Journal · 31 Aug 2026 · 20 min read
Bengaluru, a Tuesday evening in May. The kind of evening where every air conditioner in the city runs flat out, and the grid strains under the weight of it.
I was on a call. Mid-sentence. The fan overhead stopped. The lights went. For about a second and a half, the room was just dark and quiet.
This piece is also read aloud in the Desk — sign in to listen.
Then — click — the inverter kicked in. Lights back. Fan spinning again. Bijli wapas aa gayi, more or less, except the grid outside was still down. My call didn't even drop.
I didn't do anything in that second and a half. I didn't reach for a candle, or my phone's flashlight. I didn't even register the outage as something worth reacting to.
Because someone had already reacted to it. Years before it happened.
Whoever wired that inverter into the house didn't know which Tuesday in May the grid would trip. They just knew, someday, it would. So they built the switch-over before the outage ever arrived.
That's the whole story I want to tell you today. Not about power cuts. About oil.
The one idea you already own
This piece is also read aloud in the Desk — sign in to listen.
Scale that inverter up. A hundred million households. One very large, very thirsty economy. That's close to the shape of the last decade of India's energy policy.
Start with the exposure: India now buys almost nine of every ten barrels of oil it uses from abroad — 87.4% three years ago, 87.8% the next, 88.2% most recently, still climbing. Only the US and China buy more. A large share has normally travelled through the Strait of Hormuz, between Iran and Oman — no clean official number exists, but crude estimates run four in ten to just over half, depending on the year.
The bill only gets bigger from here. One of India's top credit-rating agencies puts a sustained $10-a-barrel rise at $13–14 billion in extra imports a year, and the gap between what the country earns from the world and pays back widens by roughly three-tenths of a percent of yearly output. The exposure keeps growing, too: the IEA names India the new centre of gravity for oil demand, on track to add close to 2 million barrels a day by 2035 (World Energy Outlook 2025) — almost half of every new barrel the planet will need. Whatever India builds here isn't a patch for a one-time problem — it's infrastructure for a bill that keeps growing.
A country that exposed should, on paper, be one of the fragilest large oil buyers on the planet. Instead, over the last decade, it built itself into one of the most flexible. Not the cheapest. Not the cleverest. Flexible — the way a house with an inverter isn't the house that never loses power. It's the house where losing power stops being an event.
Before I show you how, one idea to hold onto. It's the only mental model you actually need for everything that follows — and you already own it. You've used it every time you've booked a cab in the rain.
More buyers than sellers, and price goes up. More sellers than buyers, and price comes down. A surge fare at 8 PM. A discount at 2 AM. An oil tanker in the middle of a sanctions regime. Same engine, every single time. Keep it in your pocket. You'll need it four times before this is over.
This piece is also read aloud in the Desk — sign in to listen.
I'm calling the whole shape the Standby Circuit. It has four parts. Once you've seen them, you'll notice the same shape everywhere else — including, by the end, on your own chart.
The story you've probably half-heard
Here's the version that's probably reached you already, in fragments, on your feed: India got cheap Russian oil, annoyed the US into a 50% tariff, and got hit with a European ban — and somehow came out ahead anyway. Because India is smarter, Europe is stuck, and America couldn't do a thing about it.
I believed some version of this too, the first few times I saw the headlines — a satisfying story, a hero, a villain, a clean ending. It's not the one the record supports, though. The real version is better than the flattering one. Just not about anyone winning.
Between October 2025 and February 2026 — about four months — every leg of this trade got hit at once. Refiners got sanctioned. The European loophole got legislated shut, with India named by volume. A 50% tariff wall went up, priced explicitly as a penalty for the Russian purchases. Not three separate headlines — close to a coordinated attempt to shut the whole trade down in one go.
And India did not break. Not because it won a fight — because the flexibility that absorbed the hit had been built years before anyone knew what shape the hit would take.
This piece is also read aloud in the Desk — sign in to listen.
Here are the four parts that built it.
Part one: more than one place the power can come from
For most of independent India's history, its oil supply ran through a narrow set of Gulf relationships — sensible enough, geographically, until you remember the chokepoint sitting in the middle of it. The response, built quietly over years: keep adding backups. Ek se bhale do — one is good, two is safer.
India's crude suppliers went from 27 countries to 41 — around 40 as of March 2026. LNG sourcing went from 6 countries to 15. One backup that quietly grew stronger for years: the United States. Crude from American producers to India rose from 141,000 barrels a day in January 2025 to 235,000 in January 2026 — up about two-thirds in twelve months. Officials by 2026 called it deliberate policy. A cushion, not a discount hunt.
None of that widening happened because of the sanctions, the tariff, or the EU's rule change — it was already running before any of those three things existed. Nobody waits for the power cut to go shopping for an inverter.
