In its recent regulatory filings, PureTech Health uses a phrase that would make a traditional scientist shudder: “program-killing experiments.” This isn’t accidental language. It’s a precise description of what running a modern drug company actually requires, the systematic and unemotional termination of programs that don’t meet the bar, before they consume capital that belongs to the ones that do. In 2026 running a biopharmaceutical company and managing a portfolio of high-risk bets are no longer distinguishable activities. The most effective CEOs now think like fund managers. The most effective VCs act like lab heads. Understanding why this happened and what it means for where value is being created, matters whether you’re allocating capital, building a company, or evaluating one.

The CEO as Portfolio Manager: The Hub-and-Spoke Revolution

The traditional biotech model, raising $100M, chasing a single molecule for a decade, hoping for a binary outcome, is not dead, but it is increasingly the exception. In its place, the hub-and-spoke structure has emerged as what capital efficiency actually looks like in practice. Companies like PureTech and BridgeBio originate programs centrally, run early de-risking experiments internally, and then spin programs out as independent entities supported by external financing once the science reaches a meaningful inflection point. The hub retains equity in the spokes. The spokes absorb the expensive late-stage costs. The result is a structure that looks less like a drug developer and more like a portfolio with recurring optionality.

Source: L.E.K. Consulting, https://www.lek.com/sites/default/files/PDFs/2420_hub-spoke.pdf, Figure 3, p.3

What the table above makes clear is that while each firm executes the model differently, Roivant in-licenses external assets, Centessa was formed through the merger of ten private biotechs, the common logic is the same. Subsidiaries stay focused. The parent allocates capital, provides infrastructure, and retains long-term economic rights. The model incorporates best practices from both venture capital and big pharma, and in doing so begins to blur the boundary between them.

The numbers from PureTech and BridgeBio make the financial logic concrete. PureTech’s hub-and-spoke model produced one of the most scrutinised return profiles in recent biotech history: an initial allocation of $18.5 million to Karuna Therapeutics, which following its acquisition by Bristol-Myers Squibb generated over $1 billion in proceeds back to PureTech shareholders. What makes that figure meaningful is not the size of the return but the discipline it required. Karuna was de-risked early, spun out with external financing, and the hub retained equity without continuing to absorb the full cost of development. That’s not luck, it’s a repeatable process.

https://www.sciencedirect.com/science/article/pii/S135964462500296X

BridgeBio offers the scale version of the same logic. Their model is built around autonomous teams focused on individual genetic conditions, with a central hub providing clinical, regulatory, and commercial infrastructure. As of their February 2026 results, they closed their first decade having delivered three successful Phase 3 readouts in just over three months, with Attruby posting 35% quarter-over-quarter growth in net product revenue in Q4 2025 and the company hitting $502 million in full-year revenues. BridgeBio even published a peer-reviewed manuscript in Drug Discovery Today formalising the hub-and-spoke approach, an unusual move for a commercial-stage biotech, but one that signals they consider the structure itself to be part of their competitive advantage, not just the molecules inside it.

For the modern CEO, capital allocation, deciding precisely when to kill an asset, when to spin it out, and when to hold equity, is now a core scientific leadership function. It has not been delegated to the CFO. It sits at the top of the org chart, because getting it wrong in either direction destroys value: over-invest in a failing program and you exhaust the runway; kill the right asset too early and you’ve thrown away the return. The “program-killing experiment” is, in this sense, the most important experiment the organisation runs.

The VC as Operator: The CEO-Partner Era

While CEOs are becoming more financial in their orientation, VCs are becoming more operational. Flagship Pioneering is the clearest example of how far this has gone. The firm describes itself not as a venture capital firm but as a “scientific innovation engine,” and this isn’t just marketing language, it reflects a genuinely different operating model. Under their VentureLabs structure, scientists are hired internally, ideas originate from scratch within the firm, and companies are built from the inside out. Flagship didn’t back Moderna in the conventional sense; they created it in 2009. That distinction matters. It means the decision to build Moderna was made by people managing a fund, using the same judgment framework they would apply to any allocation decision.

The most recent signal of where this is heading came in January 2026, when Flagship appointed Adrian Rawcliffe as CEO-Partner at their company Etiome. Rawcliffe is not a junior hire: he led Adaptimmune to the first engineered TCR T-cell therapy approval for a solid tumour, oversaw roughly $1.3 billion in capital raises, and guided the company through its IPO. Before Adaptimmune, he spent more than 15 years at GSK in senior global finance and business development roles. The title CEO-Partner is deliberate, he is simultaneously an operating executive inside a Flagship company and a partner within the fund itself. That structural fusion is the point. The firm isn’t hiring a CEO and handing them a company; they are importing a world-class operator into the investment process because they believe those two roles are now, functionally, the same job.

Third Rock Ventures has followed a similar pattern, with partners routinely stepping into interim leadership roles at portfolio companies during critical transitions. The logic is the same: a partner who has operated understands the difference between a funding problem and a science problem in a way that a partner who has only deployed capital may not.

