Delhi shifts gears with a new EV policy | India's credit market shift

ELI5/TLDR

Delhi’s first EV policy was all carrots — subsidies, tax waivers, scrappage cash. It missed its target. The new draft policy adds sticks: hard cutoff dates after which you simply cannot register a new petrol or diesel vehicle in certain categories. Fleets first, then three-wheelers in 2027, then all two-wheelers in April 2028. Subsidies taper sharply year-on-year so dawdling gets expensive. Separately, India’s loan securitization market hit Rs 2.55 lakh cr in FY26 — more than double four years ago. NBFCs now drive 97% of it (banks pulled back to 3%), gold loans surged from nothing to 15%, and microfinance investors are demanding the safer PTC structure after a bad asset-quality year.

The Full Story

Delhi EV policy 2.0

Delhi has 87.6 lakh registered vehicles as of March 2026, and per CSE about half of locally generated pollution comes from their tailpipes. The 2020 policy aimed for 25% EV share in new registrations by 2024 and ran on incentives alone — buyer subsidies, road tax waivers, scrappage, all funded by a pollution cess. Six years on: 4.7 lakh registered EVs (5.4% of stock), 12.7% of new registrations in FY26, well above the 8% national average — but petrol still owns 73% of new registrations and 95% of the parc.

So the draft 2026 policy adds mandates. The deadlines roll out segment by segment, hitting the highest-kilometre users first.

Fleets — immediate. Delivery and aggregator fleets (Zomato, Uber, Ola, etc.) cannot induct new petrol or diesel two-wheelers, light commercial vehicles, or N1 goods carriers. Small carve-out for BS6 two-wheelers. A delivery rider does 80–100 km/day vs. a personal commuter at 15, so per-vehicle emissions impact is outsized.

Three-wheelers — 2027. Only EVs can be newly registered. CNG autos, once the clean upgrade, now exit. Existing CNG autos keep running.

Two-wheelers — April 2028. Only electric two-wheelers can be newly registered. Two-wheelers are 67% of Delhi’s vehicle stock — this is the boldest single bet in the policy.

No deadline forces existing owners to scrap. The mandate only kicks in when you buy new.

The subsidy structure is deliberately front-loaded to pull demand forward. Electric two-wheelers under Rs 2.25 lakh: Rs 10,000/kWh capped at Rs 30,000 in year 1, Rs 6,600 in year 2, Rs 3,300 in year 3. Three-wheelers: Rs 50,000 down to Rs 30,000 over three years. Goods carriers get up to Rs 1 lakh in year 1. School buses included. Steeper slopes for vehicles that drive more.

Scrapping a BS4 or older vehicle adds another layer: Rs 10,000 for two-wheelers, Rs 25,000 for three-wheelers, Rs 50,000 for goods vehicles. Three conditions: scrapper certificate of deposit, new EV bought within 6 months, payment via DBT to the registered owner only.

Road tax and registration fees: 100% waiver for EVs through the policy period, but only on cars priced under Rs 30 lakh ex-showroom (waiver runs till March 31, 2030). Above Rs 30 lakh, full tax. Strong hybrids under Rs 30 lakh get a 50% reduction — a small concession to bridge the transition.

Charging infrastructure is the obvious failure point. Every OEM must ensure at least one public charging station per dealership — three points for two/three-wheelers, two for four-wheelers. All new civil infrastructure under Delhi NCR govt must be EV charging-ready. The 2020 rule that 20% of new building parking be EV-ready stays. Delhi Transco Limited (DTL) becomes the nodal coordination agency for the public charging and swapping network.

Grid management gets serious. The Ministry of Power’s 2024 guidelines push for solar-hour charging — time-of-day tariffs make solar hours 10–20% cheaper, peak hours 10–20% more expensive. DERC has engaged with ToD structures. For apartment dwellers, the unsolved mess remains: who pays for the wiring upgrade, who hosts the charger, who controls the RWA’s veto. The policy doesn’t address this.

The current state: ~3,100 charging stations and 893 swap stations as of March 2026. The 2022 plan had targeted 18,000 charging points by 2024. Far behind. If supply doesn’t catch up to mandate-driven demand, the deadlines will get pushed.

India credit-market shift

Two rating agency reports — Crisil and CareEdge — on the FY26 securitization market. First, the vocabulary.

Securitization is when a lender bundles a pool of loans (say, an NBFC’s Rs 500 cr of vehicle EMIs) and sells it to investors for an upfront discounted amount (say Rs 480 cr). The lender gets cash today; investors collect the EMI stream over five years. The lender keeps servicing the loans for a fee, the borrower never knows.

Two routes:

Pass Through Certificates (PTCs): Loans go into a special purpose entity (SPE), which issues tranched securities. Senior tranches paid first (lower risk, lower yield), junior tranches absorb losses first (higher risk, higher yield). Plus a cash buffer and over-collateralisation (pool larger than securities issued — e.g., Rs 500 cr pool backing only Rs 450 cr of PTCs). All these layers are called “credit enhancements.”
Direct Assignment (DA): NBFC sells the pool directly to a bank, no credit enhancements (RBI prohibits them). Cheaper, faster. Only RBI-regulated entities can buy. Bank takes full credit risk.

In FY26, PTCs were ~60% of the market (an all-time high), DAs the remaining 40%.

Securitization is for performing loans. NPAs are a different market — sold to ARCs, who issue Security Receipts. The Big Short was the cautionary tale: US banks resecuritised mortgage-backed securities into CDOs, then CDOs of CDOs. RBI banned resecuritization and synthetic securitization, requires originators to hold loans 3–6 months before selling, and to retain 5–10% skin in the game. India’s market deals only in straightforward, granular retail pools.

