HSN (chapter-level) merchandise export and import data from the Ministry of Commerce's own TRADESTAT database, FY2018-19 through FY2025-26, summed across all 98 tracked HS chapters to reconstruct the national trade balance from the ground up.
The shaded gap between the lines is the trade deficit.
FY2020-21's much narrower gap is the COVID-19 import crash (India's imports fell further than its exports during lockdowns) — a one-off, not a structural improvement; the deficit resumed widening the very next year. FY2025-26 is the most recent data available (updated 19 May 2026 per the source), not necessarily a complete fiscal year.
Top 8 deficit and top 8 surplus chapters, by net trade balance.
Mineral fuels (HS27) alone accounts for a -$147.5bn deficit — India imports far more crude oil and refined fuel than it exports, by design (import-dependent energy needs). Pearls, gems & jewellery (HS71) is the second-largest deficit chapter, at -$81.1bn, driven mainly by gold and rough-diamond imports that are only partly re-exported as cut stones and jewellery. Pharmaceuticals, vehicles, cereals and apparel are the strongest surplus chapters — all comfortably export-led.
Goods balance (red) + services net (teal, stacked on top) = overall balance (marker).
Services data is India's "Non-factor Services" category from RBI's Balance of Payments "Invisibles" series (via the MoSPI connector's RBI dataset) — principally software, IT/BPO, business and travel services — converted from ₹ Crore to US$ billion using fiscal-year-average exchange rates built from this repo's own currency segment. DPIIT itself doesn't publish services-trade statistics — its main public data product is a goods import-monitoring system, not a trade-balance series. The Commerce Ministry's own DGCI&S does track services exports directly, via annual "Service Exports Reporting Form" (SERF) reports, but those are PDF/Excel releases, not a queryable database like TRADESTAT, so RBI's BoP series is used here instead — the two should tell a consistent story, but haven't been cross-checked against each other line by line.
This closes the loop on the currency and reserves segments elsewhere in this repo. Even with services trade included, India ran a current-account-style deficit every year from FY2018-19 to FY2024-25 — meaning India still needed to buy more foreign currency than it earned, just by a smaller margin than the goods numbers alone suggest. That's structural downward pressure on the rupee that has nothing to do with any single shock, and it's the backdrop the RBI has been intervening against: spending down reserves (from a $728.5bn peak to a $666.9bn trough) specifically to slow a depreciation this trade picture would otherwise make worse, not to engineer a competitive devaluation.
It also means the rupee's real depreciation (see the effective-buying-power segment) isn't just noise — a genuinely widening gap between what India earns abroad and what it spends abroad, even net of its strongest export sector, is exactly the kind of fundamental that currency markets price in over time, on top of whatever short-term shocks hit in any given month.
Checked this bulletin's TRADESTAT-derived FY2025-26 totals against three sources not previously used in this repo: the Commerce Ministry's own Trade Intelligence and Analytics (TIA) Portal (trade-analytics.commerce.gov.in), DGFT's Trade Statistics & Analytics hub, and a PIB press release covering the same fiscal year.
Imports match almost to the dollar: TIA Portal reports FY2025-26 merchandise imports of $776,013.65 million against this bulletin's own $776,013.6 million — both ultimately trace back to the same DGCI&S source data. The single largest export chapter also matches exactly: HS27 (mineral fuels) at $55,871.2 million on both. PIB's official release states FY2025-26 merchandise exports of $441.78bn and a merchandise trade deficit of $333.19bn — within a rounding error of this bulletin's $441.7bn and -$334.3bn. One discrepancy worth flagging: TIA Portal's own FY2025-26 export total reads $451.07bn, about 2% (~$9.3bn) above both this bulletin and PIB — most likely a data-vintage difference (TRADESTAT was queried 18 Jul 2026; both are marked provisional for the current fiscal year), not a methodology error, since two of the three sources agree closely. DGFT's "Import Export Data Bank" link resolves straight to tradestat.commerce.gov.in — the same source this bulletin already uses, not an independent one.
DPIIT's 2025-26 Annual Report (295 pages, dpiit.gov.in) doesn't publish a trade-balance series — consistent with what this repo found when checking DPIIT for services data — but its Industrial Entrepreneur Memorandum (IEM) Part-A investment-intention data (Appendix IV) independently corroborates the fastest-growing import chapters identified in the historical-trends bulletin: proposed investment in Electricals Equipment rose from ₹65,979 crore (2021) to ₹233,829 crore (2024), and in Fertilizers from ₹1,112 crore (2021) to ₹41,127 crore (2024) — domestic capacity-building is scaling up fastest in exactly the two sectors where import growth is also fastest. DPIIT's "24 champion sub-sectors" programme explicitly names electronics and aluminium among sectors chosen partly for "need for import substitution," using the same logic this analysis arrives at independently from the trade data alone. The report's own industry-specific sections (watches, toys, rubber goods) cite HS-code export/import figures sourced from the "Department of Commerce Trade portal" — the same DGCI&S lineage as TRADESTAT.
Underlying validation figures: data/trade_data_cross_validation_2026-07-18.json.
405 without first visiting the site to establish a session cookie and CSRF token. Once that's in place, a standard form POST works and returns two fiscal years of data per query, so 4 queries per trade direction (8 total) covered all 8 years.