Pharma and life sciences: R&D expensed, the USFDA pipeline, and price erosion

A pharma company spends years and fortunes discovering and developing drugs, builds a pipeline worth thousands of crores, and its balance sheet barely reflects any of it. It earns a fat margin on branded medicines at home and a thin, shrinking one on generics sold into the United States - then reports the two blended into a single number that flatters the weak part and hides the strong. And one morning a regulator on the other side of the world sends a letter, and a plant supplying a third of the firm's revenue is suddenly at risk of being locked out of its biggest market. Pharma is the sector where the things that matter most - the research asset, the eroding US price, the plant-level regulatory risk - are the ones least visible in the ordinary statements. illustrative

So you cannot read a pharma firm the way you read a normal manufacturer, on its blended margin and its return ratios. You have to read what the statements obscure: the segment mix, because US generics and branded home sales are opposite businesses fused into one margin; the R&D that is expensed, so the real asset sits off the balance sheet; the price erosion that forces US revenue to be constantly rebuilt; and the USFDA pipeline and plant approvals, where the defining risk lives.

Start with the segments and the expensed research, because both are hidden in the figure below.

The segments are opposite businesses. A blended margin hides that US generics and branded home-market medicines have almost inverse economics. US generics earn a thin margin that erodes: the drug is off-patent, competitors undercut, the price falls, and the only defence is scale and constant new launches. Branded home-market medicines earn a fat, stable margin, because doctors prescribe by brand and patients stay loyal, so pricing holds. A firm that is 60% US generics and 40% branded is a far more fragile business than one with the reverse mix - even at the identical blended margin. Reading the split, its size and each part's margin trajectory, is the first move.

Before that, though, check what kind of pharma firm you are holding. A formulations player sells finished medicines to patients and carries the branded-versus-generics split above. An API or CDMO player sells the active ingredient, or contract-manufactures whole molecules, to other drug makers - so it has no branded book, its revenue tracks its customers' products rather than its own, and its risk is customer concentration and a client's molecule failing, not a home-market brand. The same three lenses apply, but the segments and the risk sit in a different place, and mistaking one model for the other misreads the whole business.

R&D is expensed, so the real asset is invisible. This is where R&D expensing - the Indian treatment of most drug research as a cost of the year rather than a capitalised asset - bites. It cuts reported profit and the asset base in the years of heavy research, so a firm investing in its pipeline looks less profitable and less capital-intensive than it is. And it inflates return on capital employed, because the denominator excludes the research investment that is the firm's most valuable asset. A high ROCE at a research-heavy pharma is partly an accounting artefact; the real asset - the drugs in development and the products approved but not yet launched - sits off the balance sheet, read in the filings, not the totals.

US generics revenue is a treadmill. Price erosion - the structural fall in generic prices after patent expiry, as competitors enter - means the US book is always shrinking in price on its existing products. A firm holds its revenue only by winning approval for the next drug coming off patent fast enough to replace the erosion on the old ones. US growth is therefore a race between new launches and price erosion, and a firm whose launches slow while erosion continues will quietly lose its US business even as it looks busy. Reading launch cadence against erosion rate is how you judge whether the treadmill is being kept up.

That treadmill and the plant risk both run through the USFDA pipeline. The USFDA is the American drug regulator, whose approval a firm must hold to sell in the United States; the pipeline is the stack of pending ANDAs - Abbreviated New Drug Applications, the filing for permission to sell a generic - awaiting clearance. Two regulatory things drive a US-exposed firm. The pending ANDAs are the leading indicator of future launches and revenue. And the inspection status of the manufacturing plants is the binary risk: an adverse finding - a warning letter escalating to an import alert - can bar a plant's products from the US market for years. Because one plant often supplies a large share of US revenue, that risk is concentrated and severe, and it surfaces in the accounts only after the revenue is already lost.

Set pharma beside an ordinary consumer manufacturer - an FMCG firm makes the cleanest baseline, brand-driven and defensive like pharma's branded book, but with none of its twists - and the departure is exact.

The inversion is that pharma's most important asset and its most important risk are both largely outside the financial statements. The research that builds the pipeline is expensed, so it is off the balance sheet, making the firm look less invested and more profitable on capital than it is; the regulatory approval that gates US revenue lives in inspection reports and import alerts, so the risk crystallises in the numbers only after it has struck. A cement maker's assets and risks are on its balance sheet and in its market; a pharma firm's are in its filings, its R&D pipeline and a regulator's decisions. You have to invert the instinct to read the statements as the business - for pharma, the statements are the faint residue of a business whose real drivers sit outside them.

