How Indian services pay for imported oil

In the financial year ended March 2026, India ran a $337.3 billion deficit in goods. Its final current-account deficit was only $25.2 billion, or 0.6% of GDP. The distance between those numbers was covered by services, transfers from abroad and a less visible flow in the opposite direction: income paid to foreign owners of capital. It looks like a vanishing act. It is external accounting.

A tanker unloads crude at an Indian port. Somewhere else, a bank, laboratory or industrial group pays a company in Bengaluru, Hyderabad, Pune or Chennai to develop software, run digital infrastructure, analyse data or perform part of its research. An Indian worker in Dubai, London or Dallas sends money home. None of these transactions is legally connected. They meet in the balance of payments.

India pays for oil with code is therefore a metaphor, not barter. No barrel is exchanged for a line of software. Foreign currency is fungible: dollars earned from services and transfers help fund all external payments, including oil, electronics, machinery and gold.

That qualification does not weaken the argument. It reveals how the mechanism works.

A $337 billion gap that ends at $25 billion

The current account records four broad families of transactions between an economy and the rest of the world: goods, services, primary income and secondary income. Primary income includes cross-border wages, interest, dividends and profits. Secondary income covers current transfers without a direct quid pro quo, including much of the money sent between households.

For India’s 2025-26 financial year, from April 2025 to March 2026, the Reserve Bank of India’s standard presentation gives the following preliminary figures:

goods: -$337.288 billion;
services: +$216.615 billion;
primary income: -$48.208 billion;
secondary income: +$143.641 billion;
current account: -$25.240 billion.

Services and secondary income generated $360.256 billion net between them, more than the entire goods deficit. The negative primary-income balance opened the gap again. The RBI puts the final current-account deficit at 0.6% of GDP, the same ratio as in the previous financial year.

The chart uses one classification throughout: the RBI balance-of-payments presentation. It does not combine customs estimates with balance-of-payments adjustments.

A current-account deficit is not an unpaid invoice. It contributes to the economy’s net external financing need. Its counterpart appears in the financial account, which includes reserve assets, after allowing for the capital account and errors and omissions. The diagnosis depends on the funding. Long-term productive investment, a rise in deposits and a short-lived portfolio trade do not create the same exposure.

The $25 billion figure therefore says neither that everything is fine nor that India is simply “living beyond its means”. It says that, during this period, the net external financing need generated by current transactions remained small relative to the economy, despite an enormous goods gap.

Oil inside the goods deficit

The $337.3 billion figure is not the cost of imported oil. It covers all goods under balance-of-payments methodology. It includes a $266.9 billion deficit in general merchandise, a $1.6 billion surplus from merchanting and $72.0 billion in net imports of non-monetary gold.

A second number also appears in official releases. On 15 April 2026, India’s commerce ministry reported a customs-basis merchandise deficit of $333.19 billion for the same financial year. The two figures are not contradictory. Foreign-trade statistics and balance-of-payments data differ in coverage, timing and valuation adjustments. To preserve the accounting identity, the waterfall above uses RBI figures only.

Oil still deserves special attention. The Petroleum Planning and Analysis Cell’s 2025-26 Ready Reckoner puts India’s crude-oil import dependence at 88.7%, based on consumption and using provisional data. That structural dependence explains why a higher barrel price can move quickly through the external accounts. It also frames India’s trade-off between sugar, ethanol and energy security.

The transmission is straightforward:

higher crude price
        ↓
more dollars required for a given volume
        ↓
larger import bill and pressure on the current account
        ↓
greater demand for foreign currency and possible rupee pressure
        ↓
costlier fuel, freight and industrial inputs
        ↓
trade-offs across inflation, growth, public finances and monetary policy

The chain is not mechanical down to the last dollar. Refiners can change suppliers, volumes can slow, inventories can absorb part of the shock, discounts between crude grades can move and the government can delay or distribute the pass-through to retail prices. Services, remittances and other exports can offset part of the deterioration.

For the distinction between global benchmarks, crude grades, inventories and refining margins, see l0g’s oil-market guide.

Spring 2026 provided a live stress test. In a statement delivered on 5 June 2026 and published in the Bulletin on 22 June, the RBI said the Indian crude basket had averaged about $110 a barrel in April and May, at $114.5 in April and $106.2 in May. It also noted that a partial pass-through to petrol and diesel prices had begun. In the same statement, the central bank expected the services surplus and inward remittances to provide some comfort to the current account.

