peak has passed and therefore, NPL ratio should moderate. Though not considerably recovering, this suffices to allow banks to better measure their credit cost anticipation. With abating NPL ratio, banks NPL coverage on average should also expand. Exhibit 66. NPL ratio outlook (aggregate banks) NPL ratio (%) Q16 2Q16 216F 217F 218F BMRI BBRI BBNI BBCA Large BBTN BNGA BDMN BJBR Small Aggregate Source: Companies data; BCA Sekuritas estimates leading to upgraded earnings trajectory On an expected better operating backdrop, Indonesia banks F earnings growth is forecast to turn positive to double-digit growth rate of 14.4%/14.3%% YoY in F vs a negative and a flattish growth rates in 216F and 215, respectively. Key drivers for improved earnings growth include higher loan volume, lower credit cost, and manageable opex growth. 38

39 1Q5 4Q5 3Q6 2Q7 1Q8 4Q8 3Q9 2Q1 1Q11 4Q11 3Q12 2Q13 1Q14 4Q14 3Q15 2Q F 217F 218F F 217F 218F Exhibit 67. Aggregate s PPOP and earnings would rebound (% YoY) PPOP growth Earnings growth Improved earnings visibility (1) (2) Source: Companies data; BCA Sekuritas estimates We anticipate ROE to stabilize for the first time since 21 s consistent drops. The normalization should also be responded by normalizing P/B or a sector re-rating, in our view. The positive future ROE trend is likely to be valuation supportive for banks owing to P/B s strong correlation with ROE trajectory. Exhibit 68. followed by stabilizing ROE trend at mid-teens (%) ROE would translate into stabilizing ROE, and Source: Companies data; BCA Sekuritas estimates Exhibit 69. While banks P/B and ROE appear highly correlated (x) PBV (LHS) ROE (RHS) (%) subsequently re-rate the banking sector given normalizing ROE, which is strongly correlated with P/B.5 Source: Bloomberg; Companies data; BCA Sekuritas 39

40 3-Dec Jan Jan-16 1-Feb Feb-16 9-Mar Mar-16 6-Apr-16 2-Apr-16 4-May May-16 1-Jun Jun Jun Jul Jul-16 1-Aug Aug F 217F 218F F 217F 218F F 217F 218F F 217F 218F F 217F 218F F 217F 218F F 217F 218F F 217F 218F Exhibit 7. some banks should see a re-rating on rising ROE (%) ROE BMRI BBRI BBNI BBCA BBTN BNGA BDMN BJBR Source: Companies data; BCA Sekuritas estimates We compare our earnings projection with the Street on Indonesia banks, bringing our F aggregate earnings at pretty much in-line with the Street. While some discrepancies witnessed in some banks (up to 11%) are owing to higher credit cost, reflecting quite conservative banks target prices we have. Exhibit 71. BCA Sekuritas vs Street estimates Earnings BCAS (IDR bn) Street (IDR bn) BCAS/Street (%) 216F 217F 218F 216F 217F 218F 216F 217F 218F BMRI 14,51 17,976 22,2 14,371 18,342 21, BBRI 24,948 27,951 3,865 24,919 27,693 32, BBNI 1,46 11,413 13,9 1,223 12,664 15, BBCA 19,718 21,42 22,947 2,131 21,957 24, BBTN 2,249 2,658 3,52 2,413 2,964 3, BNGA 1,485 1,916 2,32 1,51 1,982 2, BDMN 2,883 3,532 4,26 2,215 2,546 2, BJBR 1,61 1,73 1,92 1,661 1,841 2, Total 77,441 88,578 11,266 77,442 89,988 14, Source: Bloomberg for Street forecast; BCA Sekuritas estimates Meanwhile, the negative earnings revisions for F aggregate banks continue with the recent downgrades by 2-6% post 2Q16 results, led by BMRI. Generally we think there are limited downside risks for further 216F earnings downgrades following tepid 2Q16. Exhibit F earnings revision by the Street (IDR tn) 216F aggregate earnings forecast Source: Bloomberg estimates; BCA Sekuritas Note: BNGA (limited analyst coverage) is excluded 4

41 3-Dec Jan Jan-16 1-Feb Feb-16 9-Mar Mar-16 6-Apr-16 2-Apr-16 4-May May-16 1-Jun Jun Jun Jul Jul-16 1-Aug Aug-16 Exhibit F earnings revision by the Street (IDR tn) 217F aggregate earnings forecast Source: Bloomberg estimates; BCA Sekuritas Note: BNGA (limited analyst coverage) is excluded 41

42 Investment idea and risks Prefer large banks over small banks in the longer term We like banks which will be able to sail through the expected cycles ahead. 1. To weather the NIM cycles: banks with 1) low loan participation as % of IEA, 2) retail loan mix focus, and 3) more room to reduce TD rates look better positioned. 2. To navigate the liquidity cycles: banks with 1) low LDR and 2) high CASA ratio will be better off. 3. From bottom-line cushion perspective: those with 1) higher noninterest income mix and 2) lower credit cost outlook. We have a BUY rating on BMRI, BBRI, BBNI, BBTN, and BNGA (by order of preference from the most favorite) since we believe their earnings recovery is not yet fully reflected in the share price and these banks have one of the best metrics to address the longer term structural issues. Among the large banks: BMRI is our top pick for its 1) expected strong recovery on being most leveraged to an upturn in the economy, 2) superior deposit franchise, 3) high-yield loan focus, and 4) high-sector non-interest income mix. Within the small banks: BBTN is our favorite owing to 1) structurally strong loan demand prospect, 2) understated funding strength, and 3) largely recovered asset quality. Over the longer term large banks (vs small banks) remains a dominant player for them having stronger capital base, higher non-interest income mix, wider networks, better diversified business, and better deposit franchise. Exhibit 74. Our base-case structural view over Indo banks outlook 216F 217F Beyond NIM stress phases in NIM stress intensifies NIM stress normalizes towards a new level Ample liquidity Liquidity headwinds phase in Liquidity headwinds intensify NPL volatility intensifies NPL volatility stabilizes NPL recovers significantly Adjustment cycles towards a 'new normal' Source: BCA Sekuritas estimates Exhibit 75. Indonesia banks valuation table Ticker Rating TP CP Mkt cap EPS gr. (%) P/E (x) P/B (x) ROE (%) Yield (%) (IDR) (IDR) (USDm) 16F 17F 17F 18F 17F 18F 16F 17F 16F 17F BBCA NR na 15,175 28, BBRI BUY 14,2 12, 22,539 (1.8) BMRI BUY 13,45 11,375 2,28 (28.6) BBNI BUY 6,4 5,475 7, BDMN HOLD 4, 3,81 2, BBTN BUY 2,4 1,99 1, BNGA BUY 1, 815 1, BJBR HOLD 1,575 1,675 1, Sector Overweight 85,853 (1.8) Source: Companies data; Bloomberg; BCA Sekuritas estimates * Closing prices as of 21 September 216 ** BBCA is not rated 42

43 Small banks under coverage have all outperformed JCI and JAKFIN indexes by 5-19% YTD. Small banks led by BJBR outperforming JCI the most and Danamon the least. We attribute small banks outperformance over large banks to quicker asset-quality recovery (as them having been hit by NPL cycle ahead of large banks) and NIM outperformance potential during declining interest rates. As we also anticipate asset-quality for large banks to largely recover next year, we are convinced that large banks stock prices will catch up. Exhibit 76. Stock price performance YTD, 3M, and 6M (%) YTD 6M 3M (1) (3) (7) BJBR BBTN BNGA BDMN BMRI JCI JAKFIN BBCA BBNI BBRI Source: Bloomberg; BCA Sekuritas * Closing prices as of 21 September 216 Catalysts to our OVERWEIGHT call: Resilient NIM trend. Benign credit cost. Strong loan volume. Decelerating NPL formation. Sustained cost-efficiency. Encouraging tax amnesty receipts progress. Better economic data points. Positive catalysts that we have not factored in? Successful tax amnesty realization. A speedier and significant asset-quality recovery potentials. A likelihood of a shifting deposit rate cap reference to 7-days reverse repo (currently BI rate). Risks to our call The followings are risks that could impede the share price to reach our TPs: Higher decline in NIM than we forecast. Worsening asset-quality deterioration with accelerating restructured loan formation. Higher credit cost and opex growth than we expect. Slower loan growth than forecast. 43

44 Intensifying deposit competition than envisaged. Increased regulatory forces that cap lending rate for all loan segments next year. Sharp policy rate hikes. Declining consumer price index. Global tightening. Unsuccessful tax amnesty realization. Exhibit 77. Indonesia banks ROE decomposition analysis (yearly): sustained non-interest income, improved cost efficiency, and lower credit cost key drivers for F Of avg. assets (%) F 217F 218F Net interest income Non-interest income Income yield Cost Core profit/ PPOP Credit cost Non-op. income yield Pre-tax ROA Tax rate Post-tax ROA Minorities Leverage (x) ROE Source: Companies data; BCA Sekuritas estimates Exhibit 78. Indonesia banks ROE decomposition analysis (quarterly): weakening cost control, upward trend in credit cost, and lower leverage on asset revaluation drove banks ROE lower in the past two quarters Of avg. assets (%) 3Q13 4Q13 1Q14 2Q14 3Q14 4Q14 1Q15 2Q15 3Q15 4Q15 1Q16 2Q16 Net interest income Non-interest income Income yield Cost Core profit/ PPOP Credit cost Non-op. income yield (.).. Pre-tax ROA Tax rate Post-tax ROA Minorities Leverage (x) ROE Source: Companies data; BCA Sekuritas 44

45 Industry overview summary In the short term (2H16F): macro situation has improved with stability in IDR, CAD, CPI, and commodity prices as well as loosened monetary policy helping drive up GDP. Banks as the main economic driver would enjoy the potentially up-cycle. The immediate positives from better economy, however, are yet to be felt meaningfully this year. We forecast a decent banks performance in 216F with negative earnings growth of 1.8% YoY on the back of peaking credit cost (at 212bp), NIM headwinds phasing in (-22bp YoY), and a weak loan demand (+11.6% YoY). NPL growth is likely to have peaked already in 2Q16 (assuming no kitchensinking in 2H16F) with net interest income growth bottoming out at end- 216F. Accordingly, ROE is to drop to 15.3% from a year earlier 17.7%. In the medium term (217F): on operating situation improvement coupled with continuous gradual recovery in macro, banks earnings growth is to rebound to 14.4% YoY, spurred by lower credit cost, rising fee-based income, and sustained cost-efficiency. We expect NIM headwinds to intensify and liquidity issue to phase in. NIM for aggregate banks are estimated to come off by 21bp for 217F. Yet, there are ways to mitigate NIM pressure with one of them via non-interest income focus. Net-net, net interest income growth is rebounding in 217F from 216F s bottom. Meanwhile, NPL cycles are projected to stabilize with positive credit cost outlook helping drive earnings. Both loan growth and fee-based income will likely rebound on firmer macro circumstances. Inorganic growth via M&A shall mount in light of banks exploring way in a bid to diversifying revenue to mitigate tapering top-line growth. Ultimately ROE is expected to normalize at mid-teens from consistent contractions since 21. Longer term (beyond 217F): by 218F, we see legacy of asset-quality deterioration from the last tightening periods to mostly recover. Liquidity issue should escalate as excess liquidity is entering a new normal level, which is the new low. Loan-growth-to-deposit-growth gap would narrow unless banks could sustain strong deposit growth. Large banks will likely survive for their stronger balance sheet (mainly on greater deposit franchise) and to expand their dominance over smaller banks. Deposit franchise will be key to outperform by maintaining solid liabilities spread as this matters more than asset spread. Consolidation in the banking system is envisaged to accelerate on the back of liquidity issue and this is to be most visible with the smaller banks as the targeted banks. 45

