Issuer Flash: PT Bank Mandiri (Persero) Tbk

Report date: 2026-10-27 Event date: 2026-10-23 Event title: Q3/9M2026 Results Research cutoff: 2026-10-26

1. Flash Conclusion

Bank Mandiri's September results leave the existing stable credit view intact, but the forward indicators are less reassuring than headline earnings. The government-controlled universal bank's deposit and lending franchise continues to generate earnings for loss absorption: consolidated nine-month attributable profit rose 20.7%, bank-only capital recovered from June, and aggregate liquidity ratios remained above 100%. However, higher funding costs compressed quarterly margins, while performing restructurings lifted Loans at Risk (LaR). Those developments warrant closer monitoring rather than a more positive credit assessment.

For senior creditors, the immediate earnings, capital and liquidity position remains supportive. The risk is that repricing and cost discipline prove insufficient if restructured borrowers weaken. Government control and systemic importance support confidence, but do not establish a legal guarantee or identical protection for senior and regulatory capital securities.

2. Earnings Growth Masks the Higher Price of Funding

The October 23 presentation reports consolidated nine-month profit after tax and minority interests of IDR45.5tn. Pre-provision operating profit increased 17.5%, helped by 9.4% operating-income growth and a 1.6% reduction in operating expenses. Provision expenses were 5.8% lower year on year. These results support internal loss absorption, but a quarterly reading is more restrained: net interest income fell 3.6% from Q2 as interest expense increased 20.7%, faster than the 5.1% increase in interest income. Quarterly attributable profit was almost unchanged, rising 0.6%.

Bank-only quarterly loan yield improved to 7.03% from 6.91%, yet deposit cost rose to 2.34% from 2.04%. Bank-only quarterly NIM consequently fell to 3.87% from 4.21%. Management's repricing progress is therefore visible, but it has not yet offset the higher funding burden. Monthly deposit cost was 2.36% in both July and September, with a temporary decline in August; the evidence is insufficient to call a sustained funding-cost reversal.

Source: official presentation PDF pp.19-22, 28, 32 and 36. Consolidated data exclude Bank Syariah Indonesia (BSI); historical consolidated comparisons are pro forma where stated. Bank-only NIM uses average daily balances, while consolidated NIM uses average monthly balances. Quarterly and cumulative margins should not be conflated.

Funding volumes provide some relief. Bank-only loans grew 0.89% during Q3 to IDR1,605.8tn, slower than the 2.35% deposit increase to IDR1,750.2tn, and bank-only LDR fell to 91.2% from 92.7%. CASA balances increased 4.4% sequentially. Nevertheless, bank-only wholesale funding of IDR313.0tn remained 44.2% above September 2025, despite a small quarterly decline. Better deposit growth and a lower LDR do not, by themselves, eliminate funding-price or refinancing sensitivity.

The presentation raises full-year consolidated loan-growth guidance to 9-11% from 7-9%, while retaining NIM guidance of 4.3-4.5% and credit cost of 0.6-0.8%. These are management targets, not achieved outcomes. Further expansion must be judged against funding cost, the quality of new lending and capital consumption.

3. Performing Restructurings Matter More Than the Flat NPL Ratio

Consolidated LaR rose by 93bp during Q3, principally because the current, restructured component increased to 2.83% of loans from 1.89%. The special-mention ratio was broadly unchanged and gross NPLs remained low. Performing restructurings are not equivalent to defaults, but they indicate that a larger pool requires modified repayment terms and monitoring. NPL coverage remained 235%, while broader LaR coverage fell to 35.1% from 40.5%. A substantial headline NPL reserve ratio should therefore not be treated as protection against every contingent loss in the restructured book.

The newly available detail partially resolves earlier monitoring gaps. Bank-only stage 2 loans represented 5.80% of lending, and reported corporate and commercial stage 2 shares were 6.31% and 6.60%. These accounting categories are distinct from regulatory NPL classifications. Bank-only segment NPL ratios also remained uneven: 0.25% in corporate, 0.62% in commercial, 3.19% in micro/payroll and 3.17% in consumer. The figures help locate risk, but neither a single-period stage distribution nor a low wholesale NPL ratio establishes future loss sensitivity. Consolidated annualized credit cost of 0.63% remained within guidance.

Capital and liquidity give room to manage those risks. Bank-only CAR of 18.4% exceeded the presentation's 14.68% baseline requirement and lay within its 18-20% internal target, although it remained below 19.4% at end-2025. RWA increased to IDR1,450tn from IDR1,399tn in June, so the quarterly recovery is constructive despite continued capital consumption. Aggregate LCR and NSFR also improved modestly. These buffers support the existing assessment; they do not establish currency-specific liquidity or resolve instrument-level loss-absorption terms.

4. What To Watch Next

The next test is whether improving loan yield and CASA volumes can arrest NIM pressure without accelerating risk-taking. Monitor quarterly deposit and wholesale funding costs, LDR, margin and the execution of higher loan-growth guidance together. For credit quality, follow migrations from performing restructurings and stage 2 into NPLs, segment credit costs and LaR coverage, rather than relying on the aggregate NPL ratio alone. Capital generation, RWA growth and future distributions should be tested jointly against the internal CAR target.

The results provide a limited, source-based flash check of the May 30 SSC discussion: current funding, segment/stage, capital and aggregate liquidity figures are now available. Its stress-sequence hypotheses and quantitative capital sensitivities are not thereby verified. The presentation also reports that 99.9% of bank-only FX loans were performing under its category 1-and-2 definition, but this does not demonstrate borrowers' natural hedges or currency-level funding resilience. FX maturity ladders, liquidity by currency, debt-class support and bond terms remain next-confirmation items; the discussion remains outstanding for summary-scope verification. Bank-only Tier 1 is shown as 17.2% in the ratios table and 17.3% in the capital chart, a small inconsistency requiring reconciliation before precise capital-instrument analysis. Live spreads and relative value were not assessed.

5. Sources

• Bank Mandiri, 3Q 2026 Results Presentation, Jakarta October 23, 2026, official presentation. Cover p.1 establishes the event date; pp.11, 13, 16-17 cover funding, margins and guidance; pp.19-22 cover financial and liquidity ratios; pp.28, 32-36 cover definitions, asset quality and capital; p.80 covers FX loans. Retrieved October 27; no facts first published after the October 26 research cutoff were used.
• Existing May 7 issuer summary and July 28 H1 results flash, used for the existing credit assessment and prior monitoring questions, not as September figures.
• May 30 additional discussion, used to define the limited monitoring tests above; discussion assertions are not independent evidence.