BBRI The Worst Loan Category Got Fixed. A Different One Just Took Its Place.

The Corporate Fix, and Medium's New Problem

The Q1 2018 post ended on an open question: Corporate Non-SoE's loan-mix NPL had improved to 5.83%, but a same-quarter re-segmentation (which raised the size thresholds separating Small Commercial, Medium, and Corporate loans) made it impossible to tell how much of that improvement was genuine credit-quality repair versus loans simply migrating between buckets - and, more troubling, the officially-reported Corporate segment's own net income kept falling in the same quarter its associated NPL ratio improved, a divergence the filing never explained.

This quarter answers both questions cleanly, in the same direction. Corporate Non-SoE's gross NPL fell again, to 5.38% (from 5.83% at Q1, on the same restated size thresholds), and the officially-reported Corporate segment's own income rebounded hard: Rp255,208M in Q2 alone, against just Rp134,400M in Q1 - a jump from a 7.9% net margin to 12.9%, pulling the half-year margin to 10.56%, actually above FY2017's already-weak 9.46% full-year margin for the first time this series has recorded. Corporate's six-month income now sits at 53.1% of all of FY2017's full-year total - a normal half-year run-rate - up from Q1's alarming 18.3% share. Whatever drove the divergence the Q1 2018 post flagged, it didn't persist: both the loan-mix NPL and the segment's own profitability moved the same way this quarter.

But BRI's loan-mix NPL table has six categories, and while attention was on Corporate's recovery, Medium quietly became the new problem: its gross NPL jumped from 5.20% at Q1 to 6.11% at Q2 - now the single worst-performing loan-mix category BRI discloses, ahead of Corporate Non-SoE's 5.38%. Medium's loan book barely grew over the same period (+1.8% YoY, per BRI's own materials, the slowest growth of any loan-mix category), so this isn't a fast-growing book outrunning its own underwriting - it's a stagnant book getting worse. The Special Mention (early-warning) bucket for Medium didn't spike alongside it (6.04% at Q1, 6.17% at Q2, essentially flat), which suggests this quarter's NPL jump crystallized out of loans already flagged as watch-list rather than a fresh wave of newly-troubled credit - a smaller worry than an unexplained spike would be, but still the category to watch next.

Consolidated net income attributable to owners grew 10.9% year-over-year for the half (Rp14,886,046M vs Rp13,422,688M), with operating income growing faster again at 14.2% - continuing the pattern the Q1 2018 post noted, where cost and provisioning growth finally isn't outrunning revenue growth.

The Prescription

BRI should publish the loan-mix NPL table with the same granularity it already applies to Corporate Non-SoE - a running commentary on which category is worst and why - rather than letting the "worst performer" title change hands quarter to quarter without comment in the deck itself. Medium just took over that title from Corporate Non-SoE without a single sentence of management explanation anywhere in the presentation or the filed notes; a reader has to build the comparison from six separate quarterly data points to notice it happened at all. The bank clearly tracks this internally - the loan-mix table itself proves it - so silence on a category crossing into "worst" status isn't a data problem, it's a disclosure choice, the same gap the Q1 2018 post and the FY2017 post both flagged about Corporate Non-SoE before it.

What BRI should stop doing: letting its capital ratio's post-dividend recovery quietly stall without comment. Total CAR fell again this quarter (20.74%→20.13%, Bank), a full quarter after the AGM dividend that the Q1 2018 post documented as the cause of Q1's drop - except this time it's not the dividend; risk-weighted assets grew 11.0% over the half while total capital actually fell 2.7% (see Beyond the Usual below). A capital ratio that keeps falling for a second straight quarter, for a second and different reason, deserves its own line of management commentary, not just a re-run of the CAR table.

Key Financial Metrics

H1 2018 vs. H1 2017 (P&L, consolidated), and Jun 2018 vs. Dec 2017 / Mar 2018 (balance sheet, consolidated)

FX: approximately IDR 14,404 = USD 1, the approximate Bank Indonesia JISDOR reference rate around June 29, 2018 (not disclosed in this filing itself).

