Oil and gas: reserves accounting, upstream versus downstream, and under-recoveries

An integrated oil company reports a bumper year - group profit up sharply as the crude price rises - and a reader concludes the business loves high oil prices. But inside that single figure, one business is booming, a second is being crushed by the very same crude price, and a third is quietly accumulating a subsidy the government has promised but not yet paid. An integrated oil company is not one business having a good year; it is several businesses pulling in opposite directions, blended into a number that hides more than it shows. illustrative

The parts diverge because they sit at different points in the oil chain. Upstream - exploration and production - pumps oil out of the ground and sells it, so it earns more when crude is expensive. Refining turns crude into products and earns the gross refining margin, a spread that swings on its own global cycle. Marketing sells fuels to consumers, and where those fuels are price-regulated, a crude spike squeezes it, because the cost rises but the retail price cannot follow - producing under-recoveries the government is meant to compensate. The same crude price that gilds upstream squeezes marketing, and the group number is their net. So this module reads an oil company through its segments and its cycle, never through its blended profit - because oil is deeply cyclical, and a single year, especially a peak, is the worst possible basis for a view.

The upstream-downstream split is the fundamental structure: upstream profit rises with the crude price, while downstream (refining and marketing) is driven by refining margins and regulated pricing and is often squeezed when crude is high. The two hedge each other, and the group is read by its parts. See the segments diverge first, then take them one at a time.

Upstream rises with crude. Exploration and production pumps oil and gas and sells it at prevailing prices, so its profit tracks the crude price closely. Its own cost of production per barrel is relatively fixed, so the crude price flows fairly directly to its margin. Upstream is the part of an integrated company that wants high oil prices - and in a crude spike, it is where the group's profit surge comes from.

Refining earns the gross refining margin. Refining buys crude and sells petrol, diesel and jet fuel; its profit per barrel is the GRM, the spread between product value and crude cost. GRM swings on its own global cycle of refining supply and demand, largely independent of the crude price level, so a refiner can have a great year when GRMs are wide and a poor one when they are thin. A high GRM is usually a cyclical windfall. The durable advantage in refining is a complex refinery that can process cheaper, heavier crudes and make a higher share of valuable products, earning a premium GRM through the cycle. So read the GRM against its cycle, and the refinery's complexity as the structural edge.

Marketing is squeezed by regulated pricing, and under-recoveries are the tell. Marketing sells fuels to consumers through retail outlets. When those fuels are price-regulated - the retail price capped by the government - a rise in crude raises the cost of the fuel while the retail price cannot follow, so the marketing margin is squeezed exactly when upstream is booming. The shortfall between cost and capped price is an under-recovery, which the government is meant to compensate through a subsidy. Large, rising under-recoveries mean the company is booking a receivable or expecting compensation that may be delayed, paid partly, or settled in illiquid instruments - special bonds and the like that are hard to turn into cash quickly. So a marketer's reported profit can look healthy while resting on a subsidy that has not arrived in cash: a policy-dependent receivable, not collected earnings.

The whole point is that the crude price acts on the three segments in different directions at once. Put them side by side against a high-crude year and the cross-currents are plain:

Read the composite integrated oil company across five years, and watch the third - the high-crude year - where the cross-currents are starkest. Group profit peaked at ₹24,000 crore, and a reader glancing at the group line would call it a wonderful year driven by high oil prices. Split the segments and the truth is more interesting. Upstream EBIT - earnings before interest and tax, a segment's operating profit - surged to ₹20,000 crore, more than double its normal level, because the $95 crude price flowed straight to the value of the oil it produces. But marketing EBIT was crushed to ₹3,000 crore, half its normal level, as the same high crude squeezed the margin on regulated retail fuels. And under-recoveries spiked to ₹9,000 crore, a large subsidy the government was meant to compensate. So the bumper group profit was an upstream boom partly offset by a marketing squeeze, sitting on a big, policy-dependent subsidy - not a clean win. illustrative

Read the refining segment through its own lens and the same discipline applies. Downstream refining EBIT ran ₹18,000–24,000 crore across the years, driven by the GRM, which rose from $6 to $9 a barrel over the period. A GRM of $9 is strong, and part of that refining profit is a cyclical windfall that would reverse if the global refining cycle softened. The durable question is the refinery's complexity - its ability to process cheaper, heavier crudes and make a high share of valuable products - which earns a premium GRM through the cycle. Reading the refining profit as steady would miss that a chunk of it is a wide-GRM windfall.

Then read the whole thing through the cycle. Group profit swung from ₹14,000 crore to ₹24,000 crore and back to ₹15,000 crore over five years, with crude, GRM and under-recoveries all moving. A reader who took the ₹24,000 crore peak and extrapolated it would badly overvalue the company; one who read the through-cycle level - averaging the good years and the poor - gets a fairer picture. And the ₹50,000 crore of net debt has to be read against that cyclical profit, because leverage that looks comfortable at a profit peak can look stretched at a trough. Oil is a cyclical, multi-segment, policy-exposed business, and reading it at a single point, on a single blended number, is the mistake the whole module warns against.

