A Sharia Bank Left the Group, and the Headline Number Didn't Notice
The FY2025 post closed on a bank whose Q4 profit reversal - impairment charges collapsing ~60% YoY, salary growth decelerating - still hadn't answered whether bank-only CASA» would ever stop falling, and whether the missing segment note (absent for three straight reporting periods) would return to show whether the reversal was broad-based. Bank Mandiri's reviewed interim consolidated financial statements for the three months ended March 31, 2026 answer both questions, but only after clearing a much bigger one first: on February 1, 2026, Bank Mandiri lost control of PT Bank Syariah Indonesia (Persero) Tbk (BSI), its sharia banking subsidiary, which is deconsolidated from the group's financial statements for two of this quarter's three months. The transaction wasn't a sale to an outside buyer - the filing books it as a "Difference in Transactions of Entities Under Common Control" of Rp111,953 million added to additional paid-in capital, the same accounting treatment Bank Mandiri used when its own government stake moved to the Danantara sovereign holding vehicle (noted in the Q1 2025 post) - meaning BSI moved to a related party under common state control, not to an independent acquirer. The bank's stake in BSI didn't disappear; it now sits on the balance sheet as an equity investment instead of a consolidated subsidiary, and "Investments in shares - related parties" jumped from Rp762,254 million at December 2025 to Rp27,109,885 million at March 2026 as a direct result.
Against that backdrop, consolidated net income attributable to owners grew 16.57% YoY to Rp15,384,131 million (from Rp13,197,259 million in Q1 2025), with EPS up a matching 16.68% to Rp164.99 (from Rp141.40) - a headline that reads as a clean acceleration from FY2025's Q4 reversal. But net interest, sharia and insurance income - this backfill's revenue proxy - actually fell 2.30% YoY, to Rp25,336,704 million from Rp25,933,257 million, because two months of BSI's sharia income and expense lines are simply missing from this year's column. Almost every YoY comparison on a cost or income-statement line this quarter is comparing a ten-subsidiary consolidated group against a nine-subsidiary one for two-thirds of the period, not a true like-for-like trend - a genuine, disclosed comparability issue on its own (see Beyond the Usual), but one that should sit in front of every other number in this post. Bank Mandiri's own press release routes around the problem by leading with bank-only (parent-entity) figures throughout, which were never affected by BSI's consolidation status - this post follows the same convention wherever a bank-only number is available.
What Actually Drove the Profit Growth
Operating income (Profit from Operations) grew a real 9.98% YoY to Rp19,921,857 million (from Rp18,113,820 million) - a genuine acceleration, not just a deconsolidation artifact, because operating income sits below both the (BSI-affected) revenue line and the (BSI-affected) cost lines, and the net effect skewed favorable. The clearest single driver: impairment charges on financial assets fell 29.21% YoY, to Rp2,580,327 million from Rp3,645,087 million - a continuation of Q4 2025's ~60.1% YoY collapse (per the FY2025 post), at a slower pace of improvement, but still the same direction. Salary and employee-benefit expense fell 16.36% YoY (Rp6,003,219 million vs Rp7,177,736 million) and general/administrative expense fell 10.47% YoY (Rp5,428,699 million vs Rp6,063,746 million) - but unlike the impairment figure, both cost lines are directly distorted by BSI's absence this quarter (BSI ran its own payroll and admin cost base through February and March 2025's comparative), so neither can be read as organic cost discipline the way the impairment decline can. The one Q4-2025-flagged question this quarter answers cleanly: the cost-discipline reversal held into Q1 2026, at least on credit costs, even after adjusting for what BSI's departure can and can't explain.
Key Financial Metrics
Q1 2026 vs Q1 2025 (P&L, consolidated) - see the caveat above on BSI's deconsolidation before reading any YoY change below as a pure organic trend; balance sheet compares March 31, 2026 to March 31, 2025 (both consolidated) unless noted
FX: IDR 16,994.50 = USD 1 (March 31, 2026 close, per the bank's own filed exchange-rate disclosure); IDR 16,560.00 = USD 1 (March 31, 2025 close, per the same disclosure, consistent with the Q1 2025 post).
