OPKO 2025: The Core Is Leaner, but Capital Is More Expensive

Company Overview

By the end of 2025, OPKO no longer looks like a broad diagnostics platform with a grab bag of pharmaceutical and biotech assets attached to it. After two sales to Labcorp, what remains is a much narrower company built around three very different engines: a downsized U.S. diagnostics business centered on New York, New Jersey, 4Kscore, and correctional healthcare; an international pharmaceutical base that sells products and receives profit share and royalties from NGENLA; and an early-stage development platform at ModeX, alongside partnership programs with Merck, Regeneron, BARDA, and Entera.

What is working right now is real. The company has shed non-core assets, materially reduced the BioReference cost base, and still generated meaningful partnership revenue. In 2025, revenue from transfer of intellectual property and other sources reached $79.7 million, including $31.9 million from Pfizer, $28.5 million from BARDA, $7.0 million from Regeneron, and $4.3 million from Eli Lilly. 4Kscore kept growing, and management is clearly telling investors that specialty diagnostics now matters more than the old scale story.

The problem is that the remaining core still does not fund itself. A superficial read could see $21.3 million of diagnostics operating income, a profitable third quarter, and $382.7 million of cash, cash equivalents, and restricted cash at year-end. That is incomplete. Diagnostics operating income only turned positive because the company booked a $101.6 million gain on the oncology sale. Strip that out, and diagnostics still posted an operating loss of about $80.3 million. The year-end cash balance also reflects asset sales and an expensive monetization of future Pfizer-linked cash flows.

That makes OPKO’s current bottleneck much clearer. The question is no longer just whether the company owns interesting assets. It is whether the smaller mix that remains can generate enough internal funding before high-cost capital, dilution, and R&D absorb the benefit of the cleanup. Over the next 2 to 4 quarters, the market is unlikely to reward one more promising pipeline headline by itself. It will want evidence that 4Kscore and the slimmed-down BioReference core are moving toward recurring profitability, that partnership revenue remains strong, and that the company can avoid another capital move that offsets what the buyback tried to return.

OPKO’s 2025 economic map looks like this:

From a business-quality perspective, OPKO still has a few assets that are hard to dismiss. 4Kscore is a branded diagnostic with reimbursement support and clinical traction. NGENLA is already approved in more than 50 markets. ModeX has been able to attract serious counterparties. But the risks are equally real. The company remains heavily dependent on partners for development and commercialization, Rayaldee competes against lower-cost alternatives, and management explicitly says OPKO has not historically generated sustained positive cash flow sufficient to fund all of its requirements.

Events and Triggers

The second Labcorp sale ended one chapter, but it also removed volume

First trigger: in September 2025, OPKO completed the sale of BioReference’s oncology diagnostics business to Labcorp. Cash consideration at closing totaled $192.5 million, including $19.2 million placed into escrow, and the company may still receive up to $32.5 million of performance-based earnout consideration. The sale generated a gain of $101.6 million in 2025.

This transaction helps the strategic story. The company is less complex, less operationally heavy, and more focused on what it now calls its core diagnostics business. But the transaction also means OPKO gave up revenue in exchange for time and capital. By the company’s own description, the two Labcorp divestitures together represented about $200 million of annual revenue before the sales. From here, the debate is no longer about what BioReference used to be. It is about whether what remains can stand on its own.

2025 was a year of financing, dilution, and buybacks at the same time

Second trigger: in April 2025, OPKO expanded its share repurchase authorization by another $100 million, taking total authorized capacity to $200 million. During the year, it repurchased 34.6 million shares for about $47.0 million. As of December 31, 2025, $112.7 million of authorization still remained.

But that is only half of the capital-allocation story. In April 2025, OPKO also completed exchange transactions in which $159.2 million principal amount of 2029 convertible notes was exchanged for 121.4 million shares of common stock and about $63.5 million of cash, including accrued interest. That reduced part of the maturity pressure, but it also produced a $32.6 million induced conversion expense and a $59.1 million extinguishment loss. Issued shares increased from 701.4 million to 789.9 million by year-end, a rise of about 12.6%.

That is the real read. Management is signaling that it views the stock as cheap enough to repurchase, but in the same year it also had to use both stock and cash to de-risk the balance sheet. So this is not a clean shareholder-return story. It is a mixed capital-allocation year in which one hand reduced float while the other accepted dilution to buy time.