The segment-and-cycle reading tells you how the parts respond to crude and GRM, but not where the cycle is going - and the crude price is famously unforecastable, driven by geopolitics, OPEC decisions (OPEC being the cartel of major exporting nations that coordinates supply to steer the price), and global demand that no analysis of the accounts can predict. Reading through the cycle disciplines you not to extrapolate a peak; it does not tell you when the next peak or trough will come, which is the thing that most determines a given year's profit.

Nor do the financials fully capture reserves accounting and the depletion of the upstream resource - the writing-down of the oil and gas reserve as it is pumped out, the resource sector's version of depreciation. An upstream business is consuming a finite reserve, and its value depends on how much oil and gas it has left, at what cost, and how those reserves are estimated and depleted in the accounts - all of which involve engineering judgement and can be optimistic. A company can show healthy upstream profit while its reserve base depletes faster than it is replaced, which the current profit does not reveal. The reserve-replacement ratio - how much new oil and gas it books each year against how much it produces, where below 100% means the reserve is shrinking - sits in specialist disclosures, and that is where the long-term sustainability of upstream is read.

And the under-recovery and subsidy mechanism is a moving policy target. How much the government compensates, when, and in what form has changed repeatedly, and a subsidy booked as a receivable can be delayed, disputed, cut, or paid in illiquid bonds. A regulated marketer's reported profit therefore rests on a policy judgement the accounts state at one point in time but that the government can alter. A reader can size the under-recovery exposure but cannot, from the numbers alone, be sure of its ultimate cash value.

When an oil company reports a strong year, a sharp analyst refuses the group number and reads the parts through the cycle. The question sounds like this: "Group profit was up on higher crude, but can you split the upstream gain, the GRM-driven refining, and the marketing squeeze - and how much of the refining profit is a cyclical GRM windfall versus your refinery complexity? And what are current under-recoveries and the compensation status?"

A good answer meets it head-on:

"Upstream added about ₹11,000 crore on the higher realisation - that's crude-linked and reverses if crude falls. Refining GRM was $9 against a mid-cycle $6, so roughly a third of the refining profit is a cyclical windfall; the rest reflects our refinery complexity, which earns a $2–3 premium through the cycle. Marketing was squeezed to ₹3,000 crore by regulated pricing. Under-recoveries are ₹9,000 crore, of which ₹6,000 crore is confirmed for compensation this year and the rest is under discussion. So read us through the cycle - the peak-year group profit isn't the run-rate."

It splits the segments, quantifies the cyclical windfall in refining, sizes the under-recovery and its compensation, and warns against extrapolating the peak. Contrast the evasive version:

"We're delighted to report record profits, reflecting our integrated business model, operational excellence and strong execution across the value chain. We remain confident in delivering strong shareholder returns and are well-positioned across the cycle."

No segment split, no GRM-versus-complexity breakdown, no under-recovery figure - and "well-positioned across the cycle" is asserted while the very cyclicality that makes the record profit unrepeatable is glossed over. The follow-up that settles it: "At mid-cycle crude and GRM, what would group profit be, and how much of current under-recoveries is confirmed for cash compensation?" That forces the run-rate and the subsidy cash risk into the open. If "record profits, well-positioned across the cycle" is allowed to stand, an investor capitalises a peak-crude, wide-GRM year as the norm and treats an uncompensated subsidy as collected profit. The silence is the tell - either the through-cycle profit is far below the record, or a large slice of the under-recoveries will not be paid in cash.

The first trap is reading a crude-driven profit jump as the whole company benefiting from high oil prices. The segments respond to crude oppositely - upstream gains, regulated marketing is squeezed - so a crude spike gilds one business and crushes another, and the group profit is their net. A reader who treats the group number as a clean bet on the oil price will be equally wrong when crude falls and the number drops even though marketing recovers. The segments hedge each other; reading them apart is the only way to see what the crude move actually did.

The second trap is admiring a high gross refining margin as a durable strength. GRM swings on the global refining cycle, so a wide GRM is usually a windfall that reverses, not a structural advantage. The durable edge is complexity - the ability to process cheaper crudes and make higher-value products - which earns a premium GRM through the cycle. Extrapolate a peak-GRM year's refining profit and you overvalue the business; read the GRM against its cycle and the refinery's complexity and you see which part of the profit lasts. The margin is cyclical; the complexity is the moat.

The third trap is trusting a regulated marketer's reported profit without reading under-recoveries. When a marketer sells price-capped fuels below cost, it books the shortfall as a loss compensated by a government subsidy, and large, rising under-recoveries mean the profit rests on a subsidy that may be delayed, partial, or paid in illiquid instruments. Take the profit at face value and you are treating a policy-dependent receivable as collected cash - and the cash risk is precisely largest in the high-crude years when the profit looks best. The subsidy is not the same as cash, and the under-recovery line is where that gap hides.