The headline: market hit Rs 2.55 lakh cr in FY26, up from Rs 1.13 lakh cr in FY22. More than doubled in four years. Three structural shifts underneath:

1. NBFCs took over. NBFCs originated 97% of FY26 deals, up from 74%. Bank originations collapsed from 26% to 3% — entirely because HDFC Bank pulled back. HDFC had aggressively securitized in FY25 to manage its post-merger CD ratio (it absorbed HDFC’s loans but not its deposits); once the ratio normalized, it stopped. No other bank stepped in. NBFCs need securitization structurally — they borrow short (1–3 years from banks/markets) and lend long (5–7+ years for housing). Securitization fixes the duration mismatch. After RBI raised risk weights on bank-to-NBFC lending in November 2023, this funding channel became even more critical.

2. Gold loans went from nothing to 15%. A blip in FY25, second-largest asset class behind vehicle loans in FY26. Gold loan disbursements were up 94% YoY in Q3 FY26 alone, helped by rising gold prices and tighter unsecured lending norms. The pools are nearly ideal for investors — collateral is liquid, easy to value, physically held by the lender. Tenures under 12 months. Default → auction → recovery near-certain. Most happen via the simpler DA route since the collateral itself is the protection.

3. Microfinance moved to PTCs. MFI loans held at ~12% of the market by volume, but the structure shifted hard. PTC share jumped from 30% to 69% in one year. Reason: MFI gross NPAs roughly doubled industry-wide in FY25 — investors who’d bought via DA got burned. They didn’t leave; they demanded credit enhancements.

Who’s buying? Mostly banks, partly because they have to. RBI requires 40% of credit go to priority sectors (agri, MSMEs, housing). Miss the target, you park the shortfall in NABARD at below-market returns. NBFCs lend exactly to those priority-sector borrowers banks can’t reach — so banks buy securitized pools and tick both boxes (yield + compliance). The alternative, Priority Sector Lending Certificates (PSLCs), is pure compliance with no yield. Securitized pools dominate because they pay.

The originator base broadened from 175 to over 190 entities. The top 20’s share fell from 71% to 65%. Read it as a maturing market: pricing risk more intelligently, demanding PTC armour where asset quality wobbles, staying simple where collateral is strong.

Key Takeaways

Delhi’s 87.6 lakh vehicles cause >50% of locally generated pollution; the 2020 EV policy hit 12.7% new-vehicle EV share vs. 25% target.
Hard cutoffs: fleets immediate (no new ICE inductions), three-wheelers in 2027, two-wheelers in April 2028.
Two-wheeler subsidy: Rs 30,000 → Rs 6,600 → Rs 3,300 across years 1–3. Three-wheelers: Rs 50,000 → Rs 30,000. Goods: up to Rs 1 lakh year 1.
Scrappage: Rs 10k / 25k / 50k for two-wheelers / three-wheelers / goods, conditional on buying an EV within 6 months.
Road tax waiver caps at Rs 30 lakh ex-showroom EV cars; strong hybrids under Rs 30 lakh get a 50% cut.
~3,100 charging stations and 893 swap stations vs. a 2022 target of 18,000 by 2024 — supply is the constraint.
Securitization market: Rs 2.55 lakh cr in FY26, up from Rs 1.13 lakh cr in FY22.
NBFCs = 97% of originations in FY26 (was 74%); banks crashed to 3% on HDFC pulling back post-merger.
Gold loans surged to 15% of market; gold loan disbursements up 94% YoY in Q3 FY26.
MFI PTC share jumped 30% → 69% in one year as investors demanded credit enhancement after FY25 NPA doubling.
PTCs at ~60% of total (all-time high), DAs at 40%. Originator base: 175 → 190+, top-20 share 71% → 65%.

Claude’s Take

Two stories that pair better than they look. The Delhi EV piece is the most useful regulatory development of the year — not because the subsidies are big (they aren’t, especially against FAME II’s national support), but because mandates change OEM and dealer behaviour in ways incentives don’t. Bajaj, TVS, Hero MotoCorp now have a hard date by which their Delhi books need to be majority-electric. That ripples back into capacity planning, dealer training, and battery supply contracts. Watch fleet operators first — they’re the canary, and the segment where unit economics for EVs already work.

The unsolved problem is exactly what Akira flagged: charging in apartment buildings. Delhi is dense, RWAs are obstinate, and the policy hands them a veto by silence. Without a clear right-to-charge framework — like California’s — the 2028 two-wheeler deadline ends up policing only homeowners with private parking. A wealth filter dressed as a climate policy.

The credit market story is the more revealing of the two. Securitization volumes are India’s quietest macro indicator — when they spike, NBFCs are getting squeezed elsewhere. The HDFC-driven 26% → 3% bank origination crash isn’t a story about banks abandoning the market; it’s a story about how concentrated the supply side is. One bank’s CD-ratio housekeeping moves the entire bank-originated market by 23 percentage points. That’s not a deep market. That’s a thin market with one big swing trader.

The PTC migration in microfinance is the most important number in the report. Investors aren’t running from MFI — they’re staying but charging more for the same risk. That’s how a market matures, and it’s also how funding gets quietly more expensive for the smaller MFIs who can’t structure good PTCs. Bharat Financial, Spandana, Ujjivan, CreditAccess — watch their borrowing costs in the next two quarters.

Score: 7/10. Two well-explained policy/structural pieces, accurate and grounded. Loses a point for the GE-HAL F414 tidbit being an afterthought (worth more) and for not pressing harder on the apartment-charging gap, which is genuinely the rate-limiting step for the 2028 deadline.