ModeX keeps opening doors, but not yet closing the funding gap

Third trigger: ModeX continued to generate partnership validation. In January 2025, the Merck EBV program reached a Phase 1 dosing milestone, triggering a $12.5 million payment. In October 2025, Regeneron signed a collaboration and paid a $7.0 million upfront fee. BARDA continued to fund the platform, although mRNA-related work was de-scoped in December 2025, reducing the base contract value from $110.0 million to $103.5 million. Even after that change, $49.9 million of revenue remained allocated to unsatisfied performance obligations and is expected to be recognized through February 2028.

Those are meaningful developments because they confirm outside interest in the platform. They also look less transformative once they are pushed through the income statement. What OPKO has today is still a set of upfront fees, milestones, and reimbursed development work, not a broad commercial engine that materially closes the company’s funding gap.

Efficiency, Profitability and Competition

Diagnostics shows real cost progress, but not yet a proven earnings model

The key point is that diagnostics improved for real at the cost-base level, but not enough yet at the recurring earnings level. Service revenue fell 23% to $370.3 million, mainly because about $109.6 million disappeared with the sale of lab and oncology operations and another $10.2 million was lost from lower testing volume in continuing operations. That was partly offset by a $9.4 million benefit from better reimbursement and a $3.2 million increase from 4Kscore, which grew by about 13%.

There is also visible discipline on costs. Cost of revenue in diagnostics fell 23.6%. SG&A in diagnostics fell 37.0%. BioReference’s workforce dropped to about 1,373 full-time employees from about 3,300 before the restructuring. That is not cosmetic. It is a genuine reset of the operating structure.

But once the disposal gain is stripped out, the business still does not clear the bar. Diagnostics reported $21.3 million of operating income in 2025, but without the $101.6 million gain on sale, the remaining core would still have posted an operating loss of about $80.3 million. So the company has cut the fat, but it has not yet proven that the body left behind can generate durable profitability.

From a competitive standpoint, 4Kscore is the most important asset inside the diagnostics segment. It benefited in 2025 from the FDA label expansion that allows use without digital rectal examination data, guideline support from NCCN and EAU, and a Category I CPT code that supports reimbursement. That is why management is clearly repositioning diagnostics around this franchise. The problem is scale. 4Kscore is growing, but it is still not large enough on its own to replace what the divestitures removed.

Pharmaceuticals has real revenue anchors, but R&D still overwhelms them

Pharmaceutical revenue increased only 2% to $236.6 million. Product revenue was basically flat at $156.9 million versus $155.1 million in 2024. Favorable FX added $2.9 million, offsetting weaker volume and mix in Ireland and Chile. Net Rayaldee revenue rose modestly to $29.8 million from $29.0 million. On a gross basis, Rayaldee sales were $59.4 million, but allowances and accruals still consumed 50% of gross sales.

Revenue from transfer of intellectual property and other reached $79.7 million, slightly above 2024, but the mix changed. In 2024, the revenue base included a $12.5 million Merck milestone. In 2025, that was replaced by a broader set of contributions from BARDA, Pfizer, Regeneron, and Eli Lilly. That is constructive, because it makes the revenue stream less dependent on one event. It also reinforces that the segment still leans heavily on partner economics rather than on a broad internal commercial platform.

The bigger issue remains cost absorption. Pharmaceutical R&D rose to $121.9 million, up 18%, driven mainly by higher biologics manufacturing expense, including BARDA-related work, and higher employee costs, especially at ModeX. That creates the central paradox in OPKO’s pharma segment: it generated $143.0 million of gross margin, but R&D, SG&A, and amortization consumed more than that, leaving a $98.3 million operating loss.

That would be easier to defend if OPKO were already in a clear breakout year. It is not. 2025 looks more like a year of option building. There is an EBV vaccine in Phase 1, a Regeneron collaboration, OXM and LA-PTH programs with Entera, and BARDA-supported development. But there is still no single data point that proves the segment has crossed into self-funding territory.

Cash Flow, Debt and Capital Structure

The all-in cash picture

This is a case where the all-in cash flexibility view is the right one. OPKO is not a mature business where the main debate is maintenance versus growth capex. The real question is how much cash remains after actual uses.

That 2025 bridge is straightforward. The company started the year with $445.6 million of cash, cash equivalents, and restricted cash and ended with $382.7 million. In between, it used $178.5 million in operating activities, generated $230.3 million from investing activities, and used $118.1 million in financing activities. Inside investing cash flow were $197.8 million from the Labcorp transactions and $52.2 million from the sale of equity securities. Inside financing cash flow were $47.0 million of share repurchases and $62.2 million of convertible note repurchases.

That means the year-end cash balance does not tell a clean story of internally generated liquidity funding the business. It tells a different story: asset sales and security sales funded operating burn and capital structure moves. That is not automatically bad. It does mean OPKO’s financial flexibility still depends more on monetizations and partner flows than on recurring operating cash generation.