The Shared Language: Bilingualism as the Real Moat

The common thread between a PureTech executive and a Flagship partner isn’t sector expertise or deal experience, it’s a specific kind of bilingualism. They move fluidly between the language of science (p-values, probability of technical and regulatory success, mechanism of action, trial design) and the language of capital (IRR, cash runway, dilution, portfolio concentration). Neither language is sufficient on its own. A CEO who can read a Phase 2 trial but can’t model the capital implications of a missed endpoint will over-commit to the wrong asset. A VC who can underwrite a deal but can’t evaluate a trial design will misprice the risk fundamentally.

This convergence is being accelerated by AI. According to the PwC 2026 Pharma Deals Outlook, AI-enabled diligence has become standard across major transactions, with predictive models compressing decision cycles for M&A by allowing scientific and financial evaluation of an asset to happen simultaneously rather than sequentially. When a biotech submits to an acquirer, the buyer now has access to real-time clinical and financial modelling that used to take months to assemble. This doesn’t make the bilingual operator redundant, it makes them more valuable, because the bottleneck has shifted from information gathering to interpretation.

Even large pharma is being forced into this mindset. Novartis faces approximately $4 billion in global revenue at risk from upcoming patent expirations, and their leadership must make VC-style buy-versus-build decisions, with the same cold efficiency as a seed-stage investor about which pipeline gaps to fill externally and at what price. The patent cliff, as covered in a previous piece, isn’t a finance problem or a science problem in isolation. It’s both, simultaneously, which is exactly why the bilingual leader has become such a rare and valuable asset across the entire industry.

Investor Playbook: What to Look For

For investors, this structural shift provides a practical lens for due diligence that goes beyond pipeline depth or cash position.

The first signal is the kill rate. Does the company actually shut down programs? If a CEO hasn’t discontinued a pipeline asset in the last 24 months, there is a meaningful chance they are managing to sink costs rather than to portfolio logic. Program terminations are not failures in a hub-and-spoke model, they are confirmation that the discipline is working.

The second is the equity retention structure. Look for companies that maintain meaningful ownership in their spin-outs while letting external partners, including Big Pharma, carry the cost of late-stage development. This is the exact mechanism that produced the Karuna outcome at PureTech. The hub creates and de-risks; the spoke absorbs the Phase 3 spend; the hub collects the upside. Companies that understand this structure should be able to articulate it clearly in their investor communications.

The third is the operator-to-investor ratio at the VC funds you are evaluating. A fund whose partners have collectively run clinical programmes, navigated regulatory submissions, and managed a company’s cash runway at genuine risk will make different and generally better decisions than one composed primarily of financial professionals. The Rawcliffe appointment is a useful benchmark: that’s the calibre of operator that the best funds are now bringing inside the tent, not just onto their advisory boards.

Conclusion

The role reversal described here isn’t a trend in the sense of something temporary. It’s a structural response to a market where capital is expensive, regulation is demanding, and the cost of being wrong about a single asset is severe. The CEO who treats their pipeline as a collection of scientific projects rather than a portfolio of bets will, eventually, run out of runway. The investor who cannot evaluate a trial design independently is relying on proxies that will consistently misprice risk.

PureTech’s own filings say it plainly: early program-killing experiments are the mechanism behind their greatest commercial successes. $18.5 million in, $1.1 billion out, from a process designed to be ruthless about what deserves capital and what doesn’t. In 2026, the kill decision, made fast and made correctly is the defining competency of healthcare leadership, regardless of whether you sit on the operating or the investing side of the table.

This article is for informational and educational purposes only and does not constitute financial, investment, or legal advice. I do not provide personalized investment recommendations or suggest the purchase or sale of any specific security. Investing involves substantial risk; please do your own due diligence or seek professional advice before making any financial decisions.

With over 6,000 diseases classified as rare across the world and an estimated 300 million people living with a rare disease globally (1.), this is a market far larger than many assume. Rare diseases still mean that those 300 million patients are split into extremely small patient pools, sometimes only a few hundred worldwide. Historically, these conditions have been ignored by large pharmaceutical companies because tiny patient populations combined with long and complex R&D timelines made development high risk. For investors looking at rare-disease markets, it is essential to understand how the market operates, which scientific factors matter most and how individual companies approach development.

The unmet need and treatment gap

Despite the scale of rare diseases worldwide, only a small fraction have an available treatment. Recent estimates suggest that about 90 to 95 percent of rare diseases have no approved therapy (2.). What counts as a rare disease also varies by region. In the EU, a rare disease must affect fewer than 5 in 10,000 people. In the United States, a disease is considered rare if it affects fewer than 200,000 Americans (3.).

