The idea, from first principles. Lay every rupee the company will receive and every rupee it must pay on a single time axis, bucket by bucket. Loans throw off EMIs and maturities on a schedule. Borrowings demand interest and principal on a schedule too. If contracted inflows plus on-hand liquidity cover contracted outflows in every early bucket, the company's survival is arithmetic. If an early bucket shows outflows exceeding inflows, a negative gap, then survival in that bucket depends on new money: rolling the CP, refinancing the NCD, drawing a line, selling an asset. So a negative short-bucket gap measures how much of the company's continued existence has been outsourced to other people's future decisions. Everything else about ALM follows from that. Banks live with permanent negative short gaps because deposits behaviorally stay and RBI backstops them. An NBFC has neither grace, so you read its ALM statement literally.
Advertisement
What does RBI require in the ALM statement?
The regulatory format. RBI's liquidity-risk framework for NBFCs came out in November 2019, a direct product of the IL&FS autopsy. It requires a structural liquidity statement that slots all cash flows into buckets: 1 to 7 days, 8 to 14 days, 15 to 30 days, then monthly to 3 months, 3 to 6 months, 6 to 12 months, 1 to 3 years, 3 to 5 years, over 5 years. There are two hard rules to memorize. First, cumulative negative mismatches in the first three buckets (1 to 7 days, 8 to 14 days, 15 to 30 days) may not exceed 10%, 10% and 20% of cumulative outflows in those buckets. Second, larger NBFCs must also hold a liquidity coverage ratio (LCR), meaning high-quality liquid assets against 30-day stressed net outflows, phased up to 100%. It has been fully phased since December 2024 for the covered size bands. Verify applicability in the current Master Direction. You'll find the public disclosure in the annual report's maturity-pattern note and, more usably, in rating rationales. Beyond the first month RBI prescribes board-approved internal limits instead of hard caps, so you set the threshold. Use a simple one. A lender whose cumulative gap is still negative at 12 months, with no committed lines covering the hole, is running on rollover faith.
How do you build an ALM gap analysis?
Build one. Tapti Housing Finance (synthetic, all figures ₹ crore) is a stylized pre-2019-style HFC. Its loan book is ₹40,000. Individual home loans make up 28,000 of it, contractually 15 to 20 years but behaviorally ~7 to 9 because of prepayment. Developer and construction-finance loans make up the other 12,000, with bullet repayments "scheduled" out of project refinancings. Borrowings are ₹36,000: CP 5,400 (15%), NCDs 21,600 (60%), bank term loans 7,200 (20%), NHB refinance 1,800 (5%). Cash and liquid investments come to 1,000, fixed/other assets 340, other liabilities (accrued expenses) 840, and net worth 4,500. That puts gearing (borrowings ÷ net worth) at 8.0×, normal-looking for an HFC. The balance sheet foots: 1,000 + 40,000 + 340 = 36,000 + 840 + 4,500 = 41,340. ✓
Here's the structural liquidity statement on a contractual basis, with the first bucket aggregated to a month. The regulatory statement splits that bucket into 1-7/8-14/15-30 days, and as you'll see, the aggregation flatters:
| ₹ cr | ≤1m | 1–3m | 3–6m | 6–12m | 1–3y | 3–5y | >5y | Total |
|---|---|---|---|---|---|---|---|---|
| Inflows | ||||||||
| Cash + liquid investments | 1,000 | — | — | — | — | — | — | 1,000 |
| Individual home-loan receipts | 420 | 840 | 1,260 | 2,520 | 8,000 | 5,500 | 9,460 | 28,000 |
| Developer-loan receipts | 0 | 100 | 500 | 1,400 | 5,000 | 3,500 | 1,500 | 12,000 |
| Total inflows | 1,420 | 940 | 1,760 | 3,920 | 13,000 | 9,000 | 10,960 | 41,000 |
| Outflows | ||||||||
| Commercial paper | 1,800 | 2,700 | 900 | — | — | — | — | 5,400 |
| NCD maturities | 300 | 900 | 1,500 | 2,900 | 8,000 | 5,000 | 3,000 | 21,600 |
| Bank term-loan repayments | 100 | 200 | 300 | 600 | 2,400 | 2,400 | 1,200 | 7,200 |
| NHB refinance | — | — | — | 200 | 700 | 600 | 300 | 1,800 |
| Opex & other payables | 70 | 140 | 210 | 420 | — | — | — | 840 |
| Total outflows | 2,270 | 3,940 | 2,910 | 4,120 | 11,100 | 8,000 | 4,500 | 36,840 |
| Gap (in − out) | −850 | −3,000 | −1,150 | −200 | +1,900 | +1,000 | +6,460 | +4,160 |
| Cumulative gap | −850 | −3,850 | −5,000 | −5,200 | −3,300 | −2,300 | +4,160 | |
| Cumulative outflows | 2,270 | 6,210 | 9,120 | 13,240 | 24,340 | 32,340 | 36,840 | |
| Cum. gap ÷ cum. outflows | −37% | −62% | −55% | −39% | −14% | −7% | +11% |
How do you read the gap table?