46 Jun-8 May-9 Apr-1 Mar-11 Feb-12 Jan-13 Dec-13 Nov-14 Oct-15 Sep-16 Jun-8 May-9 Apr-1 Mar-11 Feb-12 Jan-13 Dec-13 Nov-14 Oct-15 Sep-16 Jun-8 May-9 Apr-1 Mar-11 Feb-12 Jan-13 Dec-13 Nov-14 Oct-15 Sep-16 Valuation charts and forecasts tables Charts Exhibit 79. Indonesia banks 12M forward P/B (x) Std +2 = 2.5x Std +1 = 2.3x Mean = 2.1x Std -1 = 1.8x Std -2 = 1.6x De-rating (Jan-15 until 2Q16) transpired during increased macro volatility however, the trend has reversed with current situation pointing to a more stable macro economy, eventually leading to normalizing banks ROE and in turn potentially a sector re-rating! Source: Bloomberg; Companies data; BCA Sekuritas estimates Exhibit 8. Indonesia large-four 12M forward P/B (x) Std +2 = 2.8x Std +1 = 2.6x Mean = 2.3x Std -1 = 2x Std -2 = 1.8x Still running below the historical mean: higher P/B multiple on large banks vs small banks marks the market expectation over large banks stronger position to navigate the long-term structural trend Source: Bloomberg; Companies data (BMRI, BBRI, BBNI, and BBCA); BCA Sekuritas estimates Exhibit 81. Indonesia small-four banks 12M forward P/B (x) Std +2 = 2.2x Std +1 = 1.8x Mean = 1.4x Std -1 = 1.1x Std -2 =.7x offering selective small banks a compelling valuation to accumulate over the next 12M Source: Bloomberg; Companies data (BDMN, BNGA, BBTN, and BJBR); BCA Sekuritas estimates 46

47 Jun-8 May-9 Apr-1 Mar-11 Feb-12 Jan-13 Dec-13 Nov-14 Oct-15 Sep-16 Jun-8 May-9 Apr-1 Mar-11 Feb-12 Jan-13 Dec-13 Nov-14 Oct-15 Sep-16 Jun-8 May-9 Apr-1 Mar-11 Feb-12 Jan-13 Dec-13 Nov-14 Oct-15 Sep-16 Exhibit 82. Indonesia banks 12M forward P/E (x) Std +2 = 14.3x Std +1 = 12.9x Mean = 11.6x Std -1 = 1.3x Std -2 = 8.9x Current seemingly lofty P/E is driven by soft earnings growth projection, though, still lower than the earlier peak in 2Q15 with healthier earnings visibility outlook, P/E trend should be firmer ahead Source: Bloomberg; Companies data; BCA Sekuritas estimates Exhibit 83. Indonesia large-four banks 12M forward P/E (x) Std +2 = 14.3x Std +1 = 13x Mean = 11.6x Std -1 = 1.3x Std -2 = 8.9x Still lower than the earlier top, while current circumstance is firmer, a P/E catching up to the last multiple peak appears to be likely 5 Source: Bloomberg; Companies data (BMRI, BBRI, BBNI, and BBCA); BCA Sekuritas estimates Exhibit 84. Indonesia small-four banks 12M forward P/E (x) Std +2 = 15.1x Std +1 = 13.2x Mean = 11.3x Std -1 = 9.5x Std -2 = 7.6x Small banks are trading at submean which looks to be interesting to collect 5 Source: Bloomberg; Companies data (BDMN, BNGA, BBTN, and BJBR); BCA Sekuritas estimates 47

48 Forecasts tables Exhibit 85. Indonesia banks metrics (1/2) BBCA BBRI BMRI BBNI BDMN BBTN BNGA BJBR Sector Large 4 Small 4 Rating NR BUY BUY BUY HOLD BUY BUY HOLD Overweight Target price (IDR) na 14, ,4 4, 2,4 1, 1,575 Share price (IDR) 15,175 12, 11,375 5,475 3,81 1, ,675 Up(down)side (%) na (6) Total mkt cap. (USDm) 28,152 22,539 2,28 7,774 2,78 1,63 1,559 1,237 85,853 78,673 7,18 PPOP (IDR tn) E F F Net profit (IDR tn) F F F PPOP growth (1.1) (8.7) (9.3) (11.1) (79.6) (%) (5.6) (5.9) F (1.8) 15.7 (8.) F F Net profit (35.6) (26.7) (45.3) (19.8) (36.1) growth (%) (15.9) (8.2) 61.6 (81.7) (16.) 216F 9.4 (1.8) (28.6) (1.8) (4.9) F F Loan growth (%) (7.) F F F Deposit growth (%) (1.2) 2. (1.8) F F F LDR (%) F F F NIM (%) F F F Net fee income ,975 (15.3) (1.6) growth (%) (6.1) F (5.) (1.2) 217F F Non-int. income (15.7) 15.5 (21.1) (13.5) growth (%) (8.5) (3.1) 22.7 (4.2) (.3) 216F F 13.6 (.8) F Non-int. income of tot inc. (%) F F F Source: Bloomberg; Companies data; BCA Sekuritas estimates * No TP for BBCA as it is not rated ** Closing prices as of 21 September 16 48

49 Exhibit 86. Indonesia banks metrics (2/2) BBCA BBRI BMRI BBNI BDMN BBTN BNGA BJBR Sector Large 4 Small 4 Rating NR BUY BUY BUY HOLD BUY BUY HOLD Overweight Target price (IDR) na 14,2 13,45 6,4 4, 2,4 1, 1,575 Share price (IDR) 15,175 12, 11,375 5,475 3,81 1, ,675 Up(down)side (%) na (6) Total mkt cap. (USDm) 28,152 22,539 2,28 7,774 2,78 1,63 1,559 1,237 85,853 78,673 7,18 Cost to income (%) F F F Credit cost (bp) F F F NPL ratio (%) F F F NPL coverage (%) F F F P/E 216F (x) 217F F P/B 216F (x) 217F F Dividend yield 216F (%) 217F F ROA (%) F F F ROE (%) F F F CAR (%) F F F Tier (%) F F F Earnings CAGR 15-17F (6.) (%) 15-18F F BCAS/Street 216F Earnings (%) 217F F Source: Bloomberg; Companies data; BCA Sekuritas estimates * No TP for BBCA as it is not rated ** Closing prices as of 21 September 16 49

50 BANK MANDIRI BMRI IJ / BMRI.JK Company Reinitiation Financials Sector BUY Previous: - 22 Sep 216 A catch-up play Most leveraged to GDP upturn Given BMRI s exposure on its most developed and diversified business units (insurance, investment banking, and multi-finance), we are upbeat it looks poised to recover quite rapidly following our anticipation of a firmer economic growth. We project the bank s ROE to rise to 13.5%/14.9% in 217/18F vs 216F s 11.9% bottom on above-aggregate earnings growth at 23.9%/22.4% YoY. Loan growth and credit cost improvements are the key driver. Key concern of asset-quality deterioration to abate A better impending macro setting led by lower interest rates, stable IDR, and benign CPI would lead to a quicker NPL recovery. The front-loaded 3.6% credit cost in 2Q16 (1Q16: 3.1%) on aggressive asset-quality review and a better management guidance of IDR4tn loan recovery this year (vs 215: IDR3tn) should indicate an early resolution in NPL cycle. While 4.3% loans-loss reserves was the highest quarter since 1Q11, both classified/risky loans growth YoY have decelerated in 2Q16. Our credit cost assumptions are conservative with 216F/17F/18F at 2.9%/2.3%/2% vs 3.6%/2.1% in 2Q16/215. Best play to address the long-term structural issue A structural issue of tapering interest income and depleting excess liquidity are stressed in the accompanied sector report. Both factors are likely to drive NIM to a new low level. Yet, we think BMRI is the best play to address this issue since it has the best required metrics among our banks coverage to reduce the adverse impact from softening top-line. By 2Q16 it has the highest 33% non-interest income mix, lowest 67% loan to earnings-asset mix, and 3% high-yield retail loan mix (targeting a 4% mix by 22F as per the bank objective). Longer term, the bank s great deposit franchise of low-aggregate 88% LDR (vs 91% aggregate) and high-aggregate 64% CASA ratio (6%) in 2Q16 bodes well to better weather the liquidity issue. Reinitiating with GGM-based TP IDR13,45, BUY; Our top pick We assume 13.3% WACC (risk-free rate of 6.8%), 15.5% ROE, and 11% growth rate to reach our GGM-based TP. This implies a 2.1x P/B target based on 217F (between mean and +1 STD). Presently at 1.9x (mean), it is still off the earlier peaks of 2.4x (Apr-15) and 2.7x (Apr-13). Our target P/B is lower vs the last valuation peak since we are entering new normal periods. BMRI is best geared to capture the economic uptrend by registering ROE uptrend via strong earnings rebound. Downside risks include worse-than-expected trends in NPL and NIM falls, and slower loan growth than estimates. PRICE PERFORMANCE STOCK PERFORMANCE Igor Nyoman Putra igor.putra@bcasekuritas.co.id ext 178 Target Price: IDR13,45/share Current Price: IDR11,375/share Upside potential: 18.2% 52-week price range (IDR) : 7,15-11,95 Market cap (IDRtn)/(USDbn) : / 2.21 Avg Daily Turnover (IDR/USD) : / Source: Bloomberg MARKET DATA Ytd 1M 3M 12M Absolute 23.% -.7% 26.4% 32.3% JCI Return 16.3% -1.4% 9.5% 22.1% Relative 6.7%.7% 16.9% 1.2% Source: Bloomberg Exhibit 1. Financials and valuations Year-end 31 Dec (IDR bn) F 217F 218F Net interest income 39,132 45,363 46,875 48,846 53,246 Pre-prov. op. profit 31,648 38,36 37,671 4,67 44,983 Net profit 19,872 2,335 14,51 17,976 22,2 EPS (IDR) EPS growth (%) (28.6) P/E (x) BVPS (IDR) 4,4 5,17 5,453 5,992 6,652 P/B (x) DPS (IDR) Yield (%) ROE (%) SHAREHOLDERS Government : 6.% Others :.% Public : 4.% Source: Bloomberg Source: Bloomberg; Bank Mandiri, BCA Sekuritas estimates