The expensive debt sits on one of the company’s best assets

The most important balance-sheet point is not just how much debt OPKO has. It is how that debt is structured. In 2024, OPKO issued $250 million of 2044 Notes, and the year-end carrying amount in 2025 was $246.4 million. Those notes bear interest at 3-month SOFR plus 7.5%, subject to a 4.0% floor, and the weighted average interest rate at year-end was about 12.11%. That is expensive. It gets even more expensive if repaid early, because the structure includes a 3% exit fee and a make-whole that can push total payments to at least 150% of the original principal if repaid on or before July 17, 2029, or 200% after that.

Why does that matter? Because the notes are secured by OPKO’s profit share payments from Pfizer under the NGENLA arrangement. In other words, OPKO monetized one of its best and least speculative assets into high-cost debt. That improved liquidity today, but it also means a meaningful share of the future value from NGENLA is already carrying a financing burden.

The convertible exchange bought time, but at a real price

At the end of 2025, the 2029 convertible notes represented $121.4 million of principal, or $85.0 million on the balance sheet after discount and issuance costs. In April 2025, the company exchanged $159.2 million principal amount of those notes for 121.4 million shares of stock and $63.5 million of cash. That reduced part of the 2029 overhang, but it also produced immediate P&L damage and meaningful dilution.

This is a classic two-sided capital move. One side says the company reduced near-to-medium-term pressure. The other says common shareholders paid for that relief with both stock and cash at a time when recurring positive cash flow is still missing.

Book equity looks large, but a big part of the asset base is intangible

Year-end shareholder equity stood at $1.268 billion. On the surface, that looks comfortable. But the balance sheet also carried $516.3 million of intangible assets, $195.0 million of in-process research and development, and $484.3 million of goodwill. Together, that is about $1.196 billion, or roughly 61.9% of total assets. So anyone reading the headline equity line as a sign of an unusually strong balance sheet is missing the quality mix. There is substantial embedded value here, but much less hard liquidity cushion than the top-line equity number suggests.

Forward View

Before getting into the forward case, four points matter most going into 2026:

• First finding: diagnostics still has not proven recurring profitability. The 2025 reported profit came from a disposal gain, not from the remaining operating base.
• Second finding: NGENLA and the broader partner revenue base do generate real cash, but part of that value has already been pledged or monetized through expensive debt.
• Third finding: ModeX is attracting serious partners, but today it is still more a funding-and-validation platform than a proven earnings engine.
• Fourth finding: the buyback did not solve dilution. In 2025, it coexisted with dilution rather than replacing it.

2026 looks like a proof year, not a breakout year

If the next year needs a label, it is a proof year. Not a full reset year, because OPKO has already done real cleanup. But not a breakout year either, because most of the value creation case still sits in milestones that have to arrive.

In diagnostics, the first test is whether 4Kscore and the leaner New York and New Jersey lab base can absorb the revenue hole left by the oncology sale. Management itself frames the near-term goals clearly: grow 4Kscore, expand share in the core regional markets, and achieve sustained profitability through a leaner model. That is the right direction. The problem is that 2025 still does not provide the numerical proof that this transition has already landed.

The quarterly pattern makes that point even sharper. Revenue held within a relatively narrow range of $148.5 million to $156.8 million through the year, but net income was negative in every quarter except the third quarter, when the company reported $21.6 million of profit. That was the quarter in which the oncology sale closed. The fourth quarter moved back to a $31.3 million loss. In other words, quarterly profitability still depends on capital events rather than on a stable operating baseline.

The pharma pipeline now needs conversion, not just validation

On the pharma side, three things need to happen for the story to clean up. First, the Pfizer and NGENLA profit-share stream needs to keep growing beyond the $31.9 million booked in 2025. Second, ModeX and Entera programs need to continue progressing without R&D growth outpacing associated collaboration income. Third, Rayaldee needs to generate more international value through Europe, Macau, and China rather than remaining mostly a U.S. product with a heavy allowance structure.

The pipeline does have visible milestones. BARDA still carries $49.9 million of unrecognized revenue through February 2028. In February 2026, the Entera relationship was expanded to include LA-PTH, with a 50-50 ownership and cost-sharing structure, and the parties expect to file an IND in late 2026. The oral OXM program is also expected to progress following development work already underway. That is a real trigger set. It just does not yet add up to a self-funding operating model.

