The Quarter BSI's Departure Finally Shows Up in the Numbers
The Q1 2026 post covered Bank Mandiri's headline shock - losing control of PT Bank Syariah Indonesia (Persero) Tbk (BSI) on February 1, 2026 - but flagged that Q1's year-over-year comparisons were only partly distorted, since BSI was still in the consolidated group for January before the common-control transfer took effect. Q2 2026 (April through June) is the first quarter this backfill has tracked where the deconsolidation is felt for all three months on both sides of the comparison's absence - last year's Q2 2025 comparative still had BSI fully consolidated, and this year's Q2 2026 has none of it. Isolating Q2 standalone from the six-month interim filing (H1 2026 less the already-reported Q1 2026 figures) shows exactly what that looks like: net interest, sharia and insurance income - this backfill's revenue proxy - fell 13.27% YoY (Rp22,993,786 million vs Rp26,511,399 million), nearly six times sharper than H1's blended -7.84% YoY decline. And yet net income attributable to owners for the same quarter grew 33.48% YoY to Rp15,028,188 million (from Rp11,257,823 million) - the sharpest quarterly profit acceleration this backfill has recorded since Q3 2024. A bank whose top-line proxy fell double digits while its bottom line grew a third faster than a year ago is not, on its face, a business getting healthier - it's a business whose comparison base just lost a subsidiary, and the credit-cost and cost-discipline story underneath deserves more scrutiny than the headline number invites (see Beyond the Usual).
Cumulatively, H1 2026 consolidated net income attributable to owners rose 24.36% YoY to Rp30,412,319 million (from Rp24,455,082 million in H1 2025), with EPS up a similar 24.46% to Rp326.11 (from Rp262.02) - a real acceleration from H1 2025's own -7.89% YoY decline (per the H1 2025 post). Operating income (Profit from Operations) grew a genuine 16.81% YoY to Rp39,062,791 million, comfortably outrunning even the un-distorted parts of the P&L. But the H1 blend hides the quarter-to-quarter shape: Q1 2026 standalone net income was Rp15,384,131 million; Q2 2026 standalone came in essentially flat at Rp15,028,188 million (-2.31% QoQ) - so almost none of H1's YoY improvement is a Q2 acceleration on top of Q1; it's Q1's already-reported 16.57% YoY gain (per the Q1 2026 post) plus a similarly-sized Q2 gain, both flattered by the same missing-subsidiary comparison problem, just to different degrees each quarter.
The Prescription
Bank Mandiri should stop treating the Head Office segment as a catch-all for cost reallocation and start disclosing a genuine unallocated-cost bridge every quarter, not just when a segment note happens to be filed - this backfill has now tracked the pattern (lending segments improve, Head Office absorbs a disproportionate offsetting swing) across five separate reporting periods since FY2024, and a reader still can't tell how much of any quarter's headline growth is organic versus a reallocation artifact. The bank's real strength - a fast-growing, cheaply-funded core deposit base outside BSI, up double digits YoY even as the sharia subsidiary left the group - is being under-communicated by a segment presentation that keeps burying the lending businesses' genuine improvement under an opaque corporate-center number.
What it should stop doing: paying out a dividend sized to prior-year cash generation (Rp44,472,220 million this half, up 2.21% YoY) in the same period bank-only CAR fell 2.17 percentage points in a single quarter to 17.52% - the sharpest one-quarter capital decline since Q1 2025's crash from 20.10% to 17.29%. A state-controlled bank with a fixed dividend policy tied to prior-year earnings, rather than to the current quarter's actual capital consumption from credit-risk RWA growth, is exactly the setup that turns an otherwise-healthy capital position into a recurring source of investor anxiety - see Beyond the Usual below.