| Metric | Q1 2026 (IDR) | Q1 2026 (USD) | Q1 2025 (IDR) | YoY |
|---|---|---|---|---|
| Net interest, sharia & insurance income ("Net Revenue" equivalent) | Rp25,336,704M | ~$1.49B | Rp25,933,257M | ⚠️ -2.3% |
| Operating Income (Profit from Operations) | Rp19,921,857M | ~$1.17B | Rp18,113,820M | ✅ +10.0% |
| Net Income (attributable to owners) | Rp15,384,131M | ~$0.91B | Rp13,197,259M | ✅ +16.6% |
| EPS (basic, quarterly) | Rp164.99 | ~$0.0097 | Rp141.40 | ✅ +16.7% |
Operating cash flow swung to -Rp86,676,485 million for the quarter (from Q1 2025's +Rp20,242,157 million), against fixed- and intangible-asset capex of Rp570,851 million - a sharp reversal, but one directly explained by the BSI transaction: the cash-flow statement records a -Rp36,368,200 million "Impact of Loss of Control of Subsidiary" line this quarter, on top of the usual seasonal Q1 deposit outflow this backfill has tracked every year since Q1 2023. Total cash and cash equivalents at period-end is Rp255,015,412 million (~$15.00B), down from Rp379,877,189 million at the start of the period (December 31, 2025) - again, largely the BSI cash leaving the consolidated group, not an operating cash problem.
| Balance sheet metric | Mar 31, 2026 (IDR) | Mar 31, 2026 (USD) | Mar 31, 2025 (IDR) | YoY |
|---|---|---|---|---|
| Total Assets | Rp2,432,620,835M | ~$143.14B | Rp2,463,658,976M | ⚠️ -1.3% |
| Loans (gross, Kredit) | Rp1,568,084,875M | ~$92.27B | Rp1,625,286,579M | ⚠️ -3.5% |
| Total Deposits (customer deposits, consolidated) | Rp1,730,302,503M | ~$101.82B | Rp1,748,719,282M | ⚠️ -1.1% |
| Total Equity (attributable to owners) | Rp304,954,866M | ~$17.94B | N/A (Dec 2025: Rp293,750,669M) | +3.8% QoQ |
On a QoQ basis (against December 31, 2025, the balance sheet's actual filed comparative), the deconsolidation effect is far more dramatic: total assets fell 14.04% and gross loans fell 15.24% in a single quarter, purely from BSI's balance sheet leaving the group. The YoY columns above (March 2025, before BSI's departure, to March 2026, after it) understate that one-time step-change but better isolate whether the ongoing business actually shrank - and on that basis, the answer is closer to "essentially flat" than "shrinking."
Key Operational Metrics
Bank-only (Individual/parent-entity) ratios, computed from the parent-entity balance sheet and cross-checked against Bank Mandiri's own April 21, 2026 press release, which reports the same figures - the bank-only entity never consolidated BSI in the first place, so these ratios are unaffected by the deconsolidation above.
- Loan-to-Deposit Ratio (LDR) - bank-only, computed: 91.34% (Rp1,530,160,860M loans ÷ Rp1,675,220,398M deposits), up from December 2025's filed 88.92% (per the FY2025 post) - a reversal of last year's sharp full-year improvement, and consistent with this backfill's tracked pattern of a seasonal Q1 deposit outflow (see Q1 2024 post, Q1 2025 post) outpacing loan growth in the first quarter of the year.
- CASA» ratio - bank-only, computed: 71.71% (Rp1,201,246,038M demand + saving deposits ÷ Rp1,675,220,398M total deposits) - up from December 2025's 70.83% (per the FY2025 post), the first quarter-over-quarter CASA improvement this backfill has tracked since the metric began falling in mid-2025. See Beyond the Usual.
- CAR» (capital adequacy) - bank-only, filed: 19.69%, up from Q1 2025's 17.29% and from December 2025's 19.36% - a fifth straight quarter of sequential recovery from Q1 2025's crash. Consolidated CAR: 19.96%, up from Q1 2025's 18.50%.
- NPL» ratio - gross - bank-only, per the bank's press release: 0.98%, improved 3bps YoY, with an NPL Coverage Ratio of 245%.
- ROE» - after tax, per the bank's press release: 22.1%.
- BOPO», per the bank's press release: 58.0%, improved 3.48pp YoY.
- Loan growth (bank-only, per the bank's press release): +17.4% YoY to Rp1,530 trillion, ahead of the OJK-reported industry average of 9.37% YoY. Deposit growth (bank-only): +21.1% YoY to Rp1,675 trillion, ahead of the industry's 13.2%.
- Not available this quarter: a results presentation/analyst-meeting deck (not located as of this post; only the press release and financial statements) or an earnings call transcript.
Every Lending Segment Grew - So Did the Head Office's Loss
The quarterly segment note returned this quarter after being absent for three straight reporting periods (Q2 2025, Q3 2025, and FY2025, per the FY2025 post), and it answers the FY2025 post's open question directly: every one of the five core banking segments' profit grew YoY. Corporate Banking led at +65.9% (Rp4,191,382 million vs Rp2,526,099 million), followed by Government Institution +44.9% (Rp808,698 million vs Rp558,167 million), Treasury & International Banking +39.9% (Rp3,437,822 million vs Rp2,458,240 million), Commercial Banking +26.2% (Rp2,498,177 million vs Rp1,979,582 million), and Retail Banking +2.3% (Rp7,816,366 million vs Rp7,640,090 million) - the weakest of the five, but still positive, a reversal of the Q1 2025 post's finding that four of five segments had shrunk.
