Revenue Up 5.5×, a 34.9bn KRW Net Loss — What Does T&R Biofab Actually Earn On?

T&R Biofab's 2025 results carry two headlines at once. Revenue went from 4.87bn to 26.97bn KRW — a 5.5× jump — and in the same year the net loss was 34.9bn KRW with a debt-to-equity ratio of 394%. Read the first line alone and it's a growth stock; read the second alone and it's a distressed, loss-making biotech.

Both are precisely wrong. The growth is a misread because half of it was bought; the crisis is real, but it sits neither in operations nor in the size of the loss. To read this stock accurately you have to decompose three things — where the loss came from, who earned the revenue, and what the market cap is riding on. Decompose them and a gap appears: the story that floats this company (3D bioprinting, an FDA cranial implant, a 2-trillion-KRW market) and the engine that actually runs it (biosurgical, cosmetics, cost-cutting) sit in different places. This is a stock analysis, not investment advice.

Build the Bull Case at Full Strength First

The decomposition starts by cutting down the bull case — so before cutting, it's only fair to build the strongest version of it in full. And the heart of the bull case isn't cosmetics — it's an option.

Cosmetics isn't the point but the stepping stone — T&R Biofab buys that cash cow to generate cash flow, and uses the cash to carry real medical assets. Its cranial-regeneration implant (TnR CFI) has been used in more than 17,000 procedures since its 2021 domestic approval, with no reported adverse events. The losses are shrinking fast: the Q1 2026 operating loss was 0.93bn KRW, a quarter of the 3.58bn KRW a year earlier. The Street sees a swing to profit in 2026 — 46bn KRW in revenue, 1.3bn KRW in operating profit. The mezzanine wall — mezzanine being the in-between debt that carries an equity-conversion right, like a convertible bond (CB) or a bond with warrant (BW) — was held off with a rights (paid-in capital) increase. And on top of all this sits a call option named America: clear the FDA review and the global 2-trillion-KRW market management points to opens up. The bull case's real claim is this: "the market cap that once reached 543bn KRW is the probability-weighted value of that FDA option, not a bubble." Buying a real regulatory option cheap, on a company about to turn profitable — that's the ceiling of the bull case.

The data backs this framing. The point where it breaks is just one: can you price that option's odds as high as the market has? To see that, you first have to pin down what the loss and the revenue actually are.

The Loss Isn't Operating — and the Crisis Isn't Its Size

The 34.9bn KRW net loss in 2025 didn't come out of operations. The operating loss is in the 9bn KRW range. The roughly 26bn KRW between net loss and operating loss came out of the financing line. Finance costs — including derivative valuation losses — ran as high as the 40bn KRW range, by press accounts. (That finance costs exceed the net loss means offsets are mixed in — valuation gains, other income. So you can't just add the figures up.)

Go one level deeper. The nature of this derivative valuation loss matters. If the mezzanine's conversion right is structured to track the share price, then in accounting terms that right is booked as a derivative liability and throws off a valuation loss when the price rises. Which is to say: the enormous 2025 valuation loss may be a non-cash by-product of the fact that the share price climbed to a 543bn KRW market cap that year. Not liquidity drying up, but the accounting shadow of a rising price. Flinch at the "34.9bn KRW loss" number and you miss this.

So where is the real crisis? Not the size of the loss, not the valuation loss — cash. The rights (paid-in capital) increase used to fend off the mezzanine early-redemption, estimated between 22.4bn and 45bn KRW, came to 22.7bn KRW, of which only 16.8bn KRW went to debt repayment. The rest, in other words, was covered out of cash on hand. Cash was drawn down by that much. The debt-to-equity ratio is 394% (end of H1 2025); the past loss ratio has a history of running up near the designated-issue (delisting-watch) threshold (48%, 2023); and these are figures from after the 2021 expiry of the loss-requirement grace period granted to tech-special-listing firms (Korea's tech-exception IPO track) — so they aren't to be taken lightly. The crisis isn't "the loss is large," it's "cash is drying up."

Revenue Up 5.5×, Half of It Bought

Decompose the growth headline and it, too, reads differently. Revenue rose by 22.1bn KRW, from 4.87bn to 26.97bn KRW — and nearly half of that is cosmetics. Cosmetics revenue, zero in 2024, was booked at around 12bn KRW in 2025. That's the result of consolidating Blisspack — a freeze-dried-ball cosmetics OEM — acquired for about 16.7bn KRW. The other half looks to be what the core business grew (non-cosmetics revenue went from roughly 5bn to around 15bn KRW). Half bought, half earned.