What the market could misread

The market could miss two opposite things. On one side, it could reduce 2025 to a pure asset-sale story. That would be incomplete, because there is real underlying repositioning here, 4Kscore is growing, and external parties continue to validate the pipeline. On the other side, it could also assume the cleanup is already complete. That would be wrong too, because the remaining core still has not demonstrated that it can support overhead, R&D, and financing costs without help from monetizations.

That is why the next few reports need to be read through three hard checkpoints rather than through one more strategic headline: the operating loss in the remaining core needs to narrow, partnership and royalty revenue needs to hold up or improve, and the financing burden needs to stop forcing new dilution or aggressive capital moves. If OPKO can deliver two out of those three, the story will look materially cleaner. If not, 2025 will look like another year in which the company bought time without fully fixing the model.

Risks

Funding pressure remains the central risk

Management explicitly says the company has not historically generated sustained positive cash flow sufficient to offset its operating and other expenses. That is not boilerplate here. It means the real question is not how much cash is on the balance sheet today, but how quickly that cash erodes if partnership revenue, 4Kscore, and Rayaldee do not ramp faster. The Chilean and Spanish credit lines are modest in the context of the full company, with $8.5 million drawn against $30.4 million of total commitments, so they do not change the big picture.

Financing cost and dilution can come back into the story quickly

The 2044 Notes are expensive, and the 2029 convertibles are still there. Yes, the company reduced some maturity pressure through the exchange transactions. It also paid a high price to do it. If the remaining core does not improve quickly enough, it will be hard to argue that the combination of buybacks, note exchanges, and royalty-backed debt truly improved common-shareholder economics. Mostly, it bought time.

Legal and regulatory risk is not a side note

In the commitments and contingencies note, OPKO discloses an Israel Tax Authority assessment of about $246 million, including interest, against OPKO Biologics relating to 2014 through 2020 tax years. The company has appealed, the trial has concluded, and judgment is still pending. If the outcome is adverse, the effect on financial condition and cash flow could be material.

BioReference also remains subject to a Corporate Integrity Agreement signed in 2022 with a five-year term, and the company says it is responding to CIDs, subpoenas, payor audits, and document requests related to laboratory operations. There is no accrual for those matters beyond what has already been settled, but that is not the same thing as saying the risk has disappeared.

Competition, reimbursement, and FX still sit inside the model

4Kscore still depends on physician adoption and reimbursement durability. Rayaldee competes against lower-cost alternatives. NGENLA depends on Pfizer’s ability to commercialize successfully across more than 50 markets. At the same time, about 28% of 2025 revenue was denominated in currencies other than the U.S. dollar, primarily the Chilean peso and the euro, and year-end open forward contracts related to inventory purchases totaled $13.6 million. FX exposure does not break the thesis by itself, but it adds volatility to an already complex earnings model.

The local market signal is cautious, not crowded

On the local trading line, the stock does not look heavily crowded on the short side. Short float was about 0.54% at the end of March 2026, low in absolute terms even if above the sector average of 0.30%. SIR was 5.4 days, which looks more like a sign of limited local trading depth than of an aggressive short consensus. The implication is that the market is not fully positioned for failure, but it is also not offering OPKO the benefit of the doubt.

Conclusions

By the end of 2025, OPKO is structurally cleaner, but not yet economically clean. The company still owns real assets: 4Kscore, NGENLA, a validated partnership network around ModeX, and international pharmaceutical platforms with actual revenue. The central blocker, however, has not gone away. The remaining core still does not generate enough earnings and cash to remove the need for monetizations, expensive debt, or capital-structure engineering.

Current thesis in one line: OPKO has moved from an oversized diagnostics story to a narrower diagnostics-and-partnered-pharma platform, but it still has to prove that this smaller core can fund itself.

What has changed most is not just the portfolio. The conversation has shifted from what OPKO could sell to how the assets that remain actually monetize and how much of that value survives after financing costs. The strongest counter-thesis is that the post-sale core is better than the market gives it credit for, and that 4Kscore, NGENLA, BARDA, Regeneron, Merck, and Entera together can turn 2025 into a short transition year before a much cleaner operating profile emerges. That is not an unreasonable view. It simply has not yet been proven in recurring numbers.

What could change the market read over the short to medium term is a sequence of two or three quarters in which diagnostics stays close to breakeven without disposal gains, partnership revenue remains strong, and no new dilutive or highly punitive financing move is needed. What would weaken the story is a return to deep recurring losses, stagnation in 4Kscore, or further evidence that future value keeps getting sold early to fund the present.

Why does this matter? Because OPKO sits right on the line between value created in products and partnerships and value that actually remains available to common shareholders after debt, dilution, and development spending. Over the next 2 to 4 quarters, the company needs to show that its new shape brings economic discipline, not just strategic cleanup.