51 Exhibit 2. Bank Mandiri - Financials Summary Income Statement Year-end 31 Dec (IDRbn) F 217F 218F Interest income 62,638 71,57 7,984 73,32 79,375 Interest expense (23,56) (26,27) (24,19) (24,474) (26,129) Net interest income 39,132 45,363 46,875 48,846 53,246 Non-interest income 17,749 21,791 22,299 25,313 28,789 Total operating income 56,882 67,154 69,174 74,159 82,35 Operating expenses (25,233) (28,795) (31,53) (34,91) (37,52) Pre-prov. op. profit 31,648 38,36 37,671 4,67 44,983 Provisions expense (5,67) (12,21) (18,166) (16,61) (15,755) Operating profit 25,978 26,339 19,55 24,7 29,228 Non-op. inc./(exp.) Pre-tax profit 26,8 26,369 19,55 24,7 29,228 Corporate tax (5,353) (5,217) (4,96) (5,41) (6,138) Minorities (783) (817) (899) (989) (1,88) Net profit 19,872 2,335 14,51 17,976 22,2 EPS (IDR) Balance Sheet Year-end 31 Dec (IDRbn) F 217F 218F Placement with other banks 27,463 2,45 22,446 24,69 27,159 Loans 523,12 586, , , ,548 Government bonds 86,154 13,869 14,98 15,957 17,16 Other int-earning assets 122,99 91,45 11, , ,532 Financial assets impairment (2,58) (24,783) (34,979) (43,482) (51,512) Net earning assets 739, ,617 84, ,621 1,27,744 Cash 71,33 81,423 78,2 87,49 97,845 Fixed assets 8,929 9,762 1,25 1,762 11,3 Other assets 35,237 41,261 42,35 62,57 71,58 Total assets 855,4 91,63 971,336 1,87,38 1,28,469 Customer deposits 636, , ,26 84, ,639 Deposits from other banks 17,772 12,952 14,247 15,672 17,239 Borrowing and sub-debts 36,173 44,993 49,492 54,441 57,34 Other liabilities 59,868 56,24 33,899 34,238 34,58 Total liabilities 75,195 79, , ,1 1,5,762 Minorities 2,187 2,422 2,446 2,47 2,495 Equity 12, ,7 127, ,81 155,212 Total liabilities and equity 855,4 91,63 971,336 1,87,38 1,28,469 BVPS (IDR) 4,4 5,17 5,453 5,992 6,652 Breakdown Year-end 31 Dec (%) F 217F 218F As % of earning assets Gross loan Government bonds Other earnings assets As % of loan book Corporate SME/Commercial Consumer Micro Others As % of deposit Core deposit Time deposit Ratios & assumptions Year-end 31 Dec (%) F 217F 218F NIM Loan-to-deposit ratio Cost-to-income ratio Credit cost Gross NPL NPL coverage Tier-1 CAR ROA ROE Average asset yield Average funding cost Average loan yield Average deposit cost Loan growth Deposit growth EPS growth (28.6) Source: Bank Mandiri, BCA Sekuritas estimates Net interest income growth of 3%/4%/9% for 216F/17F/18F is recovering driven by improved loan growth We assume a similar focus with the bank concerning retail loan mix target Declining credit cost is in-line with our anticipation of lower gross NPL ratio 51

52 2Q12 3Q12 4Q12 1Q13 2Q13 3Q13 4Q13 1Q14 2Q14 3Q14 4Q14 1Q15 2Q15 3Q15 4Q15 1Q16 2Q F 217F 218F F 217F 218F F 217F 218F F 217F 218F F 217F 218F F 217F 218F F 217F 218F F 217F 218F Exhibit 3. ROE and earnings growth comparison within aggregate (%) ROE (LHS) Earnings growth (RHS) (% YoY) (5) (15) (25) (35) BMRI sees rising ROE and solid earnings growth for F BMRI BBRI BBNI BBCA BBTN BNGA BDMN BJBR Source: Companies data; BCA Sekuritas estimates Note: we omit BNGA s earnings growth since it is still in a restructuring phase, resulting in highly volatile trend Exhibit 4. Risky loan off the peak in 1Q16 (% YoY) Classified loan growth (LHS) Risky loan growth (LHS) (%) 6 Classified loan ratio (RHS) Risky loan ratio (RHS) Stabilizing 4 2 Both classified and risky loan growth have stabilized in the recent two quarters, bringing both ratios stable QoQ, despite of aggressive review by the bank in 2Q16 NPL (1) Source: Bank Mandiri; BCA Sekuritas Note: classified loan (NPL + SML); Risky loan (classified loan + current restructured loan) Exhibit 5. The highest/lowest non-interest income/loan of IEA (2Q16) (%) Non-interest income Loan of interest-earnings asset BMRI is leading with these two metrics for mitigating future NIM stress BMRI BBRI BBNI BBCA BDMN BNGA BBTN BJBR Source: Companies data; BCA Sekuritas 52

53 Jan-1 Jul-1 Jan-11 Jul-11 Jan-12 Jul-12 Jan-13 Jul-13 Jan-14 Jul-14 Jan-15 Jul-15 Jan-16 Jul-16 Jun-6 Sep-7 Dec-8 Mar-1 Jun-11 Sep-12 Dec-13 Mar-15 Jun-16 Exhibit 6. Focusing on retail loan (micro, business banking, consumer) (%) Micro Business banking Consumer Corporate Commercial Q16 22F 35 Source: Bank Mandiri estimates, BCA Sekuritas Exhibit 7. 12M forward P/B (x) 3. Attractive valuation, at mean and still off the previous peaks Std +2 = 2.4x Std +1 = 2.2x Mean = 1.9x Std -1 = 1.7x Std -2 = 1.5x 1. Source: Bloomberg; Bank Mandiri; BCA Sekuritas estimates Exhibit 8. 12M forward P/E (x) Std +2 = 16.3x Std +1 = 14.5x Mean = 12.6x Std -1 = 1.8x Std -2 = 8.9x In-line with peers trend on normalizing EPS growth amid current soft operating environment 6 Source: Bloomberg; Bank Mandiri; BCA Sekuritas estimates 53

54 Exhibit 9. ROE decomposition analysis (yearly) Rising F ROE driven by lower credit cost while noninterest income and cost efficiency trend are improving slightly Of avg. assets (%) F 217F 218F Net interest income Non-interest income Income yield Cost Core profit/ PPOP Credit cost Non-op. income yield Pre-tax ROA Tax rate Post-tax ROA Minorities Leverage (x) ROE Source: Bank Mandiri; BCA Sekuritas estimates Exhibit 1. ROE decomposition analysis (quarterly) Credit cost has brought down ROE consistently since 4Q14 Of avg. assets (%) 3Q13 4Q13 1Q14 2Q14 3Q14 4Q14 1Q15 2Q15 3Q15 4Q15 1Q16 2Q16 Net interest income Non-interest income Income yield Cost Core profit/ PPOP Credit cost Non-op. income yield..2 (.).... (.).. (.) (.) Pre-tax ROA Tax rate Post-tax ROA Minorities Leverage (x) ROE Source: Bank Mandiri; BCA Sekuritas 54

55 BANK RAKYAT INDONESIA BBRI IJ / BBRI.JK Company Reinitiation Financials Sector BUY Previous: - 22 Sep 216 Sustaining value-accretive growth Pursuing value-accretive micro loan growth Growth focus in government-backed credit to unbanked people (KUR) is projected to be value-accretive with F ROA/ROE on average at 4.2%/26% vs the group level at 2.6%/19%. Albeit there is an expected micro cannibalization from KUR, higher loan mix in high-yield KUR would prop up the overall blended loan yield and in turn, mitigate NIM stress. We expect micro loan to participate at 35% of total loan by 218F from 2Q16/215 33%/31%. Least risky loan segments exposure Since most of the credit-quality deteriorations in the banking system have emanated from SME and corporate segments, BBRI looks better positioned. It commands the least risky loan mix of 36% (non-soe and SME) while owning the highest excess NPL buffers (PPOP/loan + loans-loss reserves - NPL) of 4.8% excluding 18% tier-1 CAR in 2Q16. Risky loan ratio trend (classified + current restructured loan) is resilient at +11bp QoQ (+2bp YoY) vs large peers trend at 164bp QoQ (+7bp YoY) in 2Q16. Better off during low-rate milieu BBRI s largest portion of non-core funding (45.3% in 2Q16) vs peers, and fixed-rate loan dominance of c.5% of loan (micro and retail) should cushion NIM in the medium term from abating interest rate. Cuts in expensive funding rate should help NIM on high LDR (2Q16: 9%). We expect BBRI s NIM fall to be the least at 21bp next year, vs peers 24-29bp, on retail loan mix focus which is predominantly high-yield and fixed-rate. In the longer term, the superior micro banking franchise, which is nearly self-funded (17% 2Q16 LDR) with 85% CASA ratio, would help sustain growth sustainability in this segment. We reinitiate with a BUY at TP IDR14,2 Our GGM-based TP is built on 13.4% of WACC (risk-free rate of 6.8%), 17.7% of ROE, and 1% of growth rate, implying P/B target of 2.3x for 217F equivalent to a 5% premium to mean. Currently trading at -1 STD of 1.9x P/B 217F, which we deem as attractive for its 1) sustained lead in profitability and 2) NIM resilience amid declining interest rates. Downside risks include worse-than-expected trends in NPL and NIM falls, faster LDR rise, and slower loan growth than estimates. BUY for an exposure over one of the best micro banking franchise in the world. PRICE PERFORMANCE STOCK PERFORMANCE Igor Nyoman Putra igor.putra@bcasekuritas.co.id ext 178 Target Price: IDR14,2/share Current Price: IDR12,/share Upside potential: 18.3% 52-week price range (IDR) : 7,975 12,3 Market cap (IDRtn)/(USDbn) : / Avg Daily Turnover (IDR/USD) : / 24.8 Source: Bloomberg MARKET DATA Ytd 1M 3M 12M Absolute 5.%.% 15.4% 26.% JCI Return 16.3% -1.4% 9.5% 22.1% Relative -11.3% 1.4% 5.9% 3.9% Source: Bloomberg Exhibit 1. Financials and valuations Year-end 31 Dec F 217F 218F (IDR bn) Net interest income 51,49 58,279 63,24 68,475 75,49 Pre-prov. op. profit 34,84 39,43 43,916 45,48 49,233 Net profit 24,242 25,398 24,948 27,951 3,865 EPS (IDR) 983 1,3 1,11 1,133 1,251 EPS growth (%) (1.8) P/E (x) BVPS (IDR) 3,955 4,574 5,284 6,8 6,96 P/B (x) DPS (IDR) Yield (%) ROE (%) SHAREHOLDERS Government : 56.8% Public : 43.2% Source: Bloomberg Source: Bloomberg; Bank Rakyat Indonesia, BCA Sekuritas estimates