Historically, the combination of small patient populations and high scientific risk left these diseases underserved. The Orphan Drug Act in the US began shifting this dynamic by granting orphan designations that offer market exclusivity and development incentives (4.). With better technology, more efficient development pathways and broader insurance coverage, the market has become more accessible for patients and more economically feasible for companies. The resulting unmet gap creates an opportunity where investors can support innovation and benefit from the structural demand.

Figure 1. Growth in orphan drug designations over time, reflecting increased regulatory support and improving commercial feasibility for rare disease drug development.

Source: Pharmaceutical Technology, based on FDA and EMA orphan drug designation data.

Why rare diseases are now more accessible from a scientific view

The completion of the human genome and the rise of affordable sequencing have transformed detection and diagnosis. Approximately 70 percent of rare diseases are genetic and many involve single-gene mutations (5.). This makes them scientifically attractive because single-gene disorders provide a clear molecular target, unlike pathogens or multifactorial diseases. Bacteria can adapt rapidly and viruses like SARS-CoV-2 mutate frequently, forcing constant redesign of interventions. Single-gene disorders do not exhibit this type of variation, making scientific risk lower.

Many rare diseases also appear in childhood and have well documented phenotypes and natural histories, which helps design targeted, efficient and comparatively inexpensive clinical trials. Modern biotech now allows development of drug candidates faster than ever before and with trial designs that can include pre-planned adaptations. These improvements reduce both cost and risk and make rare-disease programs more predictable.

The Overlooked Bottleneck in Orphan Drugs: Manufacturing Efficiency

In rare disease drug development, scientific validity and regulatory incentives often dominate the investment narrative, but manufacturing efficiency is an equally critical and frequently underappreciated determinant of commercial success. Orphan drugs typically serve very small patient populations, which translates into small batch sizes, limited production runs, and minimal tolerance for error, making every batch failure disproportionately costly in terms of time, capital, and patient access. Enzyme-enabled manufacturing approaches can help address these constraints by improving reaction specificity, increasing yields from small batches, and reducing downstream purification requirements. In orphan-drug settings, where volumes are limited and timelines are critical, such process efficiencies can materially reduce development risk and improve economic viability. For small biotech teams operating with tight runways, long purification timelines, low yields, or inconsistent reproducibility can delay clinical progress and materially erode the economics of an otherwise promising asset. This is where process optimisation becomes a strategic advantage rather than a technical afterthought. Enzyme based manufacturing improvements, for example, can help optimise yields from small batches, reduce variability, shorten cycle times, and lower purification burdens, all of which improve throughput and reliability. Aminoverse illustrates how such enabling technologies can support rare disease programs by helping companies reach the clinic faster with a more robust and scalable manufacturing process. In a setting where small volumes magnify risk and stress, an optimised production workflow not only improves cost efficiency and revenue yield per batch, but also de-risks development timelines, benefiting patients, developers, and investors alike.

Market and pricing

The orphan-drug market is currently valued at around USD 217 billion and is projected to double within the next decade (Precedence Research). What makes this market so attractive is the combination of high pricing and limited competition. Because orphan drugs serve small populations, prices per patient are significantly higher. The few patients receiving treatment essentially carry the R&D cost that in mass-market drugs would be spread across millions of users.

Figure 2. Projected growth of the orphan drugs market from 2025 to 2034, illustrating a steady increase in market valuation reaching over $610 billion. Estimations and data from Precedence Research.

Source: Precedence Research

Additionally, orphan drugs typically receive 7 to 10 years of market exclusivity, meaning that no direct competitors can enter the market during that period (FDA and EMA orphan frameworks). This creates a monopolistic situation where one product effectively dominates an entire disease area. Patients must choose between receiving the drug or waiting many years for a generic alternative after exclusivity expires.

With this monopoly and the life changing benefit many orphan drugs provide, payers including US insurance companies, the NHS and European insurers often reimburse the majority or all of the drug cost. The willingness to pay reflects both clinical value and the societal decision that rare-disease patients should not be left untreated.

For investors, these factors make the orphan-drug sector appealing, although the same features also introduce risk. Small patient bases mean that any external shift can cause major revenue swings. Loss or denial of orphan designation can remove key protections. Ultimately, the orphan-drug market relies on a combination of political trust, regulatory consistency and detailed understanding of patient populations.

Conclusion

The orphan-drug market offers significant opportunities for investors seeking exposure to a fast growing healthcare segment. Premium pricing combined with small, well understood patient bases means each individual patient has a measurable impact on company performance. Successful orphan-drug development blends regulatory analysis, patient-population strategy and scientific excellence to create a viable product. Although the market looks promising, it requires investors to understand the political, scientific and economic factors behind each company’s pipeline as well as the customers it’s aiming to serve.

With increasing investment and scientific advancement, companies can continue reducing the clinical burden of rare diseases, improving patient outcomes while offering investors a market with unmet demand, premium pricing and regulatory protection.