That is the role of code in this story. It does not remove the oil shock. It prevents the shock from moving through the external accounts without meeting foreign-currency income on the other side.

Code is an actual balance-of-payments line

India exported $421.3 billion of services in 2025-26 and imported $204.7 billion. The net surplus was $216.6 billion.

The official category covering telecommunications, computer and information services generated:

$206.6 billion in export receipts;
$27.3 billion in payments abroad;
a $179.3 billion net surplus.

The category accounted for 49.0% of gross service exports and 82.8% of the net services surplus. That is where the code metaphor comes from. It should not reduce the entire sector to software developers. “Other business services” added $124.2 billion in receipts and a $54.6 billion surplus, spanning activities such as consulting, engineering, research, advertising and technical work.

Together, telecommunications, computer and information services and other business services made up 78.5% of gross service exports. Some of that work is outsourced. Some is performed inside multinational groups’ own Indian centres, which now cover data analytics, engineering and research and development.

An RBI study of India’s service exports describes movement towards higher-skilled work, including inside multinational Global Capability Centres. The qualification matters: an upgrading sector does not mean every job or every company is moving at the same pace. It does explain why “services” can no longer be read as a synonym for call centres.

Geographical concentration remains high. The RBI’s annual survey of software and IT-enabled service exports, published on 4 November 2025 and covering 2024-25, estimated exports at $204.7 billion, excluding sales delivered through overseas commercial presence. The United States took 52.9%, Europe 32.8%, and 72.0% was invoiced in US dollars. Off-site delivery represented 90.7% of measured exports.

This survey is not the same statistic as the 2025-26 balance-of-payments category. The period, sample and definition differ. It nevertheless exposes the risk behind the surplus: a large share of final demand still depends on US and European technology budgets.

Two buffers, two geographies

The second large source of foreign currency comes from households rather than companies.

India recorded $143.641 billion in net secondary income in 2025-26. Within that total, personal transfers reached $144.072 billion net. The broader line for transfers involving financial and non-financial corporations, households and non-profit institutions was $144.794 billion. These close figures are not interchangeable. Total secondary income is slightly lower because general government recorded a $1.154 billion deficit.

“Remittances” can also span more than one line. The RBI notes that remittance statistics may include employee compensation under primary income and personal transfers under secondary income. In India, personal transfers, mainly family maintenance sent by workers abroad and certain withdrawals from non-resident deposits, make up the largest part.

The geography does not mirror software exports. The RBI’s sixth remittance survey, published on 19 March 2025 and covering 2023-24, attributed 27.7% of inward remittances to the United States, 38.0% to the six Gulf Cooperation Council economies and 10.8% to the United Kingdom. The survey covered about 99% of the value reported by banks under family maintenance and savings, but its period and scope differ from the 2025-26 secondary-income balance.

“Other” shares are calculated as the remainder to 100. Europe in the software survey and the United Kingdom in the remittance survey follow the RBI’s published groupings.

That diversification helps, but it is not perfect insurance. A US technology slowdown can weigh on services without hitting Gulf migrant income in the same way. A Gulf crisis can damage jobs, transfers or payment channels while leaving part of Western software demand intact.

Higher oil prices may even raise income in Gulf exporters and indirectly support some transfers to India. That is a plausible partial buffer, not an established hedge. When the oil shock is caused by a conflict that also disrupts migrant employment, transport or payments, the two effects can reinforce each other instead. These annual data do not establish causality.

AI can cut hours and widen the market

Artificial intelligence reaches the centre of India’s model because it changes the cost, price and scope of services at the same time.

The first effect is adverse for repetitive work. When a developer, analyst or process worker produces more in an hour, the client may buy fewer hours. Standardised tasks can be automated, brought back in-house or subjected to tougher price competition. In a model where revenue still depends partly on the number of people billed to a project, productivity can initially compress receipts rather than expand them.

The second effect works in the opposite direction. Companies need to integrate models, clean data, secure systems, control outputs, redesign processes and maintain more complex infrastructure. AI can widen the addressable market for providers able to do that work. It can also allow an Indian team to sell more value rather than merely more hours.

The third effect is distributional. Service exports can keep rising while employment intensity slows, junior recruitment weakens or margins shift towards a small number of firms. A positive current-account outcome can coexist with a difficult labour-market transition.

