Legacy economics versus new credit risk
AUM ₹469,767cr; H1 ROE 13.25%; new project lending changes risk.
Watch: WATCH; demand realized risk-adjusted return.
Home / Research / Indian Railway Finance Corporation Limited
Legacy economics versus new credit risk
https://nsearchives.nseindia.com/corporate/IRFC_15102026153717_outcomeofmeeting15102026.pdf
https://nsearchives.nseindia.com/corporate/IRFC_15102026153717_outcomeofmeeting15102026.pdf
https://nsearchives.nseindia.com/corporate/IRFC_15102026153717_outcomeofmeeting15102026.pdf
https://nsearchives.nseindia.com/corporate/IRFC_15102026153717_outcomeofmeeting15102026.pdf
AUM ₹469,767cr; H1 ROE 13.25%; new project lending changes risk.
Watch: WATCH; demand realized risk-adjusted return.
Q2 ₹7,866cr, +23% YoY but down QoQ.
Watch: Analyze spread and funding.
PBT equals PAT on statutory basis; H1 PAT +10.3%.
Watch: Track net margin and provision.
H1 revenue +21.4%, PAT +10.3%.
Watch: Normalize finance cost and mix.
SAGE PRO Institutional Equity Research | 16 November 2026 | NSE: IRFC | Q2/H1 FY2027, ended 30 September 2026
Stance: WATCH — relatively predictable legacy earnings, uncertain diversification premium. Horizon 12–36 months. Reference close ₹83.30, 13 November 2026 (Sage PRO canonical price). All amounts ₹ crore except per-share metrics. IRFC has no subsidiary/associate/JV at September 2026, so statutory results are standalone. Original issuer deck, call and signed statutory exchange filing underpin this report. Available-source longform; full project-level credit and independently matched peer valuation QA remain open.
Key takeaways
Its advantage has been funding access, institutional linkage, asset-liability matching and low operating expense.
The current strategic pivot is to expand beyond its single-client legacy into railway-linked and other infrastructure financing. The opportunity is a larger addressable market and higher spreads. The risk is that IRFC must acquire capabilities in project appraisal, borrower risk, collateral, construction execution, covenants and portfolio concentration that differ from the original MoR model.
A nil historical GNPA record is not a sufficient reason to assign a low loss rate to every newly sanctioned project.
At September, the investor presentation reports AUM ₹469,767.03 crore, net worth ₹60,573.35 crore, H1 annualized ROE 13.25%, ROA 1.49% and NIM 1.50%. 92.36% of AUM is exposure to the MoR, while loans to companies are 7.59%. [A, slides 2–4] The legacy dominates closing AUM, even though new disbursements are shifting.
Our variant hypothesis is that the next phase can raise marginal returns without destroying the original risk profile only if the new portfolio earns sufficient spread after expected loss and capital costs. The market should not extrapolate the sovereign-like credit economics of the existing stock to non-MoR originations.
Counter-thesis: a large project pipeline may generate slower actual drawdowns, lower net spread and greater provision expense than headline sanctioned amounts suggest; the stock could be valued at a growth multiple unsupported by the still-low group return. Maintain a watch stance and require actual drawdowns, cash interest and project-risk disclosures rather than forecast railway capex alone.
| Statutory metric | Q2 FY26 | Q1 FY27 | Q2 FY27 | Change |
|---|---|---|---|---|
| Revenue from operations | 6,371.89 | 8,261.11 | 7,866.41 | +23.45% YoY; −4.78% QoQ |
| PBT/PAT | 1,776.98 | 1,927.21 | 1,958.48 | +10.21% YoY; +1.62% QoQ |
| Basic EPS (₹) | 1.36 | 1.47 | 1.50 | +10.3% YoY |
| Net worth | 56,193.85 | 58,791.95 | 60,573.35 | +7.79% YoY |
| Outstanding debt | 407,613.64 | 437,028.79 | 431,482.83 | −1.27% QoQ |
| Debt/equity | 7.25x | 7.43x | 7.12x | Lower sequential leverage |
Key takeaways
Equity face value ₹10, paid-up equity share capital ₹13,068.51 crore; derived shares ~1,306.851 crore. These exact figures allow independent per-share arithmetic, unlike an inference from market-cap data.
Revenue growth outpaced earnings growth significantly. The Q2 revenue/operation decline sequentially but small PAT increase demonstrates why gross revenue is a poor proxy for lender earnings. Financing charges, interest income, lease economics, asset origination and provision timing determine the net spread.
PAT/revenue was ~24.9% in Q2 versus ~27.9% a year earlier; this is not a net interest margin. Use net interest income and average interest-earning assets rather than this revenue ratio to forecast financial profitability.
Statutory Q2 comprehensive income ₹1,781.40 crore versus net profit ₹1,958.48 crore implies OCI −177.08 crore in the quarter. This matters for book-value growth and may reflect financial-instrument or other OCI movements; the specific composition is not independently reconciled here. Do not attribute it to a specific hedge or bond position without the audited notes. Equity value changes need to be measured using both PAT/distribution and OCI.