References

1. https://www.eurordis.org

2. https://www.nature.com/articles/s41591-023-02207-9

3. https://www.fda.gov/patients/rare-diseases-fda

4. https://www.pharmaceutical-technology.com/features/rare-disease-spotlight-tracing-the-rise-of-orphan-drug-designations-over-almost-40-years/

5. https://pubmed.ncbi.nlm.nih.gov/31527858/

Figure 1. https://www.pharmaceutical-technology.com/features/rare-disease-spotlight-tracing-the-rise-of-orphan-drug-designations-over-almost-40-years/

Figure 2. https://www.precedenceresearch.com/orphan-drugs-market

Over the past few months, there have been signs that investors have grown uneasy about the private lending market. Bankruptcies are dominating headlines, reputable leaders in the industry are warning of overinvestment (Reuters, Oct 28) and the federal reserve is backing away from quantitative tightening (QT) to relieve liquidity stress in repo markets (FOMC, Dec 2). This article will cover:

• What are the events, and how big are they?

• A list of relevant indicators of market sentiment.

• What could it all mean?

When “Senior Secured” Stops Feeling Safe

The following are descriptions of recent bankruptcies and events that have altered market sentiment:

First Brands Group (FBG) - the star of the story

FBG is a privately held automotive supplier based in Ohio. The business was built and expanded via debt-fuelled acquisitions, FBG used off-balance-sheet receivables and Special Purpose Vehicles (SPVs) to disguise how much debt it was carrying. By late September 2025, the company filed for Chapter 11 bankruptcy (The Guardian, Oct 10). Billions of its invoices were allegedly fake or double-pledged; some sat in SPVs that allowed the firm to securitise receivables and appear capital-efficient. The result: total liabilities of around $11.6bn, with roughly $2.3bn of receivables now disputed or missing with banks and private lenders absorbing the hit (Reuters, Oct 16), Lenders have arranged about $1.1bn in debtor-in-possession (DIP) financing. These loans are given to the company to support operations during the bankruptcy and are ‘super-senior’ to other claims and according to well-regarded academic research, they are considered to have low repayment risk (Eckbo, Li, Wang, 2019). Regardless, FBG’s DIP debt was reportedly trading at 30 cents on the dollar, highlighting the severity of this event (Bloomberg, Dec 11). Two of the biggest victims were Jefferies’ Point Bonita fund (with approx. $715mn of receivables exposure) and UBS’s O’Connor hedge fund units (more than $500mn of exposure), causing UBS to liquidate several investment funds.

Tricolor Holdings

Tricolor was a Dallas-based subprime auto lender, which relied on bank warehouse lines and asset-backed securities (ABS) deals, much the same securitisation ‘plumbing’ that underpins private credit. The company targeted underbanked borrowers and at its peak managed $1.4bn in receivables and 25,000 creditors (wsws, Sep 16). As performance soured, banks uncovered tens of thousands of loans allegedly being double-pledged across multiple facilities alongside major inconsistencies in the loan data (abcnews, Dec 17). Tricolor abruptly filed for Chapter 7 liquidation (complete shutdown) on September 10, leaving JPMorgan and Fifth Third Bank to absorb $340–370m of combined losses, JPMorgan alone wrote off $170mn (Reuters, Dec 17). Meanwhile, ABS investors sitting on bonds that collapsed into distressed territory.

Blackrock - HPS Investment Partners, Renovo Home Partners, CLO failing tests

HPS Investment Partners, a private credit firm now affiliated with BlackRock, extended over $400mn to a group of telecom companies against receivables that US prosecutors say have been fabricated. The loans were made from 2020 onwards and were supposedly secured by trade receivables owned by blue-chip telecoms. Of the approximately $400mn HPS provided, roughly half was funded via leverage from BNP Paribas (Financial Times, Nov 17).

In another case, Renovo Home Partners, a Dallas-based home-improvement roll-up created by Audax Group in 2022, was financed with roughly $150m of senior private debt led by BlackRock. As late as September 2025, that loan was still marked close to par by its holders, only for Renovo to suddenly go into Chapter 7 liquidation in early November. BlackRock, which held the majority of the loan, had to write its position to zero, wiping out other lenders such as Apollo’s MidCap Financial and Oaktree Capital as well (Bloomberg, Nov 10).

Renovo sat inside BlackRock Baker CLO 2021-1, a roughly $500m private-credit CLO (Collateralized Loan Obligation) that also held other weakening loans such as Pluralsight Inc. and Astra Acquisition Corp. As these credits deteriorated, the CLO repeatedly failed its over-collateralisation tests, diverting cash away from junior tranches to protect the senior notes, even prompting BlackRock to waive part of its management fees to keep the structure alive in the late October payment cycle. This outcome effectively embeds double-digit percentage losses on the loan pool and shows how one Private Equity-backed blow-up can help tip an entire securitised private-credit vehicle into distress (Bloomberg, Nov 19).