Available evidence does not settle the balance. On 5 June 2026, the RBI said service exports were holding up despite concerns about AI. That is useful contemporary evidence, not proof that the model is immune. Balance-of-payments statistics measure dollar values. They do not cleanly separate labour volume, prices, margins, employment and service quality.

The risk therefore has to be tracked through several series at once: export value, activity mix, geographical concentration, billing rates, margins, headcount and the share of higher-value work. Growth in the computer-services line alone will not answer the question.

Investment income reopens the gap

The least intuitive part of the waterfall is the $48.2 billion primary-income deficit.

Its main component is investment income: interest on debt, dividends, direct-investment profits and other returns owed abroad, minus the income Indian residents receive on their foreign assets. The net investment-income deficit was $58.1 billion in 2025-26. Positive employee compensation and other primary income narrowed the overall primary-income gap to $48.2 billion.

This is not a statistical leak. It is the other side of foreign capital accumulated in the economy. A factory, equity stake or loan brings money in when the investment is made. Later it may generate a dividend, reinvested earnings or interest recorded under primary income.

Foreign capital can accelerate investment, transfer technology and expand productive capacity. But as the stock of interest-bearing or profit-earning external liabilities grows, receipts from goods and services must rise merely to preserve the same current-account balance. India is not only paying for imports. It is also remunerating part of the capital that helped finance its development.

The manufacturing counterargument

The headline could be read as saying that India chose services instead of industry. The data do not support that conclusion.

On a balance-of-payments basis, India exported $446.1 billion of goods in 2025-26. Engineering goods, refined petroleum products and other manufactured exports contribute to that external earning capacity. Digital services also support factories, logistics and product design. India’s annual customs trade release uses a different statistical perimeter from the balance of payments, so the two series should not be added together or compared mechanically.

The issue is the gap. Goods imports reached $783.4 billion on the same basis. Some imports feed consumption; others finance investment and future production. Importing a machine or component is not economically equivalent to importing a finished consumer product. Gross trade figures alone do not reveal domestic value added.

A successful industrial strategy could reduce dependence on some inputs, increase exports and diversify foreign-currency earnings. It could also raise machinery, energy and component imports for a period. The trade deficit is not a simple gauge of industrialisation.

The other counterargument concerns funding quality. A small current-account deficit can be sustainable when it reflects productive investment and is financed through stable flows. It becomes more fragile when receipts are concentrated, liabilities are short-dated or investors can leave quickly. The current account measures the net need. External risk also requires the financial account, reserves, external debt and maturity structure.

The limits of the external cushion

The established arithmetic is strong. In 2025-26, services and secondary income offset more than 100% of India’s goods deficit. Primary income then brought the current account back to a $25.2 billion deficit. Computer and information services generated most of the services surplus. Personal transfers supplied a second large stream of foreign currency.

The interpretation should be more restrained. This structure makes India’s external position more resilient to a goods deficit than it would be without those receipts. It does not deliver energy independence, protection from a Western recession or immunity from AI. It relocates part of the risk, from the oil terminal to foreign corporate technology budgets, the host economies of the diaspora and the returns expected by owners of capital.

The largest unknown is correlation under stress. Will services, remittances and capital inflows remain strong at the exact moment oil rises and the rupee comes under pressure? The RBI’s June 2026 assessment pointed to cushioning capacity. It did not test every scenario or every duration.

India does not legally settle its oil bill with code. In its external accounts, however, software, engineering, consulting and diaspora money provide the foreign currency that keeps a vast import bill from turning automatically into a financing crisis.

That is a real strength. It does not need to be called a miracle.

Methodology

All headline amounts are current US dollars and cover India’s 2025-26 financial year, from April 2025 to March 2026. RBI Table 126 was published on 31 July 2026 and labels the figures preliminary. Totals may differ marginally because of rounding.

The waterfall uses the RBI’s BPM6 presentation only. It does not add the commerce ministry’s $333.19 billion customs-basis trade deficit to those series. Net secondary income of $143.641 billion, net personal transfers of $144.072 billion and the broader $144.794 billion private-transfer line are distinct accounting entries.

The concentration chart juxtaposes two separate surveys: software export destinations in 2024-25 and bank-reported remittance origins in 2023-24. It compares geographical distributions, not dollar amounts. The 49.0%, 82.8%, 78.5% and 106.8% ratios are l0g calculations using RBI data.

India Pays for Oil with Code