That's the good news. The Head Office segment's loss widened roughly 69.3% YoY, to Rp4,457,816 million from Rp2,633,480 million - continuing exactly the cost-reallocation pattern this backfill first documented in the FY2024 post and again in the Q1 2025 post: when the lending segments' combined profit grows, the Head Office segment's own allocated costs and unallocated items absorb a disproportionate share of the swing, understating how much of the improvement is genuinely broad-based lending strength versus a bookkeeping reallocation. The sharia subsidiary segment - now covering only BSI's operations through January 2026 before deconsolidation - fell 58.2% YoY (Rp786,021 million vs Rp1,878,796 million), the mechanical result of losing two months of the subsidiary's earnings, not a decline in BSI's own underlying performance. Insurance and other-subsidiary segments both grew modestly (+24.9% and +3.7% respectively).
Beyond the Usual
The BSI deconsolidation is disclosed plainly, but makes most YoY figures this quarter non-comparable without adjustment
Every income-statement and cash-flow line in this quarter's filing carries a footnote marker (*) pointing to Note 1g, which discloses that BSI's financial statements were deconsolidated from February 1, 2026. That's full, clean disclosure - not a hidden or misleading presentation - but it means a reader comparing this quarter's revenue, cost, or cash-flow lines against Q1 2025's without accounting for the missing two months of BSI's contribution is comparing two structurally different groups, not measuring the same business's growth. The balance sheet's QoQ comparison (against December 2025, which still included BSI for the full quarter) is affected even more sharply than the YoY comparison, since December 2025 carried BSI's full balance sheet while March 2026 carries none of it. This is worth tracking through the rest of 2026: once Q1 2027's comparative also excludes BSI, YoY comparisons become clean again, but every quarter of 2026 will carry this same asterisk.
The Head Office segment's widening loss offset almost the entire lending-segment profit gain
The five core lending segments' combined profit grew by roughly Rp3,590,267 million YoY this quarter (adding up the individual segment changes in the section above), but the Head Office segment's loss widened by Rp1,824,336 million over the same period - offsetting just over half (50.8%) of the lending segments' combined improvement before consolidated net income at the total-group level even reflects it. This is the same pattern the FY2024 and Q1 2025 posts both flagged: a genuine improvement in the operating businesses that the segment note itself shows is real, but one that's substantially muted at the headline level by how much cost gets parked in the Head Office/corporate-center segment rather than allocated back to the units that are actually driving the swing.
BSI moved to a related party, not to an outside buyer
Bank Mandiri's stake in BSI wasn't sold for cash to an independent acquirer - the equity note books the transaction as a "Difference in Transactions of Entities Under Common Control" (Rp111,953 million, added to additional paid-in capital), the identical accounting treatment used when Bank Mandiri's own government shareholding moved to the Danantara sovereign holding vehicle (noted in the Q1 2025 post). That accounting classification is only available when the transferring and receiving parties are both ultimately controlled by the same party - in this case, the Indonesian state - which means BSI's departure from Bank Mandiri's consolidated group most likely reflects a reorganisation of state-owned bank holdings under Danantara rather than an arm's-length divestment. The stake itself didn't disappear from Bank Mandiri's balance sheet: "Investments in shares - related parties" jumped from Rp762,254 million to Rp27,109,885 million in a single quarter, the accounting footprint of an equity-method stake replacing what used to be full consolidation.
Bank-only CASA rose for the first time in over a year
Every quarter from mid-2025 through the FY2025 post tracked bank-only CASA falling - from September 2025's 73.13% to December 2025's 70.83%, a run the FY2025 post explicitly flagged as "not yet resolved" even as the profit story improved. This quarter reverses that: bank-only CASA rose to 71.71%, up 0.88 percentage points from December 2025. It's a single data point, not a confirmed trend reversal, and it sits within a quarter where bank-only LDR moved the other direction (up to 91.34% from 88.92%, per above) on the usual seasonal Q1 deposit outflow - so it would be premature to call the funding-cost story resolved. But it is the first quarter this backfill has recorded where CASA moved up rather than down, worth watching for whether it holds into Q2.