Figure 1

US Private-Credit CLO Issuance Surges to New Highs (Volume and Deal Count, 2021–3Q25)

Source: PitchBook LCD, September 2025 US Private Credit Monitor (as of June 30, 2025)

The chart shows both volume and deal count are reaching new highs, which raises the question whether the recent events have slowed down deal count since these bankruptcies. If they haven’t, then their information advantage signals investors have little to worry about.

Blue Owl - an abandoned merger

Finally Blue Owl Capital, a major alternative asset manager with $250-300bn in AUM, provided a cautionary tale about liquidity. In late 2025, Blue Owl proposed merging its non-traded Business Development Company (BDC - issues loans as a direct lender to small and middle-market firms, often structuring senior-secured debt) Blue Owl Capital Corp II (around $1.7–1.8 billion in assets sold to retail investors), into its publicly traded BDC, Blue Owl Capital Corp (roughly $17bn of assets/private loans) on a 1-for-1 basis. The catch was that the listed BDC’s stock trades at a 10–20% discount to NAV, so a par-for-par swap would have imposed an immediate haircut on investors in the non-traded fund (Reuters, Nov 18). While the merger was pending, those investors were blocked from redeeming their shares until the deals closing. If they wanted out, their only path would have been through the publicly traded vehicle that the market already valued below book.

Figure 2

Price chart of Blue Owl’s publicly traded BDC - Blue Owl Corp ll

Source: LSEG Workspace (Refinitiv). Screenshot captured by the author on 23 Dec 2025.

The market reaction was swift, Blue Owl’s stock sold off, allocators and competitors pushed back hard. Consequently, Blue Owl announced its termination on November 19 (Blue Owl Capital Corp, Nov 19). All the while insisting the underlying loan portfolio was sound. Blue Owl’s Co-CEO, frames the abandoned merger as a validation/liquidation mismatch rather than a credit-quality issue. Alternatively, this merger was an attempt to turn a slow-motion run in a gated BDC into a permanent listing with a one-off haircut, preserving AUM and fee income while pushing liquidity and pricing risk onto public shareholders.

Saks Global - the most recent case

Saks Global is a privately held luxury retail group formed through the 2024 combination of Saks Fifth Avenue and Neiman Marcus. The business has historically relied on high leverage and asset-backed financing, leaving it exposed as discretionary spending softened and interest rates remained elevated. By mid-2025, the company had missed an interest payment and entered negotiations with creditors, signalling that liquidity had become the primary constraint. In anticipation of a Chapter 11 filing, lenders arranged approximately $1.25bn in DIP financing, and handed effective control of the restructuring to creditors, soon after the CEO stepped down. (Wall Street Journal, Jan 6), (Wall Street Journal, Jan 2)

Indicators marking sentiment

Although the cases above may seem like isolated frauds, they can be viewed as symptoms of a broader stress in the credit system. Lending was easy up through 2021 (into early 2022); after the Federal Reserve’s rate hikes began in 2022 and a higher-for-longer regime set in, the mood may of shifted from “How much can we lever up?” to “How do we hedge this capital structure in a 5% interest rate environment?”. At the same time, the AI-tech boom is pulling capital toward the next growth stories, and lenders are reluctant to be the ones who “miss out,” which encourages more leverage and optimistic forecasts to clear investment committees. Unfortunately, private markets operate in the shadows, so it is hard to directly observe these pressures. However, a few public proxies can shed light on possible stress:

CDSs and ETF flows

There is no CDS index for private loans, but you can see the stress in public high yield and credit indices.

• CDX HY: This an index tracking Credit Default Spreads (CDSs) for a basket of 100 High Yield bonds. When scandals like First Brands and Tricolor break, the first reaction is in US high yield spreads and the CDX HY. Even if levels are far from crisis territory, repeated spikes tell you the market is repricing leverage and refinancing risk after these significant events.

• iTraxx Crossover (Europe): The European crossover index gives a read on riskier, sub-IG corporates; it is a close liquid proxy for stressed sponsor-backed credits in the region.

• Loan and credit ETF flows: October saw roughly $1.5bn of outflows from US loan ETFs (Reuters, Oct 22).This is a hint that investors are taking chips off the table in leveraged credit, just as fraud and opacity stories hit the tape.

Figure 3

Price chart of the CDXHY.45.5Y Index, tracking basket of CDSs of High Yield U.S. mid-market companies.

Source: LSEG Workspace (Refinitiv). Screenshot captured by the author on 23 Dec 2025.

These quotes are prices, not spreads. It is a standardized CDS basket in the U.S. High Yield names which reflect the points upfront per 100 notional on a fixed 500 basis points (bps) running coupon, so prices below 100 represent wider credit spreads (>500bp) and prices above 100 (<500bps) reflect calmer conditions. Overall the chart is not intimidating, CDS spreads are tighter than previous years. Markets, on average, don’t require a larger premium for bearing high yield debt than they did before but the sharp drops around the tariff announcements and the three bankruptcies show how quickly premium can gap wider when narratives shift.