There's a second backup in this story, and it's the one that made the headlines: Russia. Before February 2022, Russian crude was a rounding error in India's import mix — roughly 2%. By FY23 it was 21.6%, by FY24 35.9%, holding around 35.8% through FY25 — for a couple of years, roughly one barrel in three landing at an Indian port was Russian.
Now apply the mental model. Why was it so cheap? Because Russia had lost almost all its buyers overnight, the day the West stopped purchasing. Same oil, same pipelines — suddenly one seller with far more supply than anyone left willing to touch it. More sellers than buyers, of course the price fell. Not generosity, not a special friendship. Demand and supply, running in reverse.
That's not a backup that got added in the calm. That's one that got massively overloaded, fast, because the price was too good to leave on the table. Which is exactly why, when Washington sanctioned two of Russia's largest oil exporters in October and November 2025, this was the one under the most strain. Indian refiners diversified hard: for the first time in two years, Russia's share fell below 25% between December 2025 and February 2026, volumes down to roughly 1.0 to 1.2 million barrels a day.
Read correctly, that isn't a retreat. It's what backups are for. The system didn't collapse when its most heavily used one got cut down. It leaned on the others — because the others already existed.
Part two: the changeover isn't free
An inverter is only as good as the moment it takes over — it has to notice the grid has failed and swap you onto stored power fast enough that you don't feel the gap. India's oil trade needed the equivalent. A way to keep paying for Russian barrels without crossing a line the G7 had drawn. The G7 had set a ceiling on what anyone could pay for Russian oil — meant to starve Russia's war chest, without cutting the world off from Russian barrels entirely. And without some bank or insurer, thousands of miles away, getting punished just for touching the deal.
What got built was genuinely inventive, even without a press conference. Indian banks toggled between the UAE's dirham, the Chinese yuan and the rupee to keep the money moving. The dirham worked especially well — pegged to the dollar without being the dollar, so payments could clear without touching the sanctioned rail.
New shipping and insurance arrangements got built alongside it. Sanctions don't stop oil from physically loading onto a ship — they stop the finance, insurance and shipping that make that loading legal and payable. Attack those three, and the oil doesn't move, even with no blockade.
None of that came free — fewer insurers and shippers were willing to touch these voyages, and when buyers outnumber sellers for a scarce service, the price goes up. Same engine, again: that apparatus carried a premium of $5 to $15 a barrel in the tightest periods.
The headline discount on Russian crude was real, but never the full number some feeds implied — a meaningful slice got eaten by the cost of the workaround itself. Worth remembering: a workaround is rarely free — muft ka kuch nahi hota. The question is only whether it's cheaper than not having one.
Part three: turning what comes in into something you can sell onward
Here's where the story gets genuinely elegant, and where it collided head-on with two governments at once. India doesn't just burn the crude it imports — it refines a large share and sells the output onward. More than 250 million tonnes of refining capacity a year makes it a top-five refining nation and a major exporter of diesel and jet fuel. Something over a quarter of what it imports — roughly 1.2 million barrels a day — isn't consumed at home: it goes in as crude, comes out as a different, more valuable product, headed elsewhere. Call it a costume change — crude converted into something a buyer who couldn't touch the raw version will happily pay for instead.
A quick, honest admission before we go further, because this is the section that humbled me most while writing it. My MBA taught me to read a balance sheet in my sleep. It never taught me to trace one barrel of crude through three countries' paperwork and back out as someone else's diesel — that took hours with a calculator and more wrong drafts than I want to admit. Coming from a modest background, nobody around me worked in energy or shipping. I built this understanding the same slow way I'm asking you to build a chart-reading eye — badly at first, then less badly.
Through 2025, one buyer took most of that output: Europe. India supplied 61.2 million barrels — 61.2% of all EU imports of refined products made from Russian crude — in 2025 alone. European buyers didn't want Russian crude, but still wanted cheap diesel — the demand never went away, just the label, once it passed through an Indian refinery. That was simply what the costume change was for.
Then the rules changed, aimed squarely at this trade. The EU's 18th sanctions package, effective 21 January 2026, barred EU buyers from importing petroleum products refined in a third country from Russian-origin crude. It covered anything refined after 21 November 2025, and required a 60-day non-Russian-crude-only run to qualify again — no mixing-and-paper-accounting workaround allowed. It wasn't subtle about who it targeted: India and Turkey, India the larger target. Almost the same window, Washington opened a second front — a 50% tariff, stacked as 25% plus 25%, from 27 August 2025, priced explicitly as a penalty for buying Russian oil. By February 2026 an interim deal cut it to 18%, India agreeing to reduce Russian purchases and buy more American — and potentially Venezuelan — crude instead. A live thread, not yet closed.