Metric	H1 2018 (IDR)	H1 2018 (USD)	H1 2017 (IDR)	YoY
Interest, Sharia and Premium Income, net ("Net Revenue" equivalent)	Rp38,661,652M	~$2,684M	Rp36,038,911M	✅ +7.3%
Other Operating Income	Rp10,196,235M	~$708M	Rp9,364,650M	✅ +8.9%
Total Other Operating Expenses	Rp19,759,696M	~$1,372M	Rp18,361,010M	⚠️ +7.6%
Operating Income	Rp18,466,599M	~$1,282M	Rp16,173,772M	✅ +14.2%
Net Income (attributable to owners)	Rp14,886,046M	~$1,034M	Rp13,422,688M	✅ +10.9%
Total Comprehensive Income (attributable to owners)	Rp10,833,082M	~$752M	Rp14,318,417M	⚠️ -24.3%¹
EPS (basic, six-month, consolidated)	Rp121.78	~$0.0085	Rp109.81	✅ +10.9%

¹ A larger version of the same bond-portfolio mark-to-market swing the Q1 2018 post flagged for that quarter alone: this half's other comprehensive income swung to a Rp5,402,377M unrealized loss on available-for-sale securities and Government Recapitalization Bonds (before tax), against a Rp1,413,680M gain in the year-ago half - not an operating result, but the securities book's own valuation move as Indonesian bond yields rose over the half.

Balance sheet metric	Jun 2018 (IDR)	Jun 2018 (USD)	Dec 2017 (IDR)	vs Dec'17	Mar 2018 (IDR)	QoQ
Total Assets	Rp1,153,228,286M	~$80.06B	Rp1,126,248,442M	✅ +2.4%	Rp1,119,240,112M	✅ +3.0%
Loans (gross, incl. sharia financing and finance lease)	Rp794,295,617M	~$55.14B	Rp739,336,520M	✅ +7.4%	Rp757,677,626M	✅ +4.8%
Total Deposits (Demand + Savings + Time)	Rp837,993,962M	~$58.18B	Rp841,656,450M	⚠️ -0.4%	Rp827,059,605M	✅ +1.3%
Total Liabilities	Rp986,548,641M	~$68.49B	Rp958,900,948M	⚠️ +2.9%	Rp958,606,454M	➖ +2.9%
Total Equity (attributable to owners)	Rp164,746,126M	~$11.44B	Rp166,748,817M	⚠️ -1.2%	Rp160,020,851M	✅ +3.0%
Total Cash and Cash Equivalents (per cash flow statement)	Rp153,856,121M	~$10.68B	Rp186,410,433M	⚠️ -17.5%	Rp175,477,339M	⚠️ -12.3%

Equity recovered from Q1's dividend-driven dip (Rp160,020,851M → Rp164,746,126M, +3.0% QoQ) as this half's retained earnings started rebuilding it, but total cash kept falling for a second straight quarter - down 12.3% QoQ on top of Q1's 5.9% drop - now a cumulative 17.5% decline since year-end, as loan growth (+7.4% since Dec 2017) continued to outpace deposit growth (-0.4% since Dec 2017). Deposits actually contracted slightly over the half even as loans grew nearly 8%, the same loans-ahead-of-deposits pattern the Q1 2018 post called a recurring first-half BRI pattern - except this year it didn't reverse by Q2 the way earlier years' Q1 dips typically did.

Net income grew double digits for a fifth straight quarter, and the divergence Q1 flagged between Corporate Non-SoE's NPL and its segment income resolved itself cleanly this quarter - but a different loan-mix category quietly took over as BRI's worst-performing one, and nobody in the filing said so (see The Corporate Fix, and Medium's New Problem above).

Key Operational Metrics

Bank-only, per BRI's own investor presentation, unless stated

• CASA^»: 59.51% (Bank, Jun 2018) vs 57.61% (Bank, Mar 2018) and 57.62% (Bank, Jun 2017) - a genuine improvement on both counts, breaking the Q1 seasonal dip pattern this series has tracked in prior years.
• Loan-to-deposit ratio (LDR)^»: 95.27% (Bank, Jun 2018) vs 92.26% (Bank, Mar 2018) and 89.76% (Bank, Jun 2017) - further above management's own target band than Q1, and unlike prior years' Q1 dips, this one kept climbing into Q2 rather than easing.
• Net Interest Margin (NIM)^»: 7.64% (Bank, Jun 2018) vs 7.49% (Bank, Mar 2018) and 8.02% (Bank, Jun 2017) - a small QoQ uptick, but still down year-over-year, continuing the gradual multi-year compression this series has tracked since 2015.
• ROA^» (before tax): 3.37% (Bank, Jun 2018) vs 3.35% (Bank, Mar 2018) and 3.31% (Bank, Jun 2017) - essentially flat, still holding the Q1 2018 post's break from the prior multi-year decline pattern.
• ROE^» (Tier 1): 19.33% (Bank, Jun 2018) vs 18.70% (Bank, Mar 2018) and 19.12% (Bank, Jun 2017) - modestly up on both counts.
• CAR^» (Total, Bank): 20.13% (Jun 2018) vs 20.74% (Mar 2018) and 21.67% (Jun 2017) - a second consecutive quarterly decline, this time not from the dividend payout that drove Q1's drop - see Beyond the Usual below. Tier 1 CAR fell to 19.14% from 19.76%.
• NPL ratio - gross (Bank): 2.33% (Jun 2018) vs 2.39% (Mar 2018) and 2.23% (Jun 2017) - improved QoQ, still slightly worse year-over-year.
• NPL ratio - net (Bank): 1.10% (Jun 2018) vs 1.16% (Mar 2018) and 1.16% (Jun 2017) - improved on both counts.
• Cost-to-income (Opex/Opr. Income, bank-only): 70.50% (Bank, Jun 2018) vs 70.43% (Bank, Mar 2018) and 72.33% (Bank, Jun 2017) - essentially flat QoQ, improved year-over-year.
• Loan-mix NPL by category (re-segmented basis): Micro 1.41%, Consumer 1.27%, Small Commercial 3.70%, Medium 6.11%, Corporate Non-SoE 5.38%, SoE 0.44% (all Bank, Jun 2018, per BRI's investor presentation, on the same restated size thresholds introduced at Q1 2018 - see The Corporate Fix, and Medium's New Problem above). Medium is now BRI's worst-performing loan-mix category, overtaking Corporate Non-SoE for the first time this series has recorded.
• Segment income (Micro/Retail/Corporate/Others/Subsidiaries): available this quarter - see Five Segments, a Rebound and a Wobble below.
• Geographic segment: Indonesia, USA, Hong Kong, Singapore, and Timor Leste - see Beyond the Usual below for Timor Leste's continued ramp.

Five Segments, a Rebound and a Wobble

BRI reports five operating segments: Micro, Retail, Corporate, Others (mainly treasury and non-lending activities), and Subsidiaries - distinct from the loan-mix NPL categories discussed above. This filing carries a genuine six-month comparative column (H1 2017), so the figures below are read against a real year-ago half, not just a share of the full prior year the way the Q1 2018 post had to.

Micro remains BRI's largest segment by income: Rp22,876,476M total income and Rp7,480,613M income for the half, a 32.7% net margin - down from FY2017's 36.5% full-year margin, and the segment's six-month income sits at 47.8% of FY2017's full-year total, slightly below a normal half-year run-rate.

Retail posted Rp17,161,959M total income and Rp4,937,008M income for the half, a 28.8% net margin, close to FY2017's 28.0% full-year margin. Its 55.1% share of FY2017's full-year net income is the closest of any segment to a clean half-year pace.

Corporate posted Rp3,690,490M total income and Rp389,608M income for the half, a 10.6% net margin - above FY2017's 9.5% full-year margin, and a sharp reversal from Q1 alone, when the segment's quarterly income was running at just 18.3% of FY2017's full-year total (see The Corporate Fix, and Medium's New Problem above). Q2 alone contributed Rp255,208M of the half's Rp389,608M, on a 12.9% margin - nearly double Q1's 7.9%.

Others posted Rp3,000,500M total income and Rp1,748,243M income for the half (58.3% margin), continuing to run as BRI's highest-margin segment, consistent with its non-lending treasury and securities business.

Subsidiaries posted Rp2,128,462M total income and Rp378,664M income for the half (17.8% margin) - but that blends a strong Q1 (26.0% margin) with a much weaker Q2 alone (Rp1,090,715M total income, just Rp108,940M income, a 10.0% margin). Subsidiaries is the one segment whose Q2 moved the opposite direction from Corporate's rebound, though this filing's notes don't break out which subsidiary drove the swing.

Ranked by six-month income as a share of FY2017's full-year total, the order has shuffled from Q1: Subsidiaries now leads at 65.8%, followed by Others (56.0%), Corporate (53.1%), Retail (55.1%), and Micro trailing at 47.8% - the segment mix looks far more evenly paced at the half-year mark than Q1's snapshot suggested, when Corporate's 18.3% share was the standout outlier.

Beyond the Usual

Medium loans quietly became BRI's new worst-performing category

Corporate Non-SoE's loan-mix NPL improved for a second straight quarter (5.83% at Q1 2018 to 5.38% at Q2 2018, on the restated size thresholds introduced this year), and the officially-reported Corporate segment's own income rebounded from Q1's weak 7.9% margin to 12.9% in Q2 alone - resolving the divergence the Q1 2018 post flagged. But over the same two quarters, Medium's NPL rose from 5.20% to 6.11%, overtaking Corporate Non-SoE as the single worst-performing loan-mix category BRI discloses - a first in this series' data. Medium's own loan book grew just 1.8% year-over-year, the slowest of any category, so this reads as a stagnant book deteriorating rather than a fast-growing one outrunning underwriting. Neither the presentation nor the filed notes mention the handoff; a reader has to build the six-quarter series themselves to notice BRI's "worst category" problem just moved to a different line of the same table. See The Corporate Fix, and Medium's New Problem above.

A second straight capital-ratio decline, this time without a dividend to blame

Bank-only Total CAR fell from 20.74% at Q1 2018 to 20.13% at Q2 2018 - a second consecutive quarterly decline, following the Rp13,048,441M AGM dividend payout the Q1 2018 post identified as that quarter's cause. This quarter's decline has a different mechanical driver: per the filing's own capital-adequacy note, risk-weighted assets grew 11.0% over the half (Rp704,515,985M to Rp782,131,078M) - largely from market-risk RWA more than doubling (Rp6,889,063M to Rp14,798,179M, +114.8%) alongside ordinary credit-risk RWA growth - while total regulatory capital actually fell 2.7% (Rp161,751,939M to Rp157,411,267M) as June's unrealized bond-portfolio losses (see the Key Financial Metrics table above) flowed through into capital. A capital ratio recovering from a dividend dip by falling for a structurally different reason the following quarter is worth tracking into Q3, not treated as a continuation of the same story.

BRI's Timor Leste branch, which generated a full year's worth of income in a single quarter at Q1 2018, kept scaling through Q2: total assets grew to Rp3,745,730M at June 30, 2018, roughly ten times December 2017's Rp379,288M, and the branch turned a Rp3,070M pre-tax profit for the half against a Rp2,360M pre-tax loss in the year-ago half. This is now a genuinely material early-stage growth story inside BRI's geographic footprint, not a rounding-error presence.

BRI disclosed a new employee compensation program this half: a Management and Employee Stock Ownership Program (MESOP) granted January 1, 2018 with a fair value of Rp541,925M, to be paid out in shares and amortized through completion no later than March 31, 2020. A new "Compensation Provision of Bonus Shares" equity line (Rp212,667M) appeared on the balance sheet this quarter as a result - the first new share-based compensation program this series has recorded at BRI since the older MSOP stock-option scheme (exercised out between 2004 and 2010).

This filing's significant-agreements note discloses one genuinely new contract signed this quarter - a six-month consulting agreement with PT Accenture for BRI's Digital Bank implementation (Rp29,500M, signed May 2, 2018) - continuing the technology-buildout pattern this series has tracked since 2016. It also discloses a 12-month telecommunications agreement with PT Telekomunikasi Selular (Rp24,915M) dated January 29, 2018 that the Q1 2018 post didn't have; that post's own Beyond the Usual section recorded four other agreements from the same window (PT PP's Gatot Subroto Tower package and three network/telecom agreements) but not this one, which only surfaced once this half's cumulative list was filed.

The bank's pending-litigation allowance grew 5.4% quarter-over-quarter (Rp1,004,354M to Rp1,058,041M) - a modest acceleration from Q1's 3.4% pace, but still far below FY2017's 58.3% full-year growth rate. Management's language on the underlying claims (ordinary contractual-compliance disputes, no material effect on operations expected) is unchanged from prior filings.

Target Valuation Range

P/E ~10.37x, P/B ~1.92x - cheaper on every multiple than any quarter this series has recorded, entirely because the stock fell faster than earnings did - not because anything in this quarter's fundamentals deteriorated.

BRI's share price closed around Rp2,582 on June 29, 2018 (the last trading session of the quarter) - down 21.1% from the Q1 2018 post's Rp3,273 close, and down 6.9% from Rp2,773 a year earlier. That decline crossed this series' usual threshold for its own dedicated discussion: the stock fell in every one of the quarter's three months (Rp3,273 at March-end, Rp2,927 in April, Rp2,800 in May, Rp2,582 in June), tracking the broader emerging-market currency selloff of mid-2018 - Bank Indonesia raised its benchmark rate three times between May and June 2018 specifically to defend the Rupiah, which weakened past Rp14,000/USD over the same window - rather than any company-specific news in BRI's own filings this quarter.

Shares outstanding remain 123,345,810,000 issued, of which 1,108,590,000 sit in treasury (122,237,220,000 outstanding) - unchanged from the Q1 2018 post.

Market cap → enterprise value	H1 2018
Share price (period-end)	Rp2,582
Shares outstanding	122,237,220,000
Market capitalization	~Rp315,617B (~$21.91B)
Total liabilities	Rp986,549B
Less: cash and equivalents	Rp153,856B
Enterprise value	~Rp1,148,310B (~$79.72B)

Total equity attributable to owners (book value) is Rp164,746B, giving the P/B ~1.92x used below.

Trailing-twelve-month EPS is approximately Rp248.90 (FY2017's Rp236.93, minus H1 2017's Rp109.81, plus this half's Rp121.78), giving a P/E^» of ~10.37x against the Rp2,582 close.

Peer-multiple sanity check	Q1 2018	H1 2018	Change
P/E»	~13.46x	~10.37x	✅ down sharply - entirely on the price decline since trailing earnings kept growing
P/B»	~2.50x	~1.92x	✅ down - both the price fall and book value per share rising as the dividend effect wore off

Both multiples compressed together this quarter, the cleanest "both compress" pattern since the price-driven declines earlier in this series. A same-period peer read is available again: BBCA's Q2 2018 post reported ROE of 17.26% against BRI's 19.33% (Tier 1), NIM of 6.05% against BRI's 7.64%, and gross NPL of 1.43% against BRI's 2.33% (both bank-only) - the same higher-margin/higher-return/higher-delinquency trade-off this series has tracked all year. BCA's P/E of ~21.9x against BRI's ~10.37x is now roughly 2.1x, the widest gap this series has recorded - wider than the ~1.66x-1.83x range earlier quarters showed, because BRI's stock fell considerably further than BCA's over the same quarter.

A full DCF still isn't included here for the same reason the Q1 2018 post gave: this quarter's loan-mix table shows the bank's worst-performing category can change hands from one quarter to the next without explanation, which isn't a stable base for multi-year credit-quality assumptions. The peer-multiple read above, alongside Beyond the Usual, is the honest valuation lens for this quarter - and on that lens, BRI looks meaningfully cheaper than it did a quarter ago, for reasons that have more to do with the Rupiah than with BRI's own results.

---