Figure 4

Bar chart showing spread range of unitranche loans held by BDCs as a percentage of total loans.

Source: PitchBook LCD, September 2025 US Private Credit Monitor (as of June 30, 2025)

This chart reinforces that spreads for loans given by BDCs are getting tighter. We can see that issues with spreads above 550 are declining while those below are increasing. Private lenders are accepting lower premia for the same risk, even though headlines and the equity discounts for these companies reflect a concern. Possible cause for this can be the slower repricing of private loans adjacent to the broader lending market. In theory, greater competition leads to more disciplined pricing as market players edge down to their marginal costs. However, these lower yields could also be an effect of behavioural forces. It is not irrational to argue that optimism bias and herding on the AI wave could be leading to a winner’s curse, causing lenders to systematically underestimate risk. If managers are optimising career risk rather than long-run Sharpe, then it is safer to be wrong with everyone else than to be early saying “no”. On the other hand, it may be hard to assure the future cash flows from this new technology but it is obvious to see it is here to stay, so why say “no”?.

Public BDCs – The Mark-to-Market Window on Private Loans

Listed BDCs are the closest thing to a daily mark on direct lending risk. Most of these vehicles hold the same kind of sponsor loans that sit in private funds, but their shares trade on exchange. Today, many BDCs sit at double-digit discounts to NAV, a market verdict that current book values are too optimistic.

Table 1

Table comparing financial ratios of publicly listed BDCs,banks and corporates.

Source: LSEG Workspace (Refinitiv). Screenshot captured by the author on 23 Dec 2025.

Table 1 shows five publicly listed funds (ARCC, BXSL, FSK, OBDC & OCSL), 2 banks (JPM & BARC), two AI companies with a Capex intensive business model (ORCL & EQIX). The ratios are specific to the funds themselves so it is inappropriate to judge the whole company since it does not include several revenue streams and costs that the aggregate business incurs. Regardless, we see that the BDCs have moderate leverage (1-1.25x D/E) yet they trade at or below book value while paying double-digit dividend yields. For all BDCs, dividend yields are higher than the 5 year average, however this is partly due to higher policy rates floating through loan coupons. At the same time, the price-to-book is falling. A consequence of a falling price is the higher yield, but a combination of the two signals that the market is demanding a higher return on providing funding to illiquid investments, at the same time as spreads on provided loans tighten. On one hand private credit has to issue cheaper loans due to competition and on the other hand they have to pay more to gain equity. This may create an incentive to lever, accept weaker covenants, amend-and-extend their loans, push PIK and concentrate in trending sectors to hit their target returns. PIK payments for the top 15 exchange traded BDCs have increased by 20% year over year in March of 2025.(Pitchbook, Aug 19), (S&P Global, Aug 1).

Repo Markets

Repo markets are systematic liquidity indicators and recently in the U.S. they have shown some signs of trouble. Moreover, the Fed has begun expanding their balance sheet with short-term T-bills (FOMC Statement, Dec 10), at a rate of approximately $40bn a month (New York Fed, Dec 10). Repo markets are the core short-term funding channel where banks and dealers borrow cash against high-quality collateral like Treasuries. In normal conditions, general collateral repo trades close to the policy rate; when repo rates spike above it or haircuts rise, it signals that access to short-term funding is getting tighter.

Figure 5

Fed Policy Rate vs SOFR: Funding Conditions Tracking the Fed Pivots (2021–2025)

Source: LSEG Workspace (Refinitiv). Screenshot captured on 24 Dec 2025.

The blue line indicates the Secured Overnight Funding Rate (SOFR) in the U.S. the white line indicates the Interest on Reserve Balances (IORB) and yellow line the Overnight Reverse Repo facility rate (ON RRP). The IORB and the ON RRP are referred to as the corridor system because, in the textbook world, they keep the overnight “risk-free” rate (SOFR) within this range. The ON RRP is the rate at which non-banks can lend cash to the Fed in exchange for treasuries. This is considered a floor because if the SOFR falls below, cash lenders can stop lending in repo and lend at ON RRP instead, which pulls SOFR back up. The IORB is the ceiling because if SOFR is above the IORB then banks with access to the IORB will prefer to keep less money sitting as reserves at the Fed and invest more in Repo markets, earning the SOFR. The opportunity cost between the two is meant to pull the SOFR back into the corridor but in practice there are a few frictions that cause it to linger above:

• Balance sheet constraints: to push SOFR down, banks/dealers would need to intermediate more repo operations (lend cash against Treasuries) that expand balance sheets and can be costly due to leverage constraints or internal risk limits.

• Segmented participants: not everyone can play the game, so the ones that can may not see the utility in a few bps due to operational constraints

• Intraday liquidity and payment needs: banks hold reserves for liquidity buffers, in uncertain times, they may not be willing to give that up.