The issuer's AUM chart shows FY26 end ₹466,148.28 crore and H1 FY27 ₹469,767.03 crore. Closing AUM is only ~0.8% above this year-end figure, despite H1 disbursements ₹32,337.77 crore. AUM stock equals prior AUM plus drawdowns less repayments and other movements. A high disbursement number can therefore coexist with modest AUM growth if older leases run down. The most important operating bridge is net AUM creation and mix, not merely gross sanctions. [A, slide 4]
The presentation's September AUM breakdown shows rolling-stock lease receivables (including NTPC BOBR rakes) 26.70%, project-asset lease receivables 51.48%, advances against railway infrastructure assets to be leased 14.23%, and loans to companies 7.59%. These sum to 100%. MoR concentration is therefore an advantage from a credit-history and institutional-access perspective but a concentration risk from a counterparty, fiscal-policy and future business-flow perspective.
Key takeaways
The opportunity is a stable large earning asset base and low operating cost. The limiting factor is that the lender cannot indefinitely increase returns by expanding a fixed-spread, low-risk model unless asset volumes or capital structure change.
Management reports FY26 disbursements about ₹35,000 crore, largely non-Indian-Railways disbursement, and expects FY27 disbursements to exceed that amount. H1 FY27 disbursements ₹32,337.77 crore already approach the prior full-year level; the issuer also discusses prospective metro, port, renewable/thermal power and road-linkage financing.
[A,B] A pipeline of potential railway infrastructure projects may span several years and involve bilateral/multilateral funding or joint financing. Do not assume that total announced railway project cost becomes IRFC's lending opportunity or immediate assets.
The next underwriting problem is project drawdown timing. Large infrastructure assets can require milestone-based disbursements; undrawn sanctions do not earn full coupon. Construction delays, permits, traffic assumptions, counterparty payment behavior, government budget allocations and refinancing can affect realized project return. Project lending can also concentrate multiple exposures in correlated state agencies, contractors or infrastructure sectors.
A useful investment model separates legacy AUM runoff, signed project drawdowns, prepayment and fee income. For each new project, compute expected risk-adjusted spread after funding, operating cost, credit/provision charge and allocated equity.
A 50bp incremental spread on ₹10,000 crore of actual average deployed assets is ₹50 crore annual gross spread before costs/loss; a 1% unexpected principal loss on the same asset amount is ₹100 crore. This asymmetry illustrates why the risk regime matters despite the legacy's zero NPA record.
Key takeaways
Staffing alone does not establish a seasoned credit record.
Management referred to new-business provisioning on greenfield and brownfield projects (1% and 0.4% respectively in its discussion), which affected the previous year's fourth-quarter profit. These are management's explanation of regulatory provisioning for relevant exposures; exact regulatory classification and the specific provision schedule must be read from original applicable rules and statutory notes, rather than generalized to every project.
Provision expense can arrive before observable credit loss. Higher-yield loans may therefore raise nominal NIM while depressing near-term PAT or changing capital needs.
IRFC's 102.56% reported capital adequacy ratio is unusually high because eligible legacy risk-weighted assets and regulatory treatments differ from ordinary NBFC portfolios. It must not be treated as total equity/assets or a comparable private retail NBFC ratio.
The issuer states exemptions to relevant prudential norms for direct MoR exposures. On 15 October the board approved a policy for transition to the NBFC Upper Layer regulatory framework. [A,C] This is significant: regulatory intensity, risk reporting, provisioning and capital requirements may evolve as business composition changes.
Verify actual transition and compliance milestones; do not assume exemptions automatically extend to new borrowers.
Debt/equity of 7.12x is a useful counterweight to the extraordinary CRAR figure. With ₹431,482.83 crore outstanding debt and equity ₹60,573.35 crore, the balance sheet is materially leveraged. Nil current GNPA does not mean nil earnings sensitivity to funding, liquidity or new project loss.
A loss of just 10bp on closing AUM represents ~₹470 crore before tax—about 6% of Q2 annualized PAT. This is illustrative and not an expected-loss forecast; the actual non-MoR risk book is much smaller and portfolio classification matters.
Key takeaways
A lender with a narrow margin is sensitive to imperfect repricing, even when absolute credit losses are small.
The group has no subsidiary structure at the reporting date, and reported H1 ROA 1.49%, ROE 13.25%. The statutory net worth increased ~₹1,781 crore sequentially, equal to the Q2 comprehensive income before the effect of any equity movements in that comparison; reconcile all distribution dates before assuming retention.
The October interim dividend ₹1.15 per share (record 23 October) is an actual declared distribution, not a prospective entitlement for a November buyer. At the reference price it is 1.38% of price for that specific distribution, not annual yield.
Capital deployment must compare the return on new loans with equity hurdle after risk weights. The historic model's low-cost funding and tax profile support profitability; the valuation question is whether new business retains these advantages after provisioning and compliance. The issuer reports no tax liability in the presented periods. Do not forecast indefinite tax exemption or assume the same tax treatment for every future activity without current legal/tax evidence.