But what pulled the company to the profit threshold is neither cosmetics nor the headline 3D bioprinting. What drove the Q1 narrowing of losses was ECM (extracellular matrix)–based biosurgical — that is, the medical-device business — helped by SG&A cut 36% as the Pangyo research lab was folded into headquarters. Cosmetics is betting on revenue ramping in earnest in the second half, and Blisspack's second plant raised capacity (CAPA) by 35%.

Look once more at the "bought half" and a bill that usually comes attached is missing — goodwill. Normally, when you acquire a company at a premium, goodwill piles up on the books, and if the subsidiary underperforms you impair that goodwill, eating into equity. T&R did the opposite. Buying Blisspack — a distressed turnaround company that had drawn disclaimers of audit opinion in 2022 and 2023 — below the fair value of its net assets, it booked not goodwill but a 6.97bn KRW bargain-purchase gain in 2024, recognized in one shot as non-operating income. So consolidated goodwill is zero, and the common scenario — "H2 ramp misses → goodwill impairment → further equity erosion" — doesn't exist for this stock. Instead the bill sits in two other places. One, that 6.97bn — readable as the skill of buying a distressed firm cheap, but either way a one-off that won't recur — flattered 2024's bottom line by that much. Two, what was actually bought was a plant running at half capacity in H1 2025 (freeze-dried-ball line 49.7%, blister 41.1%) — though that half partly reflects the second plant, just expanded by 35%, enlarging the denominator, so the H2 ramp the bull case counts on and the low utilization the bear case fears are two sides of the same number. The 2026 swing-to-profit guidance hinges on whether the H2 cosmetics ramp fills that empty capacity. The intangibles that could be impaired are in the ~1.1bn KRW range, under 5% of equity — so the real risk isn't a write-down on the books, it's whether that plant actually runs.

The Headline's Medicine and the P&L's Medicine Are Different Products

This is where readers most often slip on this stock. T&R Biofab's medical business isn't fake. The problem is that the medicine the market prices and the medicine that earns the P&L are different products.

The side that earns the P&L is ECM biosurgical; the cranial implant (CFI) is a small stream, used in 17,000 procedures domestically. Yet the headline that floats the market cap is "3D bioprinting" and "an FDA 2-trillion-KRW market." One thing to flag: CFI isn't bioprinting that prints living cells — it's a synthetic scaffold, a biodegradable polymer (PCL) and a bone-mineral component (β-TCP) printed into three-dimensional form. True bioprinting that prints cells is at a far earlier stage than that. Not only are the headline technology and the P&L product different — the headline technology itself is the least mature.

The FDA option has to be read precisely too. The 510(k) CFI filed isn't a new-drug approval (PMA) — it's a premarket notification that demonstrates "substantial equivalence" to a product already on sale. The bar is lower than a PMA, and with 17,000 incident-free real-use cases behind it, the "pass-or-reject risk" isn't as large as the market fears. A 510(k) is usually delayed by a request for more information, not denied — it gets held up, not blocked. But for the same reason, the other side holds too: a 510(k) is a clearance to sell, not revenue. To actually sell in the US, a reimbursement code, hospital adoption, and distribution have to open up separately. The option has both a high probability of passing and, even if it passes, one more step before it reaches revenue.

The Multiple Rides on an Option — and You Have to Price Those Odds Yourself

So we come back to the market cap. What turns this year's P&L is biosurgical, cosmetics, cost-cutting — but the story that justifies a premium that reached 543bn KRW is FDA and bioprinting. The medicine of the P&L and the medicine of the valuation sit in different places.

The common bear case here ("a loss-making company trades on a story") can't stop there. An option by definition sits outside current P&L, and dismissing it as "no revenue" is the static misreading that calls a call option a bubble because the current price is below the strike. The honest question isn't "is it overvalued" but "can you price that option's odds as high as the market has." And there are clear factors that cut those odds.

First, the reference point. 543bn KRW was a mid-2025 high, and the current price of 1,807 KRW sits just above the 1,425 KRW rights-issue subscription price. The market has already marked its expectations down substantially. Use the old high as the benchmark and the gap is inflated.