This means that banks that have the opportunity to earn a few bps through arbitrage do not see the benefits as worth the balance-sheet and operational constraints. This could be due to many reasons but it has not occurred with this persistent 4 month stretch in the past 4 years.

Figure 6

New York Fed Repo Operations: Daily Amount Accepted (All Collateral Types), 2006–2025

Source: Federal Reserve Bank of New York, “Repo Operations” (Repo Chart – Amount accepted; Total Security Types) on 27 Dec 2025.

The chart plots the New York Fed’s daily repo operations “amount accepted” – i.e., how much liquidity the Fed had to inject on a given day when the market demanded cash against collateral. When the bars are close to zero, it means the market is funding itself without leaning on the Fed’s repo backstop. Repo operations have remained near zero since the pandemic, and the recent tick-up coincides with the bankruptcy headlines and SOFR trading above IORB (Figure 5). Post-pandemic, liquidity in banks was vast due to the Fed’s quantitative easing (QE) increasing bank reserves, which likely helped keep repo take-up close to zero. More recently, the Fed shifted its stance, gradually tightening reserves which can make funding conditions more sensitive. Nevertheless, the plumbing still looks broadly functional: the system is not heavily leaning on the Fed. However, even a small reappearance of repo take-up is a reminder that backstops can come back into play quickly when funding conditions tighten. If those bars start to cluster and persist again, it would be the market-based confirmation of “cracks” spilling into the broader system.

What does it all mean?

The recent blow-ups can be read in one of two ways: either as early indicators of a broader fragility caused by unsustainable lending practices or as idiosyncratic failures consistent with normal ‘creative destruction’ in a maturing market, although these are just my interpretations. It is hard to judge the size of the role private markets play in this because they do not price on screen and they live in the shadows. We can only infer from public proxies, therefore, it is important to note this piece is BDC-heavy so broader conclusions are insignificant.

A useful counterpoint is how leading managers explain this divergence. In Barron’s, Blue Owl co-CEO Marc Lipschultz argues the current sell-off in listed private-credit managers is caused by investors extrapolating from a small set of headline cases (First Brands, Tricolor) into a blanket risk premium for anything labeled “private credit.” Lipschultz frames the abandoned BDC merger as a valuation/liquidity problem rather than deteriorating credit quality; but that is precisely the mechanism that exacerbates the stress but not the one that causes it. Either (i) private lenders’ tighter covenants, information rights, and sponsor scrutiny genuinely keep losses contained (today’s discounts are largely a fear premium) or (ii) the recent fraud-adjacent blow-ups are early signals that opaque collateral and securitisation “plumbing” are being stress-tested. That is why the SOFR corridor and Fed repo usage matters: it tells us whether this stays a contained “private” story or migrates into system funding.

Overall, signs of stress are beginning to emerge, even as private credit continues to grow and reach new highs. Investment in AI-related infrastructure remains robust, yet public markets increasingly appear to penalize both borrowers for incremental leverage and credit managers for perceived overextension or skepticism around the feasibility of realizing the exits embedded in buyout projections. When liquidity is abundant, risk hides easily. The real question now is whether private credit’s confidence reflects genuinely durable fundamentals, or an overextension of leverage masked by reassuring narratives. In other words, is the real AI bubble in private markets?

References

Academic Papers

Eckbo, B. E., Li, K., & Wang, W. (2019). Rent extraction by Super-Priority lenders.

https://doi.org/10.2139/ssrn.3384389

ABC News
ABC News. (2025, December 17). Fraud charges follow subprime auto lender Tricolor Holdings collapse.

Barron’s
Barron’s (2025, November 26). Blue Owl is having a rough year. Its Co-CEO says investors’ fears are ungrounded.

Bloomberg
Bloomberg News. (2025, November 10). BlackRock eyes 100% loss on private loan amid debate over marks.
Bloomberg News. (2025, November 19). BlackRock private-credit CLO fails key tests as bad loans mount.
Bloomberg News. (2025, December 11). First Brands loan hits 30 cents as fresh rescue seen as critical.

Blue Owl Capital Corp
Blue Owl Capital Corp. (2025, November 20). Press release: Termination of proposed BDC merger.

Federal Reserve / Federal Reserve Bank of New York
Federal Open Market Committee (FOMC). (2025, December 10). FOMC statement on monetary policy.
Federal Reserve Bank of New York. (2025, December 10). Statement regarding reserve-management Treasury bill purchases.
Federal Reserve Bank of New York. (2025). Repo operations: Amount accepted (all collateral types).

Financial Times

Financial Times. (2025, November 17). US probes firms that borrowed $400mn from private credit giant HPS.

Journal of Law and Economics

Eckbo, B. E., Li, K., & Wang, W. (2019, December 16). Rent extraction by super-priority lenders (Working paper, Tuck School of Business at Dartmouth)

LSEG Workspace (Refinitiv)
LSEG Workspace. (2025). CDX HY index prices; Blue Owl Capital Corp II price chart; SOFR, IORB, ON RRP data. Screenshots captured r on December 23–24, 2025.