Three reading habits come before the verdict:
1. Check the plumbing. The final cumulative gap must tie: total bucketed inflows − outflows = 41,000 − 36,840 = +4,160 = net worth (4,500) − unbucketed fixed assets (340). ✓ If a published statement doesn't roughly foot to net worth, something's been left out, usually on the outflow side. Undrawn sanctioned loans the NBFC has committed to disburse are an outflow too, and the best disclosures include them.
2. Read the shape of the whole table. Assets are long, with 60%+ of mortgage receipts beyond 3 years. Liabilities are front-loaded, with CP entirely inside 6 months and half the NCDs inside 3 years. So the gap is negative early and positive late. You're looking at borrowing short to lend long, drawn in numbers. The −850 in month 1 is 37% of the month's outflows, against a 10 to 20% regulatory tolerance, and the granular 1 to 7 day sub-bucket would look worse because CP maturities cluster there. Tapti's twelve-month cumulative hole is −5,200, so it must raise ₹5,200 crore of new money in a year merely to stand still. Its published statement almost certainly shows small, compliant gaps, because management is allowed to assume CP rollover and to place behavioral (prepayment-accelerated) asset inflows. Your job is to re-derive the table with CP rollover at zero and the behavioral generosity stripped out. We built the table above that way, because it's the one that operates in a freeze.
3. Then, and only then, count the mitigants. Start with undrawn committed bank lines. Tapti has ₹800 of them, and you should read the fine print, because "sanctioned but uncommitted" lines evaporate just when they're needed. On-hand HQLA is already counted. Last comes the unencumbered retail book eligible for securitization. Tapti has ~₹22,000 of unpledged individual mortgages, of which perhaps ₹9,000 is priority-sector-eligible and really saleable to banks inside a quarter. Mitigants turn a death sentence into a countdown. They buy months. They don't buy immunity.
What happens when the funding market freezes?
Now run the freeze: the IL&FS/DHFL mechanism in miniature. Suppose the CP market shuts today. Nothing Tapti did caused it. Someone else defaulted. In month 1, 1,800 of CP matures and can't roll, so Tapti uses its 1,000 of cash and draws its 800 of lines. It survives, at zero buffer. Months 2 to 3 bring another 3,000 net hole, and the only source left is selling loan pools. Tapti sells its cleanest retail mortgages, and because the good assets go first, the balance sheet's average quality gets worse as it shrinks. It also halts new disbursements to conserve cash.
That halt sets off two delayed detonations. The retail franchise it took twenty years to build (DSA networks, builder relationships) disperses in a quarter. Meanwhile the developer borrowers, whose projects depended on Tapti's next construction-finance tranche, stall. The "scheduled" 1,400 of developer receipts in the 6 to 12 month bucket turns into zero, which deepens the very hole the sales were filling. By month 6 to 9 the company is liquidating its earning assets to repay its liabilities. It's a melting ice cube, technically solvent, functionally finished, with GNPA on the frozen developer book about to destroy the solvency too. Illiquidity arrives first. Insolvency is the paperwork. DHFL ran this same sequence for nine months, September 2018 to June 2019, before its default was declared. Its equity was gone long before the default was.
What does safer funding cost?
The counterfactual that completes the lesson. Now re-fund the same asset book safely. Replace the 5,400 of CP with 3- and 5-year NCDs plus NHB refinance, and hold a liquidity buffer of ~1,500 against the first quarter's outflows. With that, the early buckets turn positive, and Practice 9 makes you do that arithmetic yourself. What does safety cost? A 5-year NCD prices ~50 bp over CP, and the buffer negative-carries ~200 bp, because you borrow at 8.5% and park the money in treasuries at 6.5%. Together that's 5,400 × 0.5% + 1,500 × 2.0% = ₹57 cr pre-tax and ₹42.75 cr after tax. On Tapti's ₹41,340 cr of assets that's 10.3 bp of ROA, and on its ₹4,500 cr of net worth it's 0.95 points of ROE. You'll find both reproduced in the Verify block below. So here's what the 2018 crisis came down to. Several large lenders sold their survival for about one point of ROE, and the market paid them premium multiples for it until the bill arrived. When you see an NBFC out-earning its peers on ROE with the same assets, check the ALM note before applauding. Usually that extra return is rented from the liability side.