And widen the benchmark from its own high to its peers, and the picture flips once more. Since it's loss-making, P/E is meaningless, so use price-to-sales (P/S): T&R's current market cap of 86.1bn KRW is about 3.2× its 2025 revenue of 27bn KRW, and about 1.9× against the 2026 estimated revenue of 46bn KRW. Most of the air has gone out of a peak P/S that once topped 20×. Line it up against domestic regenerative-medicine and tissue-engineering peers and it's clearer still — Hans Biomed, the only profitable one, at 3.9×, Medipost 4.3×, Corestemchemon 5.2×, Nibec 5.5×, a median around 5× — and T&R sits below that, below even the profitable peer. (The story stocks in the table — Anterogen and L&C Bio — carry an option premium baked in, so they're left out of the median and looked at separately below.)

So on a sales multiple, at least, T&R isn't expensive — if anything it's cheap. (That denominator, of course, still includes the cosmetics we called the "bought half"; discount the quality of the revenue too and the "cheap" shrinks.) But that cheapness is less an undervaluation opportunity than the market having pre-discounted the loss-making, the debt, and the dilution risk into the sales line. So the premium that remains rides not on the sales multiple but entirely on the option. What it looks like when the option pays is shown by the very names left out of the median — Anterogen and L&C Bio, trading at 25–28× on the same "approval opens a market" story. The direction of the multiple rides not on this year's revenue but on that one card.

Back to what cuts the option's odds — second, a history of the same promise repeated. The revenue estimate offered at the 2018 IPO was close to 40bn KRW, and for a while actual results came in at a fraction of that (the exact multiple depends on the benchmark). To be fair, though, 2025's 26.9bn KRW and 2026's estimated 46bn KRW are also converging on that promise, late as it is. Less a broken promise than a promise seven years late. What's worth noting, though, is that 2018's "big revenue soon," today's "FDA opens a 2-trillion market," and "the P/E gets re-rated on organoids" are the same frame — "approval opens a vast market." Not a problem unique to this stock; the shape of this promise repeats.

The More It Becomes a Cosmetics Company, the More It Needs the Regenerative-Medicine Story

At the end of the decomposition, the stock's real structure comes into view. The more the P&L tilts toward cosmetics OEM and biosurgical, the more a regenerative-medicine and FDA narrative is needed to win back a once-543bn-KRW-class premium. Because the market assigns a low multiple to cosmetics revenue and a high multiple to a regenerative-medicine option. The bull case reads this in reverse — laying boring cash generation under the option is de-risking that lowers the valuation's dependence on FDA. Both are possible readings, but as long as the market discounts cosmetics cash flow, narrative dependence doesn't fall, it rises. Not because anyone intends it, but because the structure of the premium keeps pulling that story back in.

When the narrative keeps getting reissued like that, one branch of the weight has a clear bearer — the mezzanine and the valuation loss shift onto a 1,425-KRW dilutive rights issue for existing shareholders, and when the price falls the conversion price steps down into more shares, a cascade of dilution riding along. But the one who has to answer for the promise written into the listing eight years ago is empty — the mechanism to hold the gap between estimate and result to account is fragile and almost never enforced, and there's no formal target-price consensus either, so the one who bears the cost and the one who answers for the promise stay split apart while a promise of the same shape gets issued again.

So you have to step one foot past the static "who carries it" — once you've decomposed it this way, the stock lands on what to watch and prepare for (not a verdict, but signals you can count as wrong if they fall the other way). The direction of the multiple rides not on this year's revenue but on a single card (FDA), and that card is more likely to resolve as a spectrum of delay than a "pass-or-reject" binary — because a 510(k) is usually held up by a request for more information, not denied. So the watch points for the next year are four: ① split the loss into operating and (non-cash) financing, and track the real crisis — cash headroom — through the remaining mezzanine balance and signs of further dilution; ② does H2 cosmetics revenue actually fill the capacity stuck at half (freeze-dried-ball 49.7%, blister 41.1%)?; ③ price the FDA less on the probability of passing than on whether reimbursement and market access open up after it passes; ④ the remaining mezzanine and refixing-driven cascade dilution. Which way these signals fall decides whether the multiple converges toward the peer median (around 5×) or, if the option pays, re-rates toward the story stocks (25–28×) — neither settled yet, which is exactly why they're worth watching. Not putting where the headline points (FDA, a 2-trillion-KRW market) and where the P&L turns (cosmetics, biosurgical, cost-cutting) on the same line — that's the starting point for reading these signals accurately.