56 Exhibit 2. Bank Rakyat Indonesia - Financials Summary Income Statement Year-end 31 Dec (IDRbn) F 217F 218F Interest income 75,17 85,435 88,711 95,182 13,981 Interest expense (23,681) (27,156) (25,471) (26,77) (28,573) Net interest income 51,49 58,279 63,24 68,475 75,49 Non-interest income 9,299 12,49 16,166 16,44 17,743 Total operating income 6,789 7,688 79,46 84,519 93,152 Operating expenses (26,75) (31,285) (35,49) (39,39) (43,919) Pre-prov. op. profit 34,84 39,43 43,916 45,48 49,233 Provisions expense (5,719) (8,891) (14,3) (12,62) (12,124) Operating profit 28,364 3,511 29,616 33,418 37,11 Non-op. inc./(exp.) 2,495 1,983 1,983 1,983 1,983 Pre-tax profit 3,859 32,494 31,598 35,41 39,92 Corporate tax (6,65) (7,83) (6,636) (7,434) (8,29) Minorities (12) (13) (14) (16) (17) Net profit 24,242 25,398 24,948 27,951 3,865 EPS (IDR) 983 1,3 1,11 1,133 1,251 Balance Sheet Year-end 31 Dec (IDRbn) F 217F 218F Placement with other banks 22,28 2,911 22,85 24,97 27,287 Loans 495,15 564, , ,333 86,169 Government bonds 4,34 3,816 4,198 4,617 5,79 Other int-earning assets 26, , ,226 23, ,548 Financial assets impairment (15,886) (17,18) (26,124) (32,194) (37,997) Net earning assets 711, , , ,659 1,69,86 Cash 73,654 9,49 82,671 93,187 14,518 Fixed assets 5,917 8,39 8,843 9,728 1,7 Other assets 1,531 15,382 28,92 41,414 57,83 Total assets 81, , ,3 1,11,988 1,242,18 Customer deposits 622, , , , ,27 Deposits from other banks 9,365 12,126 13,338 14,672 16,139 Borrowing and sub-debts 49,864 57,69 63,459 69,84 76,785 Other liabilities 22,667 26,489 27,218 28,1 28,844 Total liabilities 74, , ,6 951,599 1,69,975 Minorities Equity 97,56 112,833 13,362 15, 171,686 Total liabilities and equity 81, , ,3 1,11,988 1,242,18 BVPS (IDR) 3,955 4,574 5,284 6,8 6,96 Breakdown Year-end 31 Dec (IDRbn) F 217F 218F As % of earning assets Gross loan Government bonds Other earnings assets As % of loan book Corporate SME/Commercial Consumer Micro Others As % of deposit Core deposit Time deposit Ratios & assumptions Year-end 31 Dec (IDRbn) F 217F 218F NIM Loan-to-deposit ratio Cost-to-income ratio Credit cost Gross NPL NPL coverage Tier-1 CAR ROA ROE Average asset yield Average funding cost Average loan yield Average deposit cost Loan growth Deposit growth EPS growth (1.8) Source: Bank Rakyat Indonesia, BCA Sekuritas estimates Declining provisions expense next year would help cushion bottom-line from receding top-line growth Rising retail loan mix augurs well for NIM Rising tier-1 CAR on the back of internally strong capital generation 56

57 1Q13 2Q13 3Q13 4Q13 1Q14 2Q14 3Q14 4Q14 1Q15 2Q15 3Q15 4Q15 1Q16 2Q16 TD ratio High-yield loan Risky loan ROE F ROE TD ratio High-yield loan Risky loan ROE F ROE TD ratio High-yield loan Risky loan ROE F ROE TD ratio High-yield loan Risky loan ROE F ROE Exhibit 3. Retained leads vs peers (%) 2Q BBRI is leading multiple metrics against peers, allowing ROE to stay at the top BBRI BMRI BBNI BBCA Source: Companies data; BCA Sekuritas estimates * TD ratio: time deposit ratio/ expensive funding ratio ** High-yield loan is defined as retail and micro loans *** Risky loan (by loan segment) is defined as non-soe corporate and SME loans Exhibit 4. BBRI loan segment distribution (%) Corporate Medium Small commercial Consumer Micro Others F 217F 218F Expanding retail loan (especially micro within loan mix) should prop up overall blended loan yield Source: Bank Rakyat Indonesia; BCA Sekuritas estimates Exhibit 5. Stabilizing new NPL formation while credit cost policy stays prudent (%) NPL formation (LHS) Credit cost (LHS) NPL coverage (RHS) (%) NPL coverage is sustained at +14% in the past four quarters albeit challenging milieu. Source: Bank Rakyat Indonesia; BCA Sekuritas 57

58 Feb-8 Jul-8 Dec-8 May-9 Oct-9 Mar-1 Aug-1 Jan-11 Jun-11 Nov-11 Apr-12 Sep-12 Feb-13 Jul-13 Dec-13 May-14 Oct-14 Mar-15 Aug-15 Jan-16 Jun-16 Jun-7 Apr-8 Feb-9 Dec-9 Oct-1 Aug-11 Jun-12 Apr-13 Feb-14 Dec-14 Oct-15 Aug-16 Exhibit 6. Micro KUR vs non-kur profitability Micro yield (% of avg asset) KUR Non-KUR Interest income Interest expense Net interest income Non-lending income.4.8 Total operating income Operating expense Insurance fee PPOP ROA Credit cost (incl. insurance) Pre-tax ROA Tax Post-tax ROA Source: Bank Rakyat Indonesia; BCA Sekuritas estimates Micro KUR is value-accretive as compared to the bank s level ROA of 3.2% on average despite having lower profitability vs non-kur micro Exhibit 7. 12M forward P/B (x) = Std +2 = 2.8x Trading at below mean currently while consistently holding the best ROE title, warranting to trade at a higher multiple 2.5 Std +1 = 2.5x Mean = 2.2x Std -1 = 1.9x Std -2 = 1.7x 1. Source: Bloomberg; Bank Rakyat Indonesia; BCA Sekuritas estimates Exhibit 8. 12M forward P/E (x) Std +2 = 12.9x Std +1 = 11.6x Mean = 1.2x Std -1 = 8.9x Std -2 = 7.6x P/E is relatively cheap given BBRI s high profitability Source: Bloomberg; Bank Rakyat Indonesia; BCA Sekuritas estimate 58

59 Exhibit 9. ROE decomposition analysis (yearly) Declines in total income yield should pressurize bottom-line, but to be compensated by softer credit cost and good cost control for the next two years Of avg. assets (%) F 217F 218F Net interest income Non-interest income Income yield Cost Core profit/ PPOP Credit cost Non-op. income yield Pre-tax ROA Tax rate Post-tax ROA Minorities Leverage (x) ROE Source: Bank Rakyat Indonesia; BCA Sekuritas estimates Exhibit 1. ROE decomposition analysis (quarterly) PPOP ROA maintained but credit cost and lower leverage (due mostly to asset revaluation in 2Q16) depressed ROE between 4Q15-2Q16 Of avg. assets (%) 3Q13 4Q13 1Q14 2Q14 3Q14 4Q14 1Q15 2Q15 3Q15 4Q15 1Q16 2Q16 Net interest income Non-interest income Income yield Cost Core profit/ PPOP Credit cost Non-op. income yield Pre-tax ROA Tax rate Post-tax ROA Minorities. (.) (.)... Leverage (x) ROE Source: Bank Rakyat Indonesia; BCA Sekuritas 59

60 BANK NEGARA INDONESIA BBNI IJ / BBNI.JK Company Reinitiation Financials Sector BUY Previous: - 22 Sep 216 Undervalued performer Robust loan demand prospect Our bullish loan growth outlook for the bank (BBNI) is premised on gov t s increased infrastructure focus (BBNI: 23% of 2Q16 loan) and relaxed mortgage policy (1% of 2Q16 loan). An expected economic up-cycle would also favorably trickle down to SME loan where BBNI commands a quite sizeable loan mix (35% of 2Q16 loan). As reference, BBNI consistently has generated above-peers loan growth with the past two quarters each at +2% YoY. We believe strong loan growth should linger going into F with our projected above-peers loan growth at % YoY on average. PRICE PERFORMANCE Igor Nyoman Putra igor.putra@bcasekuritas.co.id ext 178 Target Price: IDR6,4/share Current Price: IDR5,475/share Upside potential: 16.9% NIM stress issues factored in; NPL concerns receding Given the bank s high vulnerability on variable-rate loan segment (c.7% as of 2Q16) and comparatively high LDR (2Q16: 91% vs 88% large-four banks), we factor in declines in NIM (-35bp) the most within covered-banks (-32bp) during F. The bottom-line impact is to be reduced by a 95bp lower of credit cost and improved loan growth. Growth in both classified and risky loans have normalized with our anticipation of peaking risky loan ratio in 2Q16. BBNI s continued strict loan underwriting in SME and SOE-infra focus should allow to de-risk its loan portfolio. Taking off from a low expectation Even after a quarter loss in 2Q15 (on kitchen-sinking), a continuation of disappointments continues with consistent falls in QoQ earnings growth. This set the market expectation at the bottom. At 1.1x on 217F P/B equating to a 6% discount to -1 STD, fully priced-in negatives should limit downside risks, in our view. Further with anticipated better economic backdrop, BBNI is poised to being one of the best recovery plays after BMRI. We reinitiate with a BUY at TP IDR6,4 We like BBNI for its positive loan growth momentum (sizeable exposure over infra-related projects), superior deposit franchise (second lowest cost of funds), rising NPL coverage, and cheap valuation. The GGM-based TP reflects 13.4% WACC (risk-free rate of 6.8%), 14.5% ROE, and 1% growth rate, implying target 217F P/B of 1.3x (a 22% discount to earlier peak) equivalent to historical mean. Downside risks are NPL slippage, restructured loan relapse, NIM declines than we estimates, and slower loan growth than expected. STOCK PERFORMANCE 52-week price range (IDR) : 3,8 5,975 Market cap (IDRtn)/(USDbn) : 12.1 / 7.78 Avg Daily Turnover (IDR/USD) : / 13.1 Source: Bloomberg MARKET DATA Ytd 1M 3M 12M Absolute 9.7% -6.4% 12.9% 26.4% JCI Return 16.3% -1.4% 9.5% 22.1% Relative -6.6% -5.1% 3.4% 4.4% Source: Bloomberg Exhibit 1. Financials and valuations Year-end 31 Dec F 217F 218F (IDR bn) Net interest income 22,376 25,56 27,726 3,1 33,52 Pre-prov. op. profit 16,69 18,39 21,281 22,549 24,563 Net profit 1,783 9,67 1,46 11,413 13,9 EPS (IDR) EPS growth (%) 13.6 (15.9) P/E (x) BVPS (IDR) 3,47 3,977 4,381 4,84 5,399 P/B (x) DPS (IDR) Yield (%) ROE (%) SHAREHOLDERS Government : 6.% Public : 4.% Source: Bloomberg Source: Bloomberg; Bank Negara Indonesia, BCA Sekuritas estimates