A 10bp annualized net margin move applied illustratively to closing AUM gives ~₹470 crore annual pre-tax sensitivity. Group NIM 1.50% implies a narrow buffer compared with retail NBFCs, so margin protection and duration are key.
If new business has materially greater risk than legacy leases, a higher spread is economically required rather than a free upside. Monitor effective borrowing yields, currency hedging, maturity gaps and counterparty exposures, and reconcile the debt/lease cash-flow schedule before declaring low risk.
The original presentation gives opening FY27 AUM of ₹466,148.28 crore, September closing AUM of ₹469,767.03 crore and H1 gross disbursements of ₹32,337.77 crore. Net stock creation was only ₹3,618.75 crore.
The implied balancing movement of ₹28,719.02 crore represents gross disbursements less net AUM growth, or approximately 88.8% of gross flows. This is a residual that can include repayments, lease run-off and other portfolio movements; it is not an independently verified cash repayment total.
The distinction changes how to evaluate management's October call guidance to exceed last year's approximately ₹35,000 crore disbursement. H1 has already achieved more than 92% of that reference, so merely passing the gross annual target would say little about net asset growth.
For an illustrative 5% increase in FY27 closing AUM from the March base, required year-end stock is ₹489,455.69 crore. That would require another ₹19,688.66 crore net stock addition in H2. If H2 residual run-off and other movements were similar to H1's ₹28,719.02 crore, implied H2 gross disbursements would need to be about ₹48,407.68 crore and full-year gross flows approximately ₹80,745 crore.
These are analyst arithmetic, not a company target or forecast. Vary the actual run-off and sanctioned-project drawdowns before using this as an earnings estimate. The new project pipeline may have a long construction schedule and a different pace of repayment from maturing railway leases.
The non-MoR loans-to-companies bucket at 7.59% of September AUM implies approximately ₹35,655 crore of exposure using the rounded presentation mix. A purely illustrative 1% principal-loss or provision stress on this bucket is about ₹357 crore before tax, roughly 4.6% of annualized H1 PAT.
This does not imply every loan is greenfield, that all company exposures have the same risk, or that the statutory provision must be 1%. It demonstrates why group-wide nil GNPA and high legacy capital adequacy cannot replace a project-by-project expected-loss assessment.
Higher spreads must compensate for the new risks after allocated capital and funding.
Key takeaways
Annualized Q2 EPS ₹6.00 implies ~13.9x. The price is paying a premium to statutory book and requires sustained returns or higher growth.
Illustrative next-year earnings model: start from H1 annualized PAT ₹7,771.38 crore. Bear/base/bull assume growth 0%/10%/20% in normalized PAT, with no modeled change in shares. Earnings become ~7,771/8,549/9,326 crore, EPS ~5.95/6.54/7.14. These assumptions are not guidance; model omits incremental OCI, tax-law changes, new equity and unexpected project losses.
| Scenario | PAT ₹cr | EPS ₹ | P/E at ₹83.30 |
|---|---|---|---|
| Bear | 7,771 | 5.95 | 14.0x |
| Base | 8,549 | 6.54 | 12.7x |
| Bull | 9,326 | 7.14 | 11.7x |
A terminal residual-income framework using sustainable ROE 11%/13%/15%, cost of equity 13%/12%/11.5% and long-run growth 4%/5%/6% yields justified P/B ~0.78x/1.14x/1.64x respectively. This is not a target or total valuation; it excludes an explicit high-growth period and uses simplified clean-surplus assumptions.
At ~1.8x current dated book, investors need a credible multi-year economic-profit phase or lower required return. A low-risk bond-like legacy is valuable, but low risk does not justify arbitrary multiple expansion.
Relevant peers PFC and REC also lend to infrastructure and have government linkage, but exposure type, borrower risk, legacy arrangements, NIM, regulatory treatment and project cycle differ. Compare normalized attributable ROE, asset quality by comparable definition, duration and cost of equity, not raw P/E alone. A full same-date verified peer valuation and project-level expected-loss model remain pending; no target price issued.
| Observable test | September evidence | What to watch |
|---|---|---|
| Net asset growth | AUM ₹469,767cr; H1 disbursements ₹32,338cr | Actual drawdowns less lease repayments, not sanctions |
| Diversification | ~7.59% loans to companies; legacy still dominates | Project mix, concentration, realized risk-adjusted spread |
| Credit discipline | Nil reported GNPA | New-project stage migration, expected loss and provisioning |
| Funding | ₹431,483cr debt; 7.12x gearing; 1.50% H1 NIM | Maturity/repricing, cost of debt, hedge and funding access |
| Capital | ₹60,573cr net worth; upper-layer transition policy | Regulatory implementation, risk-weighted capital, per-share returns |
| Valuation | ~1.80x September book; ~14x annualized H1 EPS | Sustainable ROE above hurdle, not volume alone |
Key takeaways
https://nsearchives.nseindia.com/corporate/IRFC_15102026153717_outcomeofmeeting15102026.pdf
This is dated research, not a live market quote. Scenarios and historical performance are not guarantees.