A Non-Banking Financial Company is a company registered under the Companies Act whose principal business is lending or investing, registered with RBI under the RBI Act. "Principal business" has a precise test, the 50-50 test: financial assets must exceed 50% of total assets, and income from financial assets must exceed 50% of gross income. Pass both and you must register with RBI (with carve-outs for entities regulated elsewhere: insurers, brokers, housing finance companies historically under the National Housing Bank, transferred to RBI regulation in August 2019).
How is an NBFC different from a bank?
What an NBFC is not matters more. It can't accept demand deposits, so there are no current accounts and no savings accounts. A small, closed, shrinking club of deposit-taking NBFCs and HFCs may accept term deposits, capped at 1.5× net owned funds (check the current direction). No new deposit licences have been granted in decades, and DICGC doesn't insure these deposits. An NBFC has no CRR or SLR and no branch-licensing privileges. Then comes the absence that decides everything: no standing access to RBI as lender of last resort. When a solvent bank faces a liquidity squeeze, it can repo its government bonds with RBI at the marginal standing facility and live to fight another day. When a solvent NBFC faces the same squeeze, it can sell assets, beg its bankers, or default.
After 2018 RBI did build some partial, conditional workarounds, such as special liquidity schemes routed through banks and SIDBI/NHB refinance. They were policy choices made in a crisis. No NBFC has a right to them, and if you plan as though it does, you're planning on charity.
Advertisement
Why does India have so many NBFCs?
So why does India have several hundred NBFCs that matter, and a sector holding roughly a fifth of the system's credit? RBI's Report on Trend and Progress of Banking in India tracks that share, so pull the latest. Four forces did it, layered over five decades:
1. Bank licences were scarce and banks were absent. For most of independent India's history a banking licence was nearly unobtainable. Roughly a dozen private licences went out between 1993 and 2014, and "on-tap" licensing since 2016 has attracted few takers because the regulatory load is heavy. Meanwhile enormous credit needs went unmet. Think of a truck operator in Namakkal with no income documents, or a farmer needing ₹40,000 against gold bangles in ten minutes. Or a first-time home buyer in 1978, when no Indian bank would write a 15-year mortgage. HDFC was founded in 1977 to invent the product. Finance companies grew in the gaps: Sundaram Finance (1954), Cholamandalam (1978), Shriram (1979), Muthoot's gold-loan business (roots in 1939). None of them began as a regulatory-arbitrage shop. Each was an underwriting invention, lending against cash flows and collateral that bank credit processes couldn't read.
2. Specialization really does beat generalization at the last mile. An NBFC that has financed used trucks for forty years knows what a 2019 Tata 407 resells for in each district. Its field officers can work out a fleet operator's real income from his fuel receipts and route map. A gold lender can disburse against ornaments in minutes. It can do that because valuation is standardized and the collateral is locked in its own vault. A microfinance lender runs weekly group meetings in villages no bank branch serves. Each of these is a real cost-and-information advantage, a Phase 4 moat in underwriting and distribution, and it turns into yields banks can't safely earn.
3. The regulatory treadmill ran lighter, historically. For years NBFCs enjoyed easier norms: NPA recognition at 180 days past due against 90 for banks, lower capital intensity in practice, no priority-sector obligations, no CRR/SLR drag. Most of that arbitrage has since been harmonized away. NPA norms converged to 90 days, Ind AS brought expected-credit-loss provisioning, and the 2021-23 Scale-Based Regulation framework pulled large NBFCs close to bank-grade rules. You'll meet that framework properly in the regulatory map. For now, treat "regulatory arbitrage" as a shrinking part of the story.
4. The funding plumbing turned a sector into a boom. After demonetization in November 2016, household cash flooded into bank deposits and then into debt mutual funds chasing yield. Those funds needed short paper to buy, and NBFCs were happy to issue commercial paper and short NCDs. At the same time the corporate-NPA cycle you studied with banks had eleven public-sector banks locked in PCA and retreating from lending. So NBFC credit grew ~20%+ a year through FY17 to FY18, funded increasingly by 90-day money from mutual funds, just as banks vacated the field. NBFCs' share of incremental system credit spiked. By then every ingredient of September 2018 was in place: a sector that had borrowed short from confidence-sensitive lenders to lend long into illiquid assets. Hold on to that sentence. You'll see what happened when confidence flinched in the failure anatomy.
What are co-lending and direct assignment?
One more piece of vocabulary. Since 2020 the sector's plumbing has added co-lending, where a bank funds ~80% of a loan the NBFC originates and services. It has also added ever-larger direct assignment (DA) and securitization volumes, with banks buying NBFC-originated retail pools, partly for priority-sector compliance. A new division of labor is emerging, with NBFCs becoming the underwriting-and-collections layer and banks the balance-sheet layer. That matters for your analysis in two ways. A growing share of the business economics now happens off the balance sheet, and KPI 1 on the panel is built to catch it. And the arrangement re-links NBFC risk back into the banking system, so RBI leans against the loop with risk-weight moves.