61 Exhibit 2. Bank Negara Indonesia - Financials Summary Income Statement Year-end 31 Dec (IDRbn) F 217F 218F Interest income 33,365 36,895 4,65 43,416 46,963 Interest expense (1,989) (11,335) (12,925) (13,46) (13,911) Net interest income 22,376 25,56 27,726 3,1 33,52 Non-interest income 1,336 9,458 1,43 11,45 12,672 Total operating income 32,712 35,18 38,129 41,46 45,724 Operating expenses (16,13) (16,629) (16,848) (18,911) (21,161) Pre-prov. op. profit 16,69 18,39 21,281 22,549 24,563 Provisions expense (3,642) (7,333) (8,671) (8,24) (7,64) Operating profit 12,967 11,57 12,611 14,345 17,499 Non-op. inc./(exp.) Pre-tax profit 13,524 11,466 12,815 14,55 17,74 Corporate tax (2,695) (2,326) (2,691) (3,55) (3,718) Minorities (47) (74) (78) (82) (86) Net profit 1,783 9,67 1,46 11,413 13,9 EPS (IDR) Balance Sheet Year-end 31 Dec (IDRbn) F 217F 218F Placement with other banks 13,987 17,17 18,887 2,98 21,396 Loans 262,581 38, ,57 438,338 53,536 Government bonds 43,83 47,222 49,583 52,63 54,666 Other int-earning assets 51,834 66,68 53,93 57,212 6,719 Financial assets impairment (6,87) (11,849) (17,546) (23,85) (27,327) Net earning assets 365, ,95 486, , ,989 Cash 36,33 43,823 42,699 48,521 55,111 Fixed assets 6,222 2,757 22,417 24,21 26,147 Other assets 8,893 17,66 15,833 2,322 21,56 Total assets 416,574 58, ,86 637, ,34 Customer deposits 313,893 37, , , ,595 Deposits from other banks 3,177 4,698 5,42 5,672 5,956 Borrowing and sub-debts 2,524 34,215 35,926 37,722 39,68 Other liabilities 17,958 2,824 21,898 22,42 22,925 Total liabilities 355,552 43, ,82 542,972 61,84 Minorities 4,27 4,28 4,366 4,453 4,542 Equity 56,815 74,158 81,693 9,252 1,678 Total liabilities and equity 416,574 58, ,86 637, ,34 BVPS (IDR) 3,47 3,977 4,381 4,84 5,399 Breakdown Year-end 31 Dec (IDRbn) F 217F 218F As % of earning assets Gross loan Government bonds Other earnings assets As % of loan book Corporate SME/Commercial Consumer Micro Others As % of deposit Core deposit Time deposit Ratios & assumptions Year-end 31 Dec (IDRbn) F 217F 218F NIM Loan-to-deposit ratio Cost-to-income ratio Credit cost Gross NPL NPL coverage Tier-1 CAR ROA ROE Average asset yield Average funding cost Average loan yield Average deposit cost Loan growth Deposit growth EPS growth 13.6 (15.9) Source: Bank Negara Indonesia, BCA Sekuritas estimates Earnings growth is to recover for the next two years to % YoY NIM is to fall significantly on variable-rate loan segment dominant participation Robust loan growth driven by infrastructure projects and consumer loan segment 61

62 2Q12 3Q12 4Q12 1Q13 2Q13 3Q13 4Q13 1Q14 2Q14 3Q14 4Q14 1Q15 2Q15 3Q15 4Q15 1Q16 2Q16 1Q13 2Q13 3Q13 4Q13 1Q14 2Q14 3Q14 4Q14 1Q15 2Q15 3Q15 4Q15 1Q16 2Q16 216F 217F 218F 2Q13 3Q13 4Q13 1Q14 2Q14 3Q14 4Q14 1Q15 2Q15 3Q15 4Q15 1Q16 2Q16 216F 217F 218F Exhibit 3. Loan growth comparison between BBNI vs large peers (% YoY) BNI Large-three banks (Mandiri, BRI, BCA) Strong loan growth momentum, mainly led by infrastructure, will allow BBNI to maintain lead for the next two years 5 Source: Companies data; BCA Sekuritas Exhibit 4. Bottom-line impact from NIM stress alleviated by credit cost (%) NIM (LHS) Credit cost (RHS) (%) (1) (3) (5) Source: Bank Negara Indonesia; BCA Sekuritas Exhibit 5. Classified and risky loan trends (% YoY) Classified loan growth (LHS) Risky loan growth (LHS) (%) 1 Classified loan ratio (RHS) Risky loan ratio (RHS) Peak Stabilizing 2 (2) Risky loan ratio is expected to be the peak in 2Q16 while admittedly also posing relatively higher relapse risk Source: Bank Negara Indonesia; BCA Sekuritas Note: classified loan (NPL + SML); Risky loan (classified loan + current restructured loan) 62

63 Jun-8 Nov-8 Apr-9 Sep-9 Feb-1 Jul-1 Dec-1 May-11 Oct-11 Mar-12 Aug-12 Jan-13 Jun-13 Nov-13 Apr-14 Sep-14 Feb-15 Jul-15 Dec-15 May-16 Jun-8 Aug-9 Oct-1 Dec-11 Feb-13 Apr-14 Jun-15 Aug-16 1Q13 2Q13 3Q13 4Q13 1Q14 2Q14 3Q14 4Q14 1Q15 2Q15 3Q15 4Q15 1Q16 2Q16 Exhibit 6. Growing NPL coverage with prudent credit cost policy (%) NPL formation (LHS) Credit cost (LHS) NPL coverage (RHS) (%) Credit cost has mostly exceed the new NPL formation rate, resulting in higher NPL coverage Source: Bank Negara Indonesia; BCA Sekuritas estimates Exhibit 7. 12M forward P/B (X) Std +2 = 1.8x Std +1 = 1.6x Mean = 1.3x Std -1 = 1.1x Std -2 =.9x Looks undervalued to us for its potential recovery Source: Bloomberg; Bank Negara Indonesia; BCA Sekuritas estimates Exhibit 8. 12M forward P/E (x) Std +2 = 13.7x Std +1 = 11.8x Mean = 9.8x Std -1 = 7.8x Std -2 = 5.9x Trading slightly ahead of mean, which appears to us as quite alluring given the earlier peak at 14x Source: Bloomberg; Bank Negara Indonesia; BCA Sekuritas estimates 63

64 Exhibit 9. ROE decomposition analysis (yearly) Improving credit cost should offset soft PPOP, leading ROE to 14.6% by 218F Of avg. assets (%) F 217F 218F Net interest income Non-interest income Income yield Cost Core profit/ PPOP Credit cost Non-op. income yield.1.. (.) Pre-tax ROA Tax rate Post-tax ROA Minorities.... (.) Leverage (x) ROE Source: Bank Negara Indonesia; BCA Sekuritas estimates Exhibit 1. ROE decomposition analysis (quarterly) Recent ROE fall in 2Q16 attributable to declining total income yield, exacerbated by a sharp credit cost booking more than double the size in earlier quarter Of avg. assets (%) 3Q13 4Q13 1Q14 2Q14 3Q14 4Q14 1Q15 2Q15 3Q15 4Q15 1Q16 2Q16 Net interest income Non-interest income Income yield Cost Core profit/ PPOP Credit cost Non-op. income yield (.3) (.2) (.1) (.) Pre-tax ROA (.4) Tax rate (.1) Post-tax ROA (.3) Minorities Leverage (x) ROE (2.1) Source: Bank Negara Indonesia; BCA Sekuritas 64

65 BANK TABUNGAN NEGARA BBTN IJ / BBTN.JK Company Reinitiation Financials Sector BUY Previous: - 22 Sep 216 Extending growth momentum Positive loan demand outlook Amid President Jokowi s 1mn housing program and unrelenting backlog (11.4mn units in 215 by Bureau National Statistics), we view this as genuinely positive over Bank Tabungan Negara (BBTN) s strong loan demand outlook. Presently it has strong pipelines to fund 31k housing units vs 1H16 s realized mortgage disbursement of only 1k units. This would translate into strong construction loan (2Q16: 13% of loan) and later mortgage loans (78%). We expect BBTN to register the highest loan growth amid our covered banks for F at 16-19% YoY vs aggregate s 12-14% YoY. PRICE PERFORMANCE Igor Nyoman Putra igor.putra@bcasekuritas.co.id ext 178 Target Price: IDR2,4/share Current Price: IDR1,99/share Upside potential: 2.6% Addressed liquidity strain concerns We recognize investors concerns over the BBTN s funding strain, but there exists three reasons why we think BBTN s funding strength is understated: 1) current policy allows BBTN (BUKU III) to offer a more competitive deposit rates by 25bp higher vs BUKU IV banks while it holds similar risk profile rating (RBBR) #2; 2) mortgage loan demand is largely to be funded by government s subsidy programs (liquidity facility/ FLPP and interest subsidy); and 3) potentially implementation of SOE networks (ATM and EDC) integration would help boost retail deposits. In the medium to longer term, these are positive for NIM resilience and funding sustainability. FLPP issues possibly overdone; NPL fears receding The ongoing discussed 217 FLPP proposal (BBTN potentially bearing 4% of funding vs 1% now) is margin-dilutive, but BBTN intends to sustain a c.3-375bp of spread either by reducing the funding mix or hiking the mortgage rate (from 5%). A raised FLPP mortgage rate is arduous to reach on current low-rate setting, but a better funding mix may be realistic since BBTN holds a pivotal role for low-income home buyers. NPL issues are receding (2Q16/1Q16: 3.4%/3.6%) with 216F NPL/SML guided at < 3%/1% (2Q16 SML: 13%). We are optimistic BBTN to be able to attain its guided NPL resolution. STOCK PERFORMANCE 52-week price range (IDR) : 97 / 2,1 Market cap (IDRtn)/(USDbn) : 21.7 / 1.61 Avg Daily Turnover (IDR/USD) : / 3.48 Source: Bloomberg Reinitiating with TP IDR2,4, BUY; Our top pick among small banks In the medium term, BBTN should be the net beneficiary from declining interest rates since most of its deposits mix are expensive (2Q16: 55%) with high reliance on it (2Q16 LDR: 111%). Our GGM-based TP reflects 12.8% of WACC (risk-free rate of 6.8%), 13.9% ROE, and 1% growth rate, rendering a 1.4x P/B target based on 217F (a 7% discount from +1 STD). Downside risks are worse-than-expected trends in NPL and NIM falls, and unfavorable FLPP funding mix nor mortgage rate. Exhibit 1. Financials and valuations Year-end 31 Dec F 217F 218F (IDR bn) Net interest income 5,482 6,867 7,738 8,642 9,721 Pre-prov. op. profit 2,353 3,435 3,682 3,91 4,279 Net profit 1,145 1,851 2,249 2,658 3,52 EPS (IDR) EPS growth (%) (26.7) P/E (x) BVPS (IDR) 1,159 1,31 1,48 1,681 1,911 P/B (x) DPS (IDR) Yield (%) ROE (%) MARKET DATA Ytd 1M 3M 12M Absolute 53.7% 2.6% 21.7% 92.3% JCI Return 16.3% -1.4% 9.5% 22.1% Relative 37.3% 3.9% 12.2% 7.2% Source: Bloomberg SHAREHOLDERS Government : 6.% Public : 4.% Source: Bloomberg Source: Bloomberg; Bank Tabungan Negara, BCA Sekuritas estimates