Take the composite formulations exporter. Its blended picture looks solid - ₹9,800 crore of revenue at a healthy margin - but the segments tell a different story. US generics are the largest slice, ₹5,400 crore, yet earn only a 12% EBIT margin (EBIT is operating profit, before financing costs and tax) under constant price erosion. The India branded business earns 28% on ₹3,100 crore - a little more than half the US revenue, but far more profitable and stable. Rest-of-world is ₹1,300 crore at 15%. The blended margin is thus a mix of a big, thin, eroding engine and a smaller, fat, durable one, and the quality of the business sits in that mix and its direction: a shift toward branded lifts both margin and durability; growing reliance on US generics does the opposite. illustrative

The firm expenses ₹780 crore of R&D a year - a large sum, the source of its future products - none of which appears as an asset. So reported profit is struck after that ₹780 crore, understating the cash the mature business throws off, and its return on capital is flattered because the research is not in the capital base. Admiring a high ROCE here is partly admiring an accounting effect; the real question is whether the ₹780 crore is producing a pipeline of approvals to replace the eroding US products - read in the ANDA filings, not the ratio.

Then the two US-specific reads. Price erosion means the ₹5,400 crore US book is falling in price on its existing products, so the firm must keep winning new approvals just to hold that revenue; a slowdown in launches would shrink the US business quietly, with the blended P&L lagging behind. And the plant risk is binary: were the USFDA to issue a warning letter on the plant supplying most of that US revenue, and were it to escalate to an import alert, a large, high-margin slice could be cut off for years, with nothing in the current financials to warn of it. The habit to build: always split the segments, treat the expensed R&D as the real investment and discount the flattered ROCE, watch launch cadence against erosion, and read the USFDA pipeline for future revenue and the plant inspection status for the risk.

Set the annual price erosion on the US base against the revenue new launches add, then watch the book over four years. When launches outpace erosion the book grows; when erosion wins it shrinks - even while the firm looks busy launching. This race between new approvals and the falling price on the old products is the trade-off you are really reading.

When a firm reports strong US numbers, a sharp analyst probes erosion and launches: "US revenue grew, but what was the price erosion on the base, how many launches drove the growth, and is any of it a one-off - an out-licensing deal or a limited-competition opportunity?" The question separates the durable treadmill from erosion masked by one-offs, such as an out-licensing deal, where the firm sells another company the rights to a drug for an upfront payment that lifts this year's profit but does not recur.

A good answer engages:

"Fair to break it down. Base price erosion was about high single digits, in line with the market. We offset it with eleven new launches, two of them limited-competition and not fully repeatable. There was no out-licensing income this quarter. Ex the limited-competition products, US grew low single digits, and our pipeline has 40-odd ANDAs pending, including several complex generics we think are differentiated. On plants, our largest facility cleared its last inspection with zero observations."

It quantifies erosion, splits growth into repeatable launches and one-offs, gives the pipeline's quality, and volunteers the key plant's status - everything you need to judge the durable trajectory and the risk. The evasive version does the opposite:

"We're pleased with our strong US performance, driven by our robust portfolio and best-in-class execution. Our pipeline remains healthy and we continue to invest in R&D. We're confident in our compliance culture and our relationships with regulators. Momentum across markets remains strong."

"Strong US performance" without the erosion rate, "robust portfolio" without splitting launches from one-offs, "confident in our compliance culture" in place of the specific plant status - it is exactly the answer a firm whose growth was a one-off over an eroding base, or carrying an undisclosed plant issue, would give. The follow-up that forces it open: "Excluding limited-competition and one-off products, what was base US erosion and underlying growth, and what is the inspection status of each plant supplying more than 10% of US revenue?" If "strong performance, healthy pipeline, good compliance culture" is allowed to stand, the silence is the tell.

The first trap is admiring the return ratios without adjusting for expensed R&D. Because research is expensed, a research-heavy firm's profit and asset base read low and its ROCE reads high - the return looks superb precisely because the most valuable investment is not in the capital base. Rank pharma firms on ROCE as you would manufacturers and you reward the accounting, even preferring a firm that under-invests in research (and shows a higher ROCE) to one building a real pipeline.

The second trap is reading a stable blended margin as a stable business. The blended margin is the noise; the segment mix and its direction are the signal. A firm can hold the blended number while its mix deteriorates - the eroding US book growing as a share while the durable branded book shrinks, or a one-off out-licensing deal masking underlying erosion. Only splitting the segments and stripping the one-offs reveals whether the high-margin branded engine is growing or being run down.

The third trap is ignoring the plant-level regulatory risk until it strikes. US revenue hangs on USFDA approval of specific plants; that risk is binary, concentrated, and absent from the financials until the revenue is gone. A firm can look pristine while a single plant - supplying a third of its US revenue - carries an issue an import alert could crystallise, cutting off a large, high-margin slice for years. It is precisely the kind of risk that turns a quality compounder into a broken thesis overnight.

And what the reading cannot settle: the filing count tells you how many products await approval, not whether they will be approved, when, or at what price - a crowded, commodity ANDA is worth a fraction of a first-to-file complex generic, and the count captures none of that. The inspection status can look clean right up until an adverse finding, so the risk can be sized by inspection history and revenue concentration by plant, but never forecast. And the durability of the branded franchise, and whether a therapeutic area is on the right side of medical progress, are judgements about competition and science that the accounts only gesture at.