PitchBook
PitchBook. (2025). U.S. Private Credit Monitor. September 2025 edition (data as of June 30, 2025).

PitchBook. (2025, August 19). PIK interest income surpasses $1B for top 15 BDCs in Q1 2025.

Reuters

Reuters. (2025, October 22). First Brands bankruptcy sparks sharp outflow from U.S. loan funds.
Reuters. (2025, October 16). Mapping the scale of beleaguered First Brands’ debts and its creditors.

Reuters. (2025, November 18). Blue Owl fund redemption fees sparks fresh private credit jitters.

S&P Global Market Intelligence
S&P Global Market Intelligence. (2025, August 1). Rising PIK usage reflects borrower stress in private credit.

The Guardian
The Guardian. (2025, October 10). First Brands files for Chapter 11 after accounting irregularities.

Artificial intelligence is transforming the global economy at an astonishing pace, but not every application of AI will justify the enthusiasm surrounding it. Investors looking for sustainable growth opportunities often find themselves torn between chasing the momentum of pure technology companies or identifying where AI creates genuine, measurable value. Healthcare might represent that middle ground as it is a sector where AI doesn’t just drive valuation hype, but improves fundamental outcomes, saving time, reducing costs, and, most importantly, improving patient lives.

While the returns from healthcare AI may not be as immediate or explosive as those seen in the broader tech sector, the potential for long-term efficiency gains and resilient business models makes it one of the most promising frontiers in both technology and healthcare investment.

The Broader AI Landscape: Promise and Risk

The technology industry’s AI boom has become heavily concentrated among a small network of firms. Companies such as Nvidia, OpenAI, and Oracle are now entangled in a cycle of mutual dependence where investment flows circulate rapidly between them, inflating paper valuations in ways not always supported by tangible productivity gains. When comparing market capitalisations, NVIDIA currently exceeds the combined value of more than twenty of the world’s largest pharmaceutical companies. This contrast highlights how significantly the AI boom has already been priced into technology stocks. In contrast, the pharmaceutical and broader healthcare sectors are only beginning to harness AI’s potential, through process optimisation, accelerated drug discovery, and data-driven insights. From an investor’s perspective, this suggests that while AI in technology may have already reached elevated valuations, the application of AI in healthcare represents a relatively untapped, stable, and potentially high-growth opportunity that the market has yet to fully recognise.

For investors seeking a steadier route into artificial intelligence, the real opportunity may not lie in AI as a product, but in AI as a tool, particularly in sectors where data quality, regulation, and practical impact define long-term value as it is the case in healthcare.

AI in Healthcare: Where Science Meets Stability

Healthcare offers a strong case study in how AI can move beyond hype into practical, value-driven innovation. Medical data, whether genomic sequences, imaging scans, or patient records, is already well-structured, making it ideal for advanced analytics. AI’s strength in pattern recognition is particularly well suited to medicine, where subtle correlations can guide faster diagnosis or drug discovery.

Perhaps the most visible example of AI’s impact lies in molecular biology. Deep learning models such as DeepMind’s AlphaFold have reshaped structural biology by accurately predicting three-dimensional protein structures within minutes to hours. Just a decade ago, this work might have required months of laboratory analysis. The ability to simulate protein interactions at scale enables scientists to identify disease-linked mutations, optimise drug targets, and tailor new enzyme-based therapeutics with a precision previously unimaginable.

https://www.ebi.ac.uk/training/online/courses/protein-interactions-and-their-importance/where-do-the-data-come-from/x-ray-crystallography/

This shift to in silico research forms the backbone of a much broader trend, with convergence of computational power and biological insight. It doesn’t only accelerate discovery but also dramatically lowers costs by filtering out poor candidates before they reach clinical testing. Such efficiencies hint at a healthcare model where data, rather than dollars, drive progress. For instance, when applying AI to the selection of clinical trial participants, the technology can sometimes lead to an overrepresentation of historically underrepresented groups. While counterintuitive at first, this can be a significant advantage enhancing the discovery and development of drugs that are more effective across diverse populations with underrepresented conditions for example. Moreover, when AI systems are trained to specialise in a single, well-defined area within healthcare, they can deliver substantially greater value to a company. By narrowing their focus, these models learn to identify and refine the most accurate outcomes with much greater precision, effectively reducing variability and improving both reliability and efficiency. However it is always crucial that tools are in place to audit the AI and ensure that the final decisions are human and only supported by AI in a way that helps surface details that may otherwise remain unseen.

The Investment Case for Healthcare AI

Several companies illustrate how AI is already transforming their operations. Illumina uses machine learning to refine genetic analysis, improving the speed and accuracy of next-generation sequencing. GE Healthcare applies AI to imaging diagnostics, helping physicians identify early signs of disease that might otherwise be missed. Thermo Fisher Scientific integrates predictive algorithms into laboratory workflows