66 Exhibit 2. Bank Tabungan Negara - Financials Summary Income Statement Year-end 31 Dec (IDRbn) F 217F 218F Interest income 13,374 15,567 17,71 19,411 21,355 Interest expense (7,893) (8,7) (9,972) (1,768) (11,634) Net interest income 5,482 6,867 7,738 8,642 9,721 Non-interest income 922 1,131 1,217 1,357 1,514 Total operating income 6,43 7,998 8,955 1, 11,235 Operating expenses (4,51) (4,563) (5,272) (6,98) (6,956) Pre-prov. op. profit 2,353 3,435 3,682 3,91 4,279 Provisions expense (777) (91) (836) (538) (418) Operating profit 1,576 2,534 2,846 3,363 3,862 Non-op. inc./(exp.) Pre-tax profit 1,578 2,542 2,847 3,364 3,863 Corporate tax (433) (691) (598) (76) (811) Minorities Net profit 1,145 1,851 2,249 2,658 3,52 EPS (IDR) Balance Sheet Year-end 31 Dec (IDRbn) F 217F 218F Placement with other banks 1, Loans 16, , , , ,273 Government bonds 8,238 8,231 8,313 8,396 8,48 Other int-earning assets 16,587 2,523 9,552 9,823 1,114 Financial assets impairment (1,581) (2,64) (2,42) (2,481) (2,42) Net earning assets 13, , ,25 21,6 241,195 Cash 1,292 12,168 11,229 13,16 15,116 Fixed assets 1,488 1,553 1,631 1,713 1,798 Other assets 2,191 3,12 6,961 8,238 8,364 Total assets 144, ,88 21,72 233,62 266,474 Customer deposits 16, , ,25 176,465 23,554 Deposits from other banks 1,77 3,255 3,58 3,938 4,332 Borrowing and sub-debts 18,752 2,796 23,862 27,431 3,174 Other liabilities 5,329 6,151 6,766 7,442 8,186 Total liabilities 132, , , , ,247 Minorities Equity 12,253 13,86 15,659 17,786 2,227 Total liabilities and equity 144, ,88 21,72 233,62 266,474 BVPS (IDR) 1,159 1,31 1,48 1,681 1,911 Breakdown Year-end 31 Dec (IDRbn) F 217F 218F As % of earning assets Gross loan Government bonds Other earnings assets As % of loan book Corporate SME/Commercial Consumer (non-mortgage) Subsidized mortgage Non-subsidized mortgage As % of deposit Core deposit Time deposit Ratios & assumptions Year-end 31 Dec (IDRbn) F 217F 218F NIM Loan-to-deposit ratio Cost-to-income ratio Credit cost Gross NPL NPL coverage Tier-1 CAR ROA ROE Average asset yield Average funding cost Average loan yield Average deposit cost Loan growth Deposit growth EPS growth (26.7) Source: Bank Tabungan Negara, BCA Sekuritas estimates Provisions expense is to be on the decline starting this year ahead of large banks Highly leveraged to improved domestic consumption and government-backed housing loan Mid-teens ROE to be sustained for the next two years, warranting higher multiples 66

67 1Q13 2Q13 3Q13 4Q13 1Q14 2Q14 3Q14 4Q14 1Q15 2Q15 3Q15 4Q15 1Q16 2Q F 217F 218F Exhibit 3. Loan growth comparison between BBTN vs aggregate banks (% YoY) BTN Aggregate Robust loan growth is to be driven by continued housing backlogs and construction (from extended mortgage loan), presenting positive loan demand outlook Source: Companies data; BCA Sekuritas estimates Exhibit 4. New NPL formation and coverage showing positive trends (%) NPL formation (LHS) Credit cost (LHS) NPL coverage (RHS) (%) (1) (2) (3) A very benign new NPL formation rate with credit cost staying prudent have lifted NPL coverage Source: Bank Tabungan Negara; BCA Sekuritas Exhibit 5. FLPP scenario analysis (%) FLPP scheme Scenario Old Proposed 1 2 Funding mix Government BBTN Total Scenario 1 suggests to us that there could be 7% downside risk to our earnings forecast Cost of funds (CoF) Gov't - FLPP BBTN bonds + deposit Insurance fee Blended CoF Product spread FLPP rate FLPP spread Source: Bank Tabungan Negara; BCA Sekuritas estimates 67

68 Dec-9 May-1 Oct-1 Mar-11 Aug-11 Jan-12 Jun-12 Nov-12 Apr-13 Sep-13 Feb-14 Jul-14 Dec-14 May-15 Oct-15 Mar-16 Aug-16 Dec-9 Aug-1 Apr-11 Dec-11 Aug-12 Apr-13 Dec-13 Aug-14 Apr-15 Dec-15 Aug-16 Exhibit 6. LDR vs CASA comparison (2Q16) (%) CASA ratio (LHS) Time deposit (LHS) LDR (RHS) BBCA BMRI BBRI BBNI BJBR BNGA BBTN BDMN Source: Companies data; BCA Sekuritas estimates (%) BBTN s weak funding structure would strengthen NIM resilience during the first leg of declining interest rate Exhibit 7. 12M forward P/B (x) At mean, P/B looks set to re-rate on the back of structurally robust loan growth trickling down to great earnings visibility Std +2 = 1.8x Std +1 = 1.5x Mean = 1.1x Std -1 =.8x Std -2 =.5x Source: Bloomberg; Bank Tabungan Indonesia; BCA Sekuritas estimates Exhibit 8. 12M forward P/E (x) Std +2 = 13.5x Std +1 = 11.2x Mean = 8.9x Std -1 = 6.6x Std -2 = 4.3x Trading at a multiple off the earlier peak in 1Q14 before the prelude of tightening monetary policy, yet the monetary policy has turned favorably now Source: Bloomberg; Bank Tabungan Negara; BCA Sekuritas estimates 68

69 Exhibit 9. ROE decomposition analysis (yearly) Rising ROE powered by subdued credit cost and rising leverage expected over F Of avg. assets (%) F 217F 218F Net interest income Non-interest income Income yield Cost Core profit/ PPOP Credit cost (.) Non-op. income yield.1.1 (.) (.) (.) (.) Pre-tax ROA Tax rate Post-tax ROA Minorities Leverage (x) (7.5) ROE (8.7) Source: Bank Tabungan Negara; BCA Sekuritas estimates Exhibit 1. ROE decomposition analysis (quarterly) Recent quarters of ROE drop determined by a downward credit cost trajectory despite of subdued PPOP ROA Of avg. assets (%) 3Q13 4Q13 1Q14 2Q14 3Q14 4Q14 1Q15 2Q15 3Q15 4Q15 1Q16 2Q16 Net interest income Non-interest income Income yield Cost Core profit/ PPOP Credit cost Non-op. income yield. (.) (.) (.) (.) (.) Pre-tax ROA Tax rate Post-tax ROA Minorities Leverage (x) ROE Source: Bank Tabungan Negara; BCA Sekuritas 69

70 BANK CIMB NIAGA BNGA IJ / BNGA.JK Company Initiation Financials Sector BUY Previous: - 22 Sep 216 Moving forward Transformation progress on track Within the bank (BNGA) s three-years focus over F, it aims at 1) diversifying income sources into non-interest income via transaction and digital banking, 2) increased cost disciplined, 3) enlarged CASA, 4) renewed risk management, and 5) working capital loan focus. Most of the aforementioned points have made a solid development in 1H16 with noninterest income mix at 23% (vs 21% avg in the past 2 years), 53% cost-toincome ratio (vs 56% avg), 3.4% cost-to-avg asset (vs %), CASA mix at 49% (vs 42% avg), and 4% NPL ratio (vs 2Q15 s peak 4.4%). Our talks with the bank s CFO reveal the current soft economy will likely recuperate next year while expressing optimism over tax amnesty receipts. Better stock liquidity is one positive catalyst post enlarged free floats via stock dividends. PRICE PERFORMANCE Igor Nyoman Putra igor.putra@bcasekuritas.co.id ext 178 Target Price: IDR1,/share Current Price: IDR815/share Upside potential: 22.7% Tapping the burgeoning domestic consumption The bank is genuinely focused on the working capital loan with a 218F goal skewing towards consumer and MSME loans at 6% (vs 49% now). These two segments are high-yielding loans and most leveraged to an up-cycle. On better domestic consumption and easy monetary policies; mortgage (13% of 2Q16 loan), credit card and personal (6%), and auto (1%) will likely grow buoyantly next year. We expect consumer loan to grow 3%/7% YoY, being the growth driver for BNGA s total loan growth at 1%/6% YoY in 216F/17F. Single-digit rate push and NPL less of concerns BNGA is one of the banks with the lowest loan yield among coverage. Our talks with the management suggests that interest rates for auto (mostly single rate already), mortgage (single rate), and corporate (single rate) are less adversely vulnerable to regulator s push on single-digit rate. SME rates (19% of 2Q16 loan) is 1.7% and comparatively lower than peers. As for asset quality, we notice a lower new NPL formation with continuous prudent credit cost policy leading to higher NPL coverage at 114% as of 2Q16. Lower credit cost this year (-6bp YoY) and next year (-3bp YoY) will drive earnings growth to jump substantially by 226%/29%/21% YoY in 216F/17F/18F. STOCK PERFORMANCE 52-week price range (IDR) : 48 1,9 Market cap (IDRtn)/(USDbn) : 2.48 / 1.56 Avg Daily Turnover (IDR/USD) : / 1.52 Source: Bloomberg We initiate with a BUY at TP IDR1, BNGA s positive progress operationally and nearly completed NPL recovery are not fully appreciated by the market with 217F P/B at.6x (14% below mean) vs target GGM-based P/B at.8x (slightly above mean). We are incorporating 11.8% WACC (6.8% risk-free rate), 1% ROE, and 5% long-term growth rate. We like the bank s continued prudent practice with rising NPL buffers, on-track transformation development, and high-yield loan focus in the longer term. Downside risks are higher credit cost, slower loan growth, and higher opex. Exhibit 1. Financials and valuations Year-end 31 Dec F 217F 218F (IDR bn) Net interest income 1,689 11,386 11,122 11,267 11,887 Pre-prov. op. profit 6,47 6,46 6,253 6,415 6,843 Net profit 2, ,485 1,916 2,32 EPS (IDR) EPS growth (%) (45.3) (81.7) P/E (x) BVPS (IDR) 1,132 1,141 1,2 1,276 1,369 P/B (x) DPS (IDR) Yield (%) ROE (%) MARKET DATA Ytd 1M 3M 12M Absolute 37.% -16.% 63.7% 52.3% JCI Return 16.3% -1.4% 9.5% 22.1% Relative 2.7% -14.6% 54.1% 3.3% Source: Bloomberg SHAREHOLDERS CIMB Group Sdn Bhd : 96.9% Public : 3.1%% Source: Bloomberg Source: Bloomberg; CIMB Niaga, BCA Sekuritas estimates