There's a second dial in this same story, easy to miss because it never wears the language of geopolitics: India also taxes its own refined-fuel exports, a tax it raises or lowers every couple of weeks depending on how tight domestic supply feels. Raise it, and staying home looks like the better trade — sellers keep more of the barrel inside the country, supply loosens, nobody has to ban an export. This isn't new or gentle. The tax went up in July 2022, after the invasion of Ukraine. It was scrapped in December 2024. It returned on diesel and jet fuel in March 2026, as prices climbed again on the West Asia conflict — the stated reason was keeping fuel at home, not a tariff countermove. I won't hand you a motive the record doesn't support. It moves fast: in the first three weeks of August 2026 alone, the diesel export tax went from ₹15.5 a litre, to ₹25.5, to nil — three settings inside one month. Not a typo, not chaos — a dial somebody adjusts as conditions shift. The lesson was never the number; it's that the dial exists, and somebody's hand never leaves it. Living. Recalibrating.
And here's the number that tells you whether this held. In July 2026, India's product exports hit a ten-month high — on rebounding European purchases, not collapsing demand. That only makes sense one way: refiners had somewhere else to source non-Russian crude, fast enough to clear that new 60-day window and keep selling into the very market the EU had just tried to close. That "somewhere else" is part one — redundant lines built years before Brussels drafted an eighteenth package, the reason part three could adapt in months, not years.
Part four: needing less from a grid already struggling
A standby circuit isn't only about more supply. Sometimes it's about needing less from the grid already under strain — the fan and the lights running while the AC waits its turn.
India's version is the ethanol-blending programme — mixing plant-based ethanol into petrol so pumps sell less pure petrol and more blend. It's been fast: supply has gone from 38 crore litres in the programme's first year to just over 1,000 crore litres now, more than 26 times as much. Blending reached roughly 20% average by mid-2025, a 2030 target arrived five years early, and went nationwide in April 2026 — pumps now sell E20 as standard. Every litre of ethanol in the tank is a litre of petrol the pump didn't pull from imported crude. Eleven years of that adds up — real forex savings, a farmer-income bump.
Now the honest arithmetic. Seedhi baat — it cuts both ways. Petrol isn't most of a barrel, it's roughly a sixth; diesel, LPG, jet fuel make up the rest. E20 blending a fifth of the petrol pool touches a fifth of a sixth of the barrel, not a fifth of the oil bill — and ethanol carries less energy per litre besides. The honest number lands between roughly 2% and 3.5% of crude imports a year — not a 20% cut in oil, but not trivial: against a roughly quarter-billion-tonne annual import bill, that's around $4 billion a year, close to independent estimates, and unlike a one-off trade deal, it doesn't expire. It compounds instead of fading. Small percentage. Real, permanent number.
That saving didn't come free. The nationwide rollout landed with a genuine backlash — something like 200,000 petitions alleging mileage loss and vehicle damage. Here's where I need to correct something, because the correction is its own lesson. After a June 2026 court hearing, it was widely reported that the government's top lawyer had called E20 an "experiment" in the Supreme Court — the line travelled everywhere, fitting what a lot of angry drivers already suspected: that they were the test subjects. The Attorney General's office formally denied it on 30 June 2026, stating that no such submission was made. What actually happened was procedural — a request to bundle similar ethanol-allocation petitions from different High Courts to be heard together, with the Court holding things steady while that got sorted. That's the whole submission.
I'm telling you this not to defend anyone, but because it's the cleanest example here of the exact thing I keep asking you to watch for: a quote travelled faster than the transcript, and the correction reached a fraction as many people. Read only the first version and your grievance points at the wrong sentence — but the real grievance, real drivers with real mileage complaints, never needed an invented quote to be legitimate. Field trials reported no compatibility issue; the government states a real mileage dip of 2–4%, and the Supreme Court dismissed a legal challenge over lack of consumer choice — but a dismissed case doesn't erase the grievance, any more than a corrected misquote erases the frustration that made it spread.
There's a second, harder cost alongside it: sugar. In the third week of August 2026, ex-mill sugar prices in Maharashtra pushed past ₹58 a kilo — a fresh record — with retail nationwide averaging ₹63–65, and ₹70–75 in some cities, before easing after the government stepped in. It's easy, and wrong, to blame ethanol — cane pulled from factory to fuel tank. The causation runs the other way: two of the three ethanol prices paid to mills have been frozen since late 2022, the third only nudged up in January 2025 — nowhere close to what mills must pay farmers for cane, which climbed from ₹230 to ₹355 per hundred kilos over the same decade. Mills had every reason to keep making sugar, little reason to divert cane to ethanol: when the fuel buyer won't pay enough and the sugar buyer will, a mill sells sugar. Not sabotage — just where the better price is.