Bank Mandiri Shares Fell Further as the Rupiah Kept Weakening
The FY2025 post closed at Rp5,100 on December 30, 2025, a 15.9% Q4 rebound from September's Rp4,400 low. That rebound partially reversed this quarter: shares fell to Rp4,820 at end-January, recovered to Rp5,275 at end-February, then closed the quarter at Rp4,720 on March 31, 2026 - a -7.5% net move for the quarter and a -9.2% decline from March 27, 2025's Rp5,200 close. Zooming out to the two-year window ending this quarter (April 2024 through March 2026), shares peaked at Rp7,125 in August 2024 and troughed at Rp4,400 in September 2025 - a 38.2% peak-to-trough decline, comfortably clearing this playbook's threshold for a dedicated price section, and shares remain well below that August 2024 peak even after Q4 2025's rebound. 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 Rp16,994.50/USD at March 31, 2026 - continuing the multi-quarter depreciation trend this blog has tracked since 2024.
Coverage at a Glance
| Metric | Q1 2026 | Q1 2025 | YoY | Why it matters |
|---|---|---|---|---|
| Net income (attributable to owners) | Rp15.38T | Rp13.20T | +16.6% | Real growth, but sits atop a structural break - see below |
| Net interest, sharia & insurance income | Rp25.34T | Rp25.93T | -2.3% | Revenue proxy fell only because two months of BSI's contribution are missing this year |
| Impairment charges | Rp2.58T | Rp3.65T | -29.2% | Continues Q4 2025's ~60% collapse, at a slower pace - the one genuinely like-for-like cost line |
| Bank-only CASA | 71.71% | N/A (Sep-25: 73.13%) | +0.88pp QoQ | First quarterly rise this backfill has tracked, after five straight quarters of decline |
| Bank-only LDR | 91.34% | ~94.18% (computed) | -2.84pp | Rose QoQ from Dec-2025's 88.92% on the usual seasonal Q1 deposit outflow |
| Bank-only CAR | 19.69% | 17.29% | +2.40pp | Fifth straight quarter of sequential recovery from Q1 2025's crash |
| Head Office segment loss | Rp4.46T | Rp2.63T | +69.3% | Offset ~51% of the five lending segments' combined profit gain |
| Share price (Mar 31, 2026 close) | Rp4,720 | Rp5,200 | -9.2% | Reversed part of Q4 2025's rebound; still well below Aug-2024's Rp7,125 peak |
Target Valuation Range
Market cap Rp440.5 trillion (~$25.92B), ~1.44x P/B, ~7.5x P/E. Bottom line: fairly valued to modestly cheap on trailing multiples, with the caveat that this quarter's earnings quality is genuinely harder to read than usual - real credit-cost improvement and broad-based lending-segment growth, offset by a Head Office allocation swing and a structural deconsolidation that makes next quarter's comparisons the real test of whether the underlying business (ex-BSI) is actually growing.
Using trailing-twelve-month net income attributable to owners of approximately Rp58,480,822 million (FY2025's Rp56,293,950 million less Q1 2025's Rp13,197,259 million plus Q1 2026's Rp15,384,131 million), implied TTM EPS of approximately Rp626.58, and the post-split share count of 93,333,333,332:
| Market cap → book value | Q1 2026 |
|---|---|
| Share price (period-end) | Rp4,720 |
| Shares outstanding | 93,333,333,332 |
| Market capitalization | Rp440.5 trillion (~$25.92B) |
| Total equity (book value) | Rp304,954,866M |
| Book value per share | Rp3,267.37 |
| Peer-multiple sanity check | Q4 2025 | Q1 2026 | Change |
|---|---|---|---|
| P/B | 1.62x | 1.44x | down |
| P/E | 8.5x | 7.5x | down |
P/E: ~7.5x, using TTM EPS of ~Rp626.58 against the Rp4,720 close on March 31, 2026. P/B: ~1.44x, using book value per share of approximately Rp3,267.37 (Rp304,954,866M total equity attributable to owners ÷ 93,333,333,332 shares). Market capitalization: approximately Rp440.5 trillion (~$25.92B). That's cheaper on both multiples than the FY2025 post's ~8.5x P/E and ~1.62x P/B - the share price fell further this quarter even as TTM earnings kept growing, so both multiples compressed together. The open questions carried forward from the FY2025 post split into one resolved and one still open: the segment note's return confirms the Q4 2025 cost-discipline reversal was broad-based across every core lending segment, not concentrated in one - but the Head Office offset means a reader still can't take the consolidated headline number as a clean read of how much the operating businesses actually improved. And whether bank-only CASA's first quarterly rise in over a year is a real funding-cost stabilization or a one-quarter blip won't be confirmed until at least Q2 2026 - the same "confirm it's structural, not a one-quarter favor" test this blog has now applied to Mandiri's cost base for two straight quarters running.
PT Bank Mandiri (Persero) Tbk's reviewed interim consolidated financial statements as of and for the three-month period ended March 31, 2026 (with comparatives for March 31, 2025 and December 31, 2025), and the bank's April 21, 2026 press release, via Bank Mandiri's investor relations page.