71 Exhibit 2. Bank CIMB Niaga - Financials Summary Income Statement Year-end 31 Dec (IDRbn) F 217F 218F Interest income 2,813 22,319 2,781 2,471 21,337 Interest expense (1,123) (1,932) (9,658) (9,24) (9,451) Net interest income 1,689 11,386 11,122 11,267 11,887 Non-interest income 3,22 2,895 3,258 3,529 4,84 Total operating income 13,712 14,281 14,381 14,796 15,971 Operating expenses (7,34) (8,236) (8,127) (8,381) (9,128) Pre-prov. op. profit 6,47 6,46 6,253 6,415 6,843 Provisions expense (3,449) (5,353) (4,263) (3,849) (3,738) Operating profit 2, ,99 2,565 3,15 Non-op. inc./(exp.) 242 (122) (1) (11) (12) Pre-tax profit 3,2 57 1,98 2,554 3,93 Corporate tax (856) (142) (495) (639) (773) Minorities (1) () () () () Net profit 2, ,485 1,916 2,32 EPS (IDR) Balance Sheet Year-end 31 Dec (IDRbn) F 217F 218F Placement with other banks 4,53 3,89 3,398 3,738 4,112 Loans 176, , ,51 19,227 27,294 Government bonds 11,484 16,5 16,21 16,372 16,536 Other int-earning assets 21,86 19,66 2,91 22,714 25,57 Financial assets impairment (6,69) (7,93) (8,1) (7,915) (7,68) Net earning assets 27,172 28, , , ,391 Cash 19,144 18,845 16,881 18,266 2,277 Fixed assets 2,71 2,978 3,127 3,283 3,447 Other assets 4,775 8,827 17,133 2,762 24,23 Total assets 233, , ,61 267, ,346 Customer deposits 174, , , , ,814 Deposits from other banks 2,4 5,654 6,22 6,842 7,184 Borrowing and sub-debts 17,588 22,756 24,734 26,72 27,51 Other liabilities 1,3 1,139 1,24 1,343 1,446 Total liabilities 24,715 21,17 218, , ,945 Minorities Equity 28,447 28,679 3,164 32,8 34,4 Total liabilities and equity 233, , ,61 267, ,346 BVPS (IDR) 1,132 1,141 1,2 1,276 1,369 Breakdown Year-end 31 Dec (IDRbn) F 217F 218F As % of earning assets Gross loan Government bonds Other earnings assets As % of loan book Corporate SME/Commercial Consumer Micro Others As % of deposit Core deposit Time deposit Ratios & assumptions Year-end 31 Dec (IDRbn) F 217F 218F NIM Loan-to-deposit ratio Cost-to-income ratio Credit cost Gross NPL NPL coverage Tier-1 CAR ROA ROE Average asset yield Average funding cost Average loan yield Average deposit cost Loan growth Deposit growth 6.7 (1.8) EPS growth (45.3) (81.7) Source: Bank Bank CIMB Niaga, BCA Sekuritas estimates We expect the net profit dip last year to be the bottom with strong net profit growth in F Expanding profitability metrics are on the card Forecasting limited loan yield declines given the 2 nd lowest loan yield within the 8 largest Indonesia banks 71

72 Jun-12 Sep-12 Dec-12 Mar-13 Jun-13 Sep-13 Dec-13 Mar-14 Jun-14 Sep-14 Dec-14 Mar-15 Jun-15 Sep-15 Dec-15 Mar-16 Jun-16 1Q13 2Q13 3Q13 4Q13 1Q14 2Q14 3Q14 4Q14 1Q15 2Q15 3Q15 4Q15 1Q16 2Q16 Exhibit 3. New NPL formation trends lower coupled with continued prudent credit cost practice, bringing NPL coverage higher (%) NPL formation (LHS) Credit cost (LHS) NPL coverage (RHS) (%) (.5) Rising NPL coverage and higher credit cost trend than new NPL formation comfort us Source: Bank CIMB Niaga; BCA Sekuritas Exhibit 4. NPL growth and SML growth trends improving (% YoY) SML growth NPL growth (1) (3) (1.7) (5) (12.3) A substantial contraction in both NPL and SML growth marks a positive development over NPL clean-ups attempts Source: Bank CIMB Niaga; BCA Sekuritas Exhibit 5. Potentially single-digit rate implementation not a major issue for BNGA (loan yields) (%) Q16 216E BNGA should be one of the best sheltered banks on single-digit rate implementation when the issue reemerges BBCA BNGA BMRI BBNI BBTN BDMN BBRI BJBR Source: Companies data; BCA Sekuritas 72

73 Dec-7 Aug-8 Apr-9 Dec-9 Aug-1 Apr-11 Dec-11 Aug-12 Apr-13 Dec-13 Aug-14 Apr-15 Dec-15 Aug-16 Dec-11 Jul-12 Feb-13 Sep-13 Apr-14 Nov-14 Jun-15 Jan-16 Aug-16 Exhibit 6. Mortgage loan mix and market share comparison (Jun-16) (%) Market share Loan mix BTN BCA BNI Mandiri BNGA BRI Bank bjb Mortgage significance will allow BNGA to enjoy the expected upturn in mortgage loan growth next year following multiple relaxed mortgage policies Excluding BBTN (subsidized mortgage specialist), BNGA commands the 2 nd highest loan mix stemming from mortgage Source: Bank Indonesia; Companies data; BCA Sekuritas Exhibit 7. 12M forward P/B valuation (x) (.1) Std +2 = 1x Std +1 =.9x Mean =.7x Std -1 =.5x Std -2 =.3x Recent correction makes valuation attractive to accumulate Source: Bloomberg; Bank CIMB Niaga; BCA Sekuritas estimates Exhibit 8. 12M forward P/E valuation (x) Std +2 = 38.2x Std +1 = 27.9x Mean = 17.6x Std -1 = 7.2x Recent EPS improvement and stability have made P/E valuation compelling; BUY since it is still trading below mean (1) Std -2 = -3.1x Source: Bloomberg; Bank CIMB Niaga; BCA Sekuritas estimates 73

74 Exhibit 9. ROE decomposition analysis (yearly) Bottom-line to be propped up by better top-line and benign credit cost Of avg. assets (%) F 217F 218F Net interest income Non-interest income Income yield Cost Core profit/ PPOP Credit cost Non-op. income yield (.) (.2) (.)... (.).1 (.1) (.) (.) (.) Pre-tax ROA Tax rate Post-tax ROA Minorities Leverage (x) ROE Source: Bank CIMB Niaga; BCA Sekuritas estimates Exhibit 1. ROE decomposition analysis (quarterly) ROE kept increasing from 3Q15 s bottom underpinned by NIM and credit cost momentum, expecting these bottom-line driver to persist next year Of avg. assets (%) 3Q13 4Q13 1Q14 2Q14 3Q14 4Q14 1Q15 2Q15 3Q15 4Q15 1Q16 2Q16 Net interest income Non-interest income Income yield Cost Core profit/ PPOP Credit cost Non-op. income yield (.) (.).. (.).4 (.) (.) (.) (.1) (.). Pre-tax ROA Tax rate Post-tax ROA Minorities... (.) (.) Leverage (x) ROE Source: Bank CIMB Niaga; BCA Sekuritas estimates 74

75 BANK DANAMON BDMN IJ / BDMN.JK Company Reinitiation Financials Sector HOLD Previous: - 22 Sep 216 Not yet there A tepid turnaround pace As details on post-restructuring steps (which is close to a conclusion) remain vague, we could only expect a slow turnaround for the bank. Recovering loan growth to 9% YoY next year (vs 216F: 5% YoY), retained cost control at 4.3% of avg asset (vs 4.6%), and lower credit cost of 2.9% (vs 3.3%) should allow ROE to rise to 9.5% vs 215 s bottom 7.2%. Not only is ROE the second lowest among coverage, but also lower vs our WACC assumption. High-yield mass-market contracting; SME challenged; Auto key BDMN s less focus on high-yield mass-market loan, presently at 22% of loan vs 2Q15/2Q14 at 26%/3%, is likely to cap NIM. Meanwhile concentration over SME loan would put the bank in an intense competition with other midto-large banks (i.e. BMRI, BBRI, BBNI, BBCA, BNLI IJ-not rated, PNBN IJ-not rated), since the market is congested already. Renewed burden on the bank s SME lending rate (at low-teens) is imminent given products availability from KUR (9% rate) and single-digit rate SME (9.75%; ticket size < IDR5bn) offered by the three SOE banks (ex. BBTN). We only expect auto loan growth (2Q16 36% of loan), from Adira unit (ADMF IJ-NR), to slightly rebound +1% YoY (216F: 5% YoY) next year on better domestic consumption from relaxed LTV and low interest rates. Renewed operational headwinds On top of NIM stress emanating from SME competition, we expect operational headwinds to mount next year when BDMN embarks on growing more rapidly. High LDR (2Q16: 118%; +5% YoY) and high reliance on time deposit ratio (59%; +3% YoY) would give NIM benefits in the medium term from contracting interest rates. We see rising 217F opex slightly to 1.7% YoY (216F/215: -5%/-11% YoY) as it has been focused on cost rationalization. Longer term, structurally depleting system excess liquidity will challenge its NIM and loan growth sustainability on weak funding base. We reinitiate with HOLD at TP IDR4, Our neutral outlook on BDMN is on account of 1) the continued slow turnaround, 2) renewed operational headwinds towards the new normal (new low level of NIM and excess liquidity), and 3) a cheap valuation. Our GGMbased target 217F P/B of 1.x (slightly below mean) assumes 11.8% WACC (risk-free rate of 6.8%), 11.6% ROE, and 6% growth rate. We like BDMN on potential turnaround while benefiting from low-rate milieu, but doubtful over the longer term ability to address the new normal issues. Upside risks are buoyant auto loan growth, lower credit cost and cost growth, and better NIM. Downside risks are slower loan growth, sharp NIM falls, and high cost growth. Exhibit 1. Financials and valuations Year-end 31 Dec F 217F 218F (IDR bn) Net interest income 13,911 13,887 12,461 12,937 14,288 Pre-prov. op. profit 8,49 9,25 8,36 8,924 9,874 Net profit 2,65 2,391 2,883 3,532 4,26 EPS (IDR) EPS growth (%) (35.5) (8.2) P/E (x) BVPS (IDR) 3,42 3,54 3,751 4,9 4,32 P/B (x) DPS (IDR) Yield (%) ROE (%) PRICE PERFORMANCE STOCK PERFORMANCE Igor Nyoman Putra igor.putra@bcasekuritas.co.id ext 178 Target Price: IDR4,/share Current Price: IDR3,81/share Upside potential: 5.% 52-week price range (IDR) : 2,665-4,22 Market cap (IDRtn)/(USDbn) : / 2.78 Avg Daily Turnover (IDR/USD) : /.95 Source: Bloomberg MARKET DATA Ytd 1M 3M 12M Absolute 19.1% 11.4% 8.2% 5.8% JCI Return 16.3% -1.4% 9.5% 22.1% Relative 2.7% 12.8% -1.3% -16.3% Source: Bloomberg SHAREHOLDERS Asia Financial (Pte) ltd : 67.4% JPMCB-Franklin Templeton : 6.7% Others :.% Public : 25.9% Source: Bloomberg Source: Bloomberg; Bank Danamon, BCA Sekuritas estimates