Ordinary households pay more for sugar either way. The government's response: a duty-free import quota of a million tonnes, and a pivot toward surplus rice as feedstock for 2026/27, instead of leaning further on cane — an attempt to loosen the bind, not proof it's fixed. Sugar output is actually expected to rise, around 15% in 2026 to roughly 35 million tonnes, on a good monsoon. Not a supply-collapse story — a pricing and incentive story, still open, which is why I won't tell you this front came free.
And here's where two of these four parts turn out to be one part wearing two hats. Every litre of ethanol blended domestically is a litre of petrol that no longer has to be sold at home — it can leave the country instead. Less demand at home becomes more supply for export, and reports through 2026 point to exactly that: fuel exports climbing to multi-year highs. Part four isn't separate from part three. It's upstream of it.
When all four got tested at once
Put the four parts next to each other, and the shape stops looking like four separate news stories. Backups built over a decade, quietly, before anyone needed them. A changeover with a real cost every time it happens. A costume change that turned one binding constraint — can't sell Russian crude directly — into an exportable product, with a domestic dial re-tuned every fortnight. And a demand-side cut that arrived five years early, worth a real, compounding few billion dollars a year, with a mileage backlash, a misreported quote, and a sugar bill to show for it.
Then, in the space of about four months — October 2025 to February 2026 — all four parts got tested at once. The refiner sanctions hit the backups directly. The EU's eighteenth package hit the costume change by name. And the 50% tariff hit the whole trade's economics.
If this system had been built to survive any one of those individually, it would likely have failed when all three arrived together. It didn't.
Russia's share fell, but it didn't collapse to zero — the country kept buying elsewhere without a fuel-security event, and the tariff got renegotiated down to 18% within months. Product exports to Europe, the very trade the EU tried to shut down, hit a ten-month high five months after the ban took effect. Refining capacity kept expanding through it all — another 32 million tonnes due online by early 2027, with plans reaching toward 450 million tonnes by 2030.
That is not a story about outsmarting three governments at once — it would be dishonest to tell it that way. Nobody "won" a confrontation here. A country still paid a real premium on its own workaround. Its citizens paid a record sugar price through August 2026.
The honest version is quieter than the triumphant one. I think it's more useful, too: the country wasn't tested and found strong. It was tested and found flexible — because the flexibility had been built years before anyone knew what shape the test would take.
Back to your own chart
Here's why I wanted to walk you through all four parts instead of handing you the headline version — and why I stopped to correct that one quote instead of quietly deleting it. A line about an "experiment" moved faster than the transcript that corrected it. Not a side story I owed you as a footnote — the main story, wearing an ethanol costume.
The shape is the same one you're training yourself to see on a much smaller screen, every single day: every time a piece of news moves a stock or an index — a policy announcement, an earnings print, a tariff, a ban — almost everyone watches the arrow. Up or down. How much. How fast. That's the signal, and a signal is someone else's eyes.
What actually decides how far that move can carry, and how much it costs to absorb, was never the headline — it's whatever standby circuit was already wired into the price before the headline arrived: the buying already there, the supply already lined up, redundancy built by people who had no idea which particular Tuesday the test would land on.
You don't get to see India's four fronts on a price chart. But you can practice the same eye on something you already follow.
Try this tonight (5 minutes): Open a chart of anything on your watchlist — an index, a stock, whatever you already track. Find the last time it moved sharply on a piece of news. Then look at what price was doing in the weeks before that headline, not after. Was it already sitting at a level it kept returning to? Already building a base? Already thin on one side? You're not trying to predict the next headline. Nobody can, and this isn't that kind of exercise. You're training your eye to separate the two things that get flattened into one story every time: the move everyone's discussing, and the structure that quietly decided what the move would cost.
Structure is your own.
That's what stayed with me from that Tuesday evening in Bengaluru, more than the second and a half of darkness. The outage didn't decide anything. The wiring nobody saw — done years earlier, by someone who wasn't trying to be clever in the moment, just trying to make sure the moment, whenever it came, wouldn't decide everything on its own — did that.
Read for that wiring. On your own chart. And in the next headline that tries to tell you the whole story.
The freedom is in the read.
A note on sources: the import-dependency figures (87.4%, 87.8%, 88.2%) are from India's Petroleum Planning & Analysis Cell (PPAC); Hormuz-routing estimates draw on Kpler vessel-tracking data; the 2035 oil-demand projection is from the IEA's World Energy Outlook 2025; the EU product-import rule is Council Regulation (EU) No 833/2014, Article 3ma; the E20 mileage field trials were run by IOCL, ARAI and SIAM. Figures are as reported on the dates named and are not updated after publication.
— Anubhav
Written first. Dated. Left standing.