Key Financial Metrics
H1 2026 vs H1 2025 (P&L, consolidated), plus Q2 2026 standalone vs Q2 2025 standalone (H1 less Q1) - see the caveat above before reading either comparison as a pure organic trend
FX: IDR 17,880.00 = USD 1 (June 30, 2026 close, per the bank's own filed exchange-rate disclosure); IDR 16,675.00 = USD 1 (December 31, 2025 close, per the FY2025 post) - the Rupiah weakened a further 7.2% in the six months since, on top of the multi-quarter depreciation this blog has tracked since 2024.
| Metric | H1 2026 (IDR) | H1 2026 (USD) | H1 2025 (IDR) | YoY |
|---|---|---|---|---|
| Net interest, sharia & insurance income ("Net Revenue" equivalent) | Rp48,330,490M | ~$2.70B | Rp52,444,656M | ⚠️ -7.8% |
| Operating Income (Profit from Operations) | Rp39,062,791M | ~$2.18B | Rp33,442,833M | ✅ +16.8% |
| Net Income (attributable to owners) | Rp30,412,319M | ~$1.70B | Rp24,455,082M | ✅ +24.4% |
| EPS (basic, six-month) | Rp326.11 | ~$0.018 | Rp262.02 | ✅ +24.5% |
| Metric (Q2 standalone = H1 less Q1) | Q2 2026 (IDR) | Q2 2025 (IDR) | YoY |
|---|---|---|---|
| Net interest, sharia & insurance income | Rp22,993,786M | Rp26,511,399M | ⚠️ -13.3% |
| Operating Income | Rp19,140,934M | Rp15,329,013M | ✅ +24.9% |
| Net Income (attributable to owners) | Rp15,028,188M | Rp11,257,823M | ✅ +33.5% |
Operating cash flow was -Rp103,052,329 million for the half (from H1 2025's +Rp97,316,260 million), against fixed- and intangible-asset capex of Rp1,528,129 million. Isolating Q2 2026 standalone (H1 less Q1's already-reported -Rp86,676,485 million operating cash flow) puts the quarter's operating cash flow at roughly -Rp16,375,844 million - still negative, but a much smaller drag than Q1's, which had carried the one-off -Rp36,368,200 million "Impact of Loss of Control of Subsidiary" cash-flow line. Total cash and cash equivalents at period-end is Rp264,212,231 million (~$14.78B), up 3.61% from Q1 2026's Rp255,015,412 million, though still down sharply from December 31, 2025's Rp379,877,189 million (largely BSI's cash leaving the consolidated group in February, per the Q1 2026 post).
| Balance sheet metric | Jun 30, 2026 (IDR) | Jun 30, 2026 (USD) | Jun 30, 2025 (IDR) | YoY |
|---|---|---|---|---|
| Total Assets | Rp2,527,568,193M | ~$141.37B | Rp2,514,684,369M | ✅ +0.5% |
| Loans (gross, consolidated) | Rp1,630,495,652M | ~$91.19B | Rp1,655,987,678M | ⚠️ -1.5% |
| Total Deposits (customer, consolidated) | Rp1,763,167,223M | ~$98.61B | Rp1,582,776,360M | ✅ +11.4% |
| Total Equity (attributable to owners) | Rp284,593,401M | ~$15.92B | N/A (Dec 2025: Rp293,750,669M) | -3.1% vs Dec-25 |
Total assets and equity are essentially flat to down YoY even with BSI gone from the group - deposits actually grew double digits YoY, meaning the core bank (ex-BSI) added more in new deposits than BSI's departure removed from the balance sheet. Gross loans fell only 1.5% YoY, a far milder move than assets' post-deconsolidation shrinkage in Q1's QoQ comparison - most of BSI's own loan book evidently sat inside the "Sharia" segment column rather than materially shrinking the group's core lending franchise.
Key Operational Metrics
Bank-only (Individual/parent-entity) ratios, filed in the bank's June 30, 2026 published financial statements (OJK format) - the bank-only entity never consolidated BSI in the first place, so these ratios are unaffected by the deconsolidation, though CASA and LDR below are computed directly from the parent-entity balance sheet.
Funding quality
- Loan-to-Deposit Ratio» (LDR) - bank-only, filed: 92.73%, up from March 2026's 91.34% (per the Q1 2026 post) and from June 2025's 90.22% - continuing to climb for a third straight quarter tracked in this backfill.
- CASA» ratio - bank-only, computed: 70.49% (Rp1,205,490,188M demand + saving deposits ÷ Rp1,710,032,217M total deposits) - down from March 2026's 71.71% and now even below December 2025's 70.83%, reversing the one quarter of improvement the Q1 2026 post had flagged as tentative. See Beyond the Usual.
- Net Open Position (NOP) - overall: 2.34%, up from 1.44% a year earlier - still comfortably within the regulatory ceiling, but the largest NOP this backfill has recorded.
Capital and credit quality
- CAR» (capital adequacy) - bank-only, filed: 17.52%, down sharply from March 2026's 19.69% and from June 2025's 18.35% - the sharpest single-quarter capital decline since Q1 2025's crash. Consolidated CAR: 17.88%, down from June 2025's 19.45% and December 2025's 20.43%. See Beyond the Usual.
- NPL» ratio - gross, filed: 0.98%, improved from 1.08% a year earlier. Net NPL: 0.40%, essentially flat YoY (0.39%).
Profitability
- ROE», filed: 24.28%, up from 21.06% a year earlier - the strongest bank-only ROE this backfill has recorded.
- ROA, filed: 3.10%, up from 2.89% a year earlier.
- NIM», filed: 4.34%, down from 4.61% a year earlier - continuing the multi-year margin compression this blog has tracked since 2024, now happening even as the balance sheet shrank BSI's higher-margin sharia financing out of the bank-only base.
- Cost-to-Income Ratio» (CIR), filed: 35.60%, sharply improved from 43.40% a year earlier - the single most favorable YoY ratio move in this table, though see the caveat in the opening section about how much of any cost-line improvement this year is organic versus BSI's own cost base leaving the comparison.
Not available this quarter: an NPL Coverage Ratio figure, a results presentation/analyst deck, or an earnings call transcript - none were located as of this post; only the OJK-format published financial statements and the interim consolidated financial statements themselves.
Five Segments, One Head Office Offset That Finally Stopped Growing
Bank Mandiri reports five core banking segments - Corporate Banking, Commercial Banking, Government & Institutional Banking, Retail Banking, and Treasury & International Banking - plus a Head Office/corporate-center segment, a Sharia associate-entity column (BSI, now equity-accounted rather than consolidated), an Insurance segment, and other-subsidiary and elimination columns. Comparing H1 2026 to H1 2025 cumulatively:
- Corporate Banking: net income Rp8,528,776 million, +46.25% YoY (from Rp5,831,912 million) - the strongest of the five core segments.
- Treasury & International Banking: Rp2,165,309 million, +77.75% YoY (from Rp1,218,174 million) - the fastest percentage grower, though off a smaller base; isolating Q2 2026 standalone (H1 less the already-reported Q1's Rp3,437,822 million) shows the segment actually swung to a loss of roughly Rp1,272,513 million for the quarter alone - a reminder that a strong cumulative number can still hide a weak or negative standalone quarter for a segment whose income is driven by trading and treasury positioning rather than steady interest income.
- Government & Institutional Banking: Rp1,401,897 million, +24.82% YoY (from Rp1,123,077 million).
- Commercial Banking: Rp5,076,809 million, +21.92% YoY (from Rp4,164,197 million).
- Retail Banking: Rp15,499,547 million, +11.18% YoY (from Rp13,940,958 million) - the largest segment by profit and the slowest grower of the five, the same pattern the Q1 2026 post flagged.
Combined, the five lending segments' profit grew +24.33% YoY, from Rp26,278,318 million to Rp32,672,338 million - a Rp6,394,020 million absolute gain. The Head Office segment's loss widened only 0.50% YoY, from Rp5,499,771 million to Rp5,527,373 million - a Rp27,602 million move, essentially flat. That's a genuine break from the pattern this backfill has tracked in every segment-note quarter since FY2024 (most recently the Q1 2026 post's finding that the Head Office loss widened 69.3% YoY, offsetting just over half the lending segments' combined gain): this half, for the first time, the Head Office offset barely moved at all, meaning almost the entire Rp6.39 trillion lending-segment improvement flows through to the consolidated number largely intact rather than being substantially muted at the corporate-center level. Whether this is a genuine change in how costs get allocated or a one-half anomaly won't be confirmed until Q3 2026's segment note, if filed.
The Sharia associate column (BSI, equity-accounted from February 2026) shows Rp786,020 million for H1 2026, against Rp3,741,223 million a year earlier when BSI was still fully consolidated - not a decline in BSI's own performance, but the mechanical effect of the group now booking only its equity-method share of BSI's profit rather than the subsidiary's full income statement. Insurance grew from Rp193,056 million to Rp643,081 million (+233.2% YoY), and other subsidiaries grew modestly (+12.25% YoY, Rp1,355,939 million to Rp1,522,006 million).
Beyond the Usual
Bank-only capital adequacy fell 2.17 percentage points in a single quarter, the same quarter a large dividend went out
Bank-only CAR fell from March 2026's 19.69% to June 2026's 17.52% - the sharpest one-quarter decline this backfill has recorded outside Q1 2025's crash from 20.10% to 17.29%. Consolidated CAR moved similarly, from December 2025's 20.43% to June 2026's 17.88%. The timing lines up with a Rp44,472,220 million dividend payment recorded in this half's cash-flow statement (up 2.21% YoY from H1 2025's Rp43,510,539 million) plus continued growth in credit-risk risk-weighted assets (bank-only credit RWA rose from Rp1,208,049,003 million at December 2025 to Rp1,292,848,460 million at June 2026). Both capital ratios remain comfortably above regulatory minimums (a filed minimum CAR based on risk profile of 9.73% this quarter), so this isn't a solvency concern - but it's the second time in six quarters this backfill has tracked a rapid capital drawdown at Bank Mandiri, and a dividend policy that doesn't flex with the quarter's actual capital consumption is a real, recurring pattern rather than a one-off.
Bank-only CASA reversed its one-quarter improvement and is now below where it stood in December
The Q1 2026 post flagged bank-only CASA's rise to 71.71% - the first quarterly improvement after five straight quarters of decline dating back to mid-2025 - as "a single data point, not a confirmed trend reversal." This quarter confirms that caution was warranted: CASA fell back to 70.49%, below even December 2025's 70.83%. The underlying driver hasn't changed since this blog first flagged it in the 9M 2025 post: bank-only time deposits (Deposito) grew faster than low-cost demand and savings deposits again this half, continuing the shift toward more expensive funding that has pressured net interest margin (down to 4.34% from 4.61% a year earlier) for over a year now.
Related-party exposure hit a new high, driven by state food-security and infrastructure lending
Total assets with related parties reached 29.23% of total consolidated assets at June 2026, up from 25.66% at December 2025 and 25.35% a year earlier - the highest ratio this backfill has recorded. Part of the jump is mechanical (BSI's equity stake, now booked as a related-party investment rather than eliminated on consolidation, added Rp27,271,566 million on its own), but the loan book itself is also concentrating further: lending to PT Agrinas Pangan Nusantara (Persero) (formerly PT Yodya Karya (Persero), a food-security-focused state vehicle) grew to Rp55,000,000 million from Rp46,697,487 million at December 2025 and from nothing a year earlier, while lending to Perum Bulog (the state food-logistics agency) grew more than 5x in the half, from Rp2,223,843 million to Rp11,976,157 million. Unused committed loan facilities extended to related parties also rose to Rp158,639,817 million from Rp128,582,195 million at December 2025 - now 52.2% of the bank's total unused loan-facility liability. None of this is disclosed as anything other than ordinary related-party lending under common state control, and Bank Mandiri's broader related-party loan book (a figure this backfill has tracked since FY2024) isn't disclosed at this cadence to size precisely - but the concentration of new lending growth in a small number of state food-security vehicles is worth watching as a funding-quality theme distinct from the bank's ordinary corporate and retail book.
A planned shareholder spin-off of Mandiri Sekuritas' asset-management arm became a sale to Danantara instead
An "Assets and liabilities held for sale" line appeared on the balance sheet for the first time in this backfill (Rp346,221 million at June 2026, up from Rp253,774 million at December 2025), representing PT Mandiri Manajemen Investasi (MMI), an asset-management subsidiary of PT Mandiri Sekuritas. The footnote reveals a change of plan mid-stream: as of December 2025, Bank Mandiri's stated intention was to transfer MMI's shares directly to Bank Mandiri's own shareholders within twelve months - effectively a spin-off. On April 1, 2026, that plan changed: Mandiri Sekuritas instead signed a Conditional Share Purchase Agreement with PT Danantara Asset Management (DAM), the state's asset-management holding vehicle, to transfer 99.93% of MMI to DAM for Rp1,024,317 million - a related-party sale rather than a distribution to public shareholders. The unconditional Share Purchase Agreement was then signed on July 22, 2026 (disclosed as a subsequent event in this filing), formalizing the transfer. It's a small transaction relative to the group's balance sheet, but it's a second data point (after BSI itself) of a state-owned financial asset being routed to Danantara rather than to the market or to Bank Mandiri's own shareholders directly.
Q2's treasury swing shows how much a "cumulative" segment number can hide
The Treasury & International Banking segment's H1 2026 profit of Rp2,165,309 million looks like solid growth against H1 2025's Rp1,218,174 million (+77.75%) - but isolating Q2 2026 standalone shows the segment actually lost roughly Rp1,272,513 million for the quarter alone, since Q1 2026 had already booked Rp3,437,822 million on its own. A segment whose income depends on trading positions and market-making rather than steady interest spread can swing from a strong quarter to a loss the very next one, and a reader who only sees the half-year cumulative number would never know Q2 alone was negative.
Bank Mandiri Shares Hit a Fresh Two-Year Low as the Rupiah Kept Weakening
The Q1 2026 post closed at Rp4,720 on March 31, 2026, itself a 9.2% YoY decline. That decline continued and accelerated through Q2: shares fell to Rp4,390 at end-April, Rp4,080 at end-May, and closed the quarter at Rp3,850 on June 30, 2026 - an -18.4% net move for the quarter and a -21.1% decline from June 30, 2025's Rp4,880 close. Zooming out to the two-year window ending this quarter (July 2024 through June 2026), shares peaked at Rp7,125 in August 2024 and just made a new trough at this quarter's Rp3,850 close - a -46.0% peak-to-trough decline, the steepest this backfill has recorded and a clear break below every low point tracked since (September 2025's Rp4,400, February 2025's Rp4,600). The currency backdrop kept moving the same direction it has all backfill: the Rupiah weakened further from Rp16,675/USD at December 31, 2025 to Rp17,880.00/USD at June 30, 2026 - a 7.2% depreciation in six months, continuing the multi-quarter trend this blog has tracked since 2024 and adding a second headwind (currency, on top of the share-price decline) for any USD-based investor.
Coverage at a Glance
| Metric | H1 2026 | H1 2025 | YoY | Why it matters |
|---|---|---|---|---|
| Net income (attributable to owners) | Rp30.41T | Rp24.46T | +24.4% | Real growth cumulatively, but Q2-standalone alone (+33.5% YoY) is now the fully BSI-affected comparison |
| Net interest, sharia & insurance income | Rp48.33T | Rp52.44T | -7.8% | Q2-standalone fell 13.3% YoY - the first quarter where BSI's absence is felt for all three months on both sides |
| Bank-only CAR | 17.52% | 18.35% | -0.83pp | Fell 2.17pp QoQ from March's 19.69% - sharpest quarterly capital drop since Q1 2025 |
| Bank-only CASA | 70.49% | N/A (Dec-25: 70.83%) | -1.22pp vs Dec-25 | Reversed Q1 2026's one-quarter improvement; now below December's level |
| Bank-only LDR | 92.73% | 90.22% | +2.51pp | Third straight quarter of sequential tightening |
| Head Office segment loss (H1 YoY) | Rp5.53T | Rp5.50T | +0.5% | First half in this backfill where the Head Office offset barely moved, letting lending-segment gains flow through |
| Related-party share of total assets | 29.23% | 25.35% | +3.88pp | New high, driven partly by BSI's equity stake and growing SOE-linked lending |
| Share price (Jun 30, 2026 close) | Rp3,850 | Rp4,880 | -21.1% | Fresh two-year low; -46.0% peak-to-trough since Aug-2024 |
Target Valuation Range
Market cap Rp359.3 trillion (~$20.10B), ~1.26x P/B, ~5.77x P/E. Bottom line: cheap on trailing multiples by this backfill's own history, but the cheapness reflects real, not merely sentimental, concerns - a capital ratio moving the wrong way, a funding-cost metric (CASA) that reversed its one bright quarter, and a share price that just broke every low this blog has tracked. This isn't yet a "buy the dip" setup; it's a "confirm the capital and funding trends stabilize" setup.
Using trailing-twelve-month net income attributable to owners of approximately Rp62,251,187 million (FY2025's Rp56,293,950 million less H1 2025's Rp24,455,082 million plus H1 2026's Rp30,412,319 million), implied TTM EPS of approximately Rp667.32, and the issued share count of 93,333,333,332:
| Market cap → book value | Q2 2026 |
|---|---|
| Share price (period-end) | Rp3,850 |
| Shares outstanding | 93,333,333,332 |
| Market capitalization | Rp359.3 trillion (~$20.10B) |
| Total equity (book value) | Rp284,593,401M |
| Book value per share | Rp3,049.93 |
| Peer-multiple sanity check | Q1 2026 | Q2 2026 | Change |
|---|---|---|---|
| P/B | 1.44x | 1.26x | down |
| P/E | 7.5x | 5.77x | down |
P/E: ~5.77x, using TTM EPS of ~Rp667.32 against the Rp3,850 close on June 30, 2026. P/B: ~1.26x, using book value per share of approximately Rp3,049.93 (Rp284,593,401M total equity attributable to owners ÷ 93,333,333,332 shares). Market capitalization: approximately Rp359.3 trillion (~$20.10B). That's cheaper on both multiples than the Q1 2026 post's ~7.5x P/E and ~1.44x P/B - TTM earnings kept growing (from ~Rp58.5 trillion to ~Rp62.3 trillion) while the share price fell far faster, so both multiples compressed hard. Against Indonesia's other large conventional banks (BCA, BRI), Bank Mandiri now trades at a meaningfully wider discount than it has through most of this backfill's history - a discount a bull case would attribute to temporary BSI-transition noise and a bear case would attribute to the CAR and CASA trends above being the start of a genuine capital-and-funding-quality problem, not noise. The Head Office segment's stopped-growing offset (see Beyond the Usual) is the one genuinely encouraging structural signal this half - if it holds into Q3 2026, it would mean the lending businesses' real improvement is finally reaching the headline number rather than being muted, which would argue for the cheap multiple closing rather than persisting.
A full multi-year DCF isn't built here - Bank Mandiri's earnings this half are distorted by a one-off deconsolidation, not a stable base to project forward - so the scenarios below are illustrative, applying a stated P/E multiple to the current TTM EPS of ~Rp667.32 rather than modeling multi-year cash flows:
| Scenario | Key assumption | Multiple applied | Implied price |
|---|---|---|---|
| Bear | CAR keeps falling and CASA keeps reversing; market prices in a genuine capital/funding problem, de-rating further below today's already-compressed multiple | 5.0x P/E | Rp3,337 (~$187) |
| Base | Capital and funding trends stabilize but don't fully recover; multiple partially reverts toward Q1 2026's 7.5x, landing roughly midway | 6.5x P/E | Rp4,338 (~$243) |
| Bull | BSI-transition noise is confirmed temporary, CAR/CASA stabilize by Q3 2026, and the multiple reverts toward the 8.0-8.5x range this backfill has traded at in recent history | 8.0x P/E | Rp5,339 (~$299) |
| Current (period-end close) | Actual Q2 2026 close | 5.77x P/E (actual) | Rp3,850 (~$216) |
PT Bank Mandiri (Persero) Tbk's reviewed interim consolidated financial statements as of and for the six-month period ended June 30, 2026 (with comparatives for June 30, 2025, December 31, 2025 and December 31, 2024), and the bank's OJK-format published financial statements for the same period.