76 Exhibit 2. Bank Danamon - Financials Summary Income Statement Year-end 31 Dec (IDRbn) F 217F 218F Interest income 22,991 22,421 2,268 21,163 23,95 Interest expense (9,81) (8,534) (7,87) (8,226) (8,87) Net interest income 13,911 13,887 12,461 12,937 14,288 Non-interest income 4,766 4,616 4,885 5,179 5,697 Total operating income 18,676 18,53 17,346 18,116 19,985 Operating expenses (1,627) (9,478) (9,4) (9,192) (1,112) Pre-prov. op. profit 8,49 9,25 8,36 8,924 9,874 Provisions expense (3,986) (5,82) (4,378) (4,126) (4,12) Operating profit 4,63 3,944 3,928 4,797 5,772 Non-op. inc./(exp.) (51) (662) Pre-tax profit 3,554 3,282 3,978 4,847 5,822 Corporate tax (871) (812) (985) (1,2) (1,441) Minorities (78) (78) (11) (115) (121) Net profit 2,65 2,391 2,883 3,532 4,26 EPS (IDR) Balance Sheet Year-end 31 Dec (IDRbn) F 217F 218F Placement with other banks 11,473 9,659 11,591 12,75 14,25 Loans 17,89 99,652 17, , ,853 Government bonds 6,65 6,916 7,242 7,583 7,94 Other int-earning assets 52,468 53,699 54,285 58,843 64,932 Financial assets impairment (3,976) (4,485) (4,945) (4,833) (4,856) Net earning assets 173, , , ,19 213,895 Cash 13,125 12,239 11,369 12,774 14,331 Fixed assets 2,49 2,559 2,687 2,821 2,963 Other assets 6,435 7,819 11,882 14,45 15,58 Total assets 195,79 188,57 22,96 222, ,768 Customer deposits 116, ,142 12, , ,745 Deposits from other banks 2,429 1,827 2,193 2,412 2,653 Borrowing and sub-debts 27,14 22,8 26,389 29,28 3,376 Other liabilities 16,627 14,73 16,377 19,9 22,289 Total liabilities 162, , , ,518 25,63 Minorities Equity 32,78 33,932 35,951 38,423 41,45 Total liabilities and equity 195,79 188,57 22,96 222, ,768 BVPS (IDR) 11,473 9,659 11,591 12,75 14,25 Breakdown Year-end 31 Dec (IDRbn) F 217F 218F As % of earning assets Gross loan Government bonds Other earnings assets As % of loan book Corporate SME/Commercial Consumer Micro Others As % of deposit Core deposit Time deposit Ratios & assumptions Year-end 31 Dec (IDRbn) F 217F 218F NIM Loan-to-deposit ratio Cost-to-income ratio Credit cost Gross NPL NPL coverage Tier-1 CAR ROA ROE Average asset yield Average funding cost Average loan yield Average deposit cost Loan growth 2.7 (7.) Deposit growth 6.7 (1.2) EPS growth (35.5) (8.2) Source: Bank Danamon, BCA Sekuritas estimates Opex trend is benign Key impetus for ROE uptrend would emanate from tapering credit cost trend A strong %YoY EPS growth is off a low base in

77 Jun-7 Dec-7 Jun-8 Dec-8 Jun-9 Dec-9 Jun-1 Dec-1 Jun-11 Dec-11 Jun-12 Dec-12 Jun-13 Dec-13 Jun-14 Dec-14 Jun-15 Dec-15 Jun F 217F 218F F 217F 218F F 217F 218F F 217F 218F F 217F 218F F 217F 218F F 217F 218F F 217F 218F Exhibit 3. Future ROE comparison with other banks (%) ROE Rising ROE trend but still substantially lower than other banks BMRI BBRI BBNI BBCA BBTN BNGA BDMN BJBR Source: Bank Danamon; BCA Sekuritas Exhibit 4. SME loan leading overall loan growth (% YoY) SMEC (SME + Commercial) Mass-market Total loan Mass-market loan (which contains auto loan/micro at 1%/4% of 2Q16 total loan) will continue to be weighed down by contracting highyield micro loan book 2 1 (1) (2) 6 (8) (13) Source: Bank Danamon; BCA Sekuritas Exhibit 5. Danamon commands the weakest deposit franchise (2Q16) (%) CASA ratio (LHS) Time deposit (LHS) LDR (RHS) (%) BBCA BMRI BBRI BBNI BJBR BNGA BBTN BDMN Weak deposit franchise is to support the bank s NIM over the medium term Source: Companies data; BCA Sekuritas 77

78 Dec-7 Jan-9 Feb-1 Mar-11 Apr-12 May-13 Jun-14 Jul-15 Aug-16 May-7 Apr-8 Mar-9 Feb-1 Jan-11 Dec-11 Nov-12 Oct-13 Sep-14 Aug-15 Jul F 217F 218F Exhibit 6. Projecting a renewed opex trend, but still manageable vs peers, on a more buoyant pace next year (% YoY) Cost/avg. asset (RHS) Opex growth (LHS) (%) (5) (1) (15) () (11) (5) On cost rationalization attempts, cost-to-avg asset has improved significantly (and will remain sustained in our forecast) from the past s 6-7% Source: Bank Danamon; BCA Sekuritas estimates Exhibit 7. 12M forward P/B (x) At present the bank is trading meaningfully below the forward book value Std +2 = 1.7x Std +1 = 1.4x Mean = 1.2x Std -1 =.9x Std -2 =.7x Source: Bloomberg; Bank Danamon; BCA Sekuritas estimates Exhibit 8. 12M forward P/E (x) Std +2 = 19.7x Std +1 = 16.9x Mean = 14.1x Std -1 = 11.3x Std -2 = 8.5x and at around -1 STD by P/E as the market seems to be awaiting a successful turnaround on the bank s business restructuring Source: Bloomberg; Bank Danamon; BCA Sekuritas estimates 78

79 Exhibit 9. ROE decomposition analysis (yearly) Better credit cost and rising leverage would dictate the ROE movement Of avg. assets (%) F 217F 218F Net interest income Non-interest income Income yield Cost Core profit/ PPOP Credit cost Non-op. income yield (.2).1 (.4) (.6) (.5) (.5) (.) (.3) (.3)... Pre-tax ROA Tax rate Post-tax ROA Minorities Leverage (x) ROE Source: Bank Danamon; BCA Sekuritas estimates Exhibit 1. ROE decomposition analysis (quarterly) Better income yield and benign credit cost have lifted ROE momentum Of avg. assets (%) 3Q13 4Q13 1Q14 2Q14 3Q14 4Q14 1Q15 2Q15 3Q15 4Q15 1Q16 2Q16 Net interest income Non-interest income Income yield Cost Core profit/ PPOP Credit cost Non-op. income yield (.) (.) (.) (.1) (.1) (.9). (.1) (.) (1.3).. Pre-tax ROA Tax rate Post-tax ROA Minorities Leverage (x) ROE Source: Bank Danamon; BCA Sekuritas 79

80 BANK JABAR BANTEN BJBR IJ / BJBR.JK Company Initiation Financials Sector HOLD Previous: - 22 Sep 216 Profitable but pricey Progress of de-risking attempts at a standstill? To resolve the legacy NPL, Bank Jabar Banten (BJBR) focuses on hastening the write-offs and asset recoveries. It aims at a flattish loan growth for micro and mortgage this year on high NPL. Risks are to be reduced by renewed underwriting standard, a cooperation with insurance companies and rural banks (linkage). The 3.5% 2Q16 NPL ratio has been sinking for five quarters straight, but recently a new NPL formation rate sharply surged (+3.3% QoQ). Under-provisioned on less prudent credit cost policy As of 2Q16 loans-loss reserves ratio/absolute was at 1.5%/IDR971bn which is the lowest quarter since 2Q7/2Q13. We think BJBR is attempting to smooth earnings by keeping credit cost low with 2Q16/215 at.5%/.2%, the lowest period ever recorded. The credit cost is unusually low and should catch up with the recent pick-ups in non-civil servant loans (2Q16 commercial +42% YoY/17% of loan) in the medium term. 2Q16 NPL coverage dipped to 43% on aggressive write-offs to keep a lid on NPL. In a normal state credit cost should return to 1% levels to sustain coverage of 7%, posing 12% downside risk to our earnings forecast. We see.5% credit cost and 49% NPL coverage for 217F on BJBR s willingness to be at ease on coverage ratio policy. BJBR is the only bank we expect to see rising credit cost ahead. Operational headwinds Key loan growth driver civil servants (2Q16: +75% YoY)/pensioners (9% YoY) look unsustainably high with the former relying on deeper market penetration (already high at 75%) and the latter on salary hike (regulator dependency). Rivalry in this market is intense, competing with i.e. BMRI, Bank Tabungan Pensiunan (BTPN IJ-not-rated) and BBRI. Recent earnings jump in 2Q16 (+11% YoY) is due to unsustainable shots (unusually low credit cost, +19% YoY LDR surge, and time deposit rate cut), in our view. These factors leave limited room for further utilization, signifying operational headwinds next year. Initiating at TP IDR1,575 with HOLD Our neutral view is on 1) spiking valuations (YTD 19% outperforming JCI and P/B at a 13% premium from +2 STD), 2) impending operational headwinds for the next 12M, and 3) highest ROE. Our GGM-based TP implies 1.4x 217F P/B (slightly sub +2 STD) on 12.3% WACC (6.8% risk-free rate), 15.4% ROE, and 5% growth. Recent Bank Banten news pose minor downside risks if Banten s province pulled out their IDR6bn/IDR2.5tn loan/deposit from BJBR (2%/3% of 2Q16). Upside risks are better loan growth, lower credit cost, and better NIM. Downside risks are higher credit cost and slower loan growth. PRICE PERFORMANCE STOCK PERFORMANCE Igor Nyoman Putra igor.putra@bcasekuritas.co.id ext 178 Target Price: IDR1,575/share Current Price: IDR1,675/share Downside potential: 6.% 52-week price range (IDR) : 585 1,76 Market cap (IDRtn)/(USDbn) : / 1.24 Avg Daily Turnover (IDR/USD) : / 2.83 Source: Bloomberg MARKET DATA Ytd 1M 3M 12M Absolute 121.9% 1.2% 67.5% 153.8% JCI Return 16.3% -1.4% 9.5% 22.1% Relative 15.5% 11.6% 58.% 131.7% Source: Bloomberg Exhibit 1. Financials and valuations Year-end 31 Dec F 217F 218F (IDR bn) Net interest income 4,462 4,976 5,667 6,37 7,289 Pre-prov. op. profit 1,99 1,872 2,226 2,53 2,87 Net profit 1,14 1,377 1,61 1,73 1,92 EPS (IDR) EPS growth (%) (19.8) P/E (x) BVPS (IDR) ,12 P/B (x) DPS (IDR) Yield (%) ROE (%) SHAREHOLDERS East Java Government : 61.9% Banten Government : 13.1% Others :.% Public : 25.% Source: Bloomberg Source: Bloomberg; Bank Jabar Banten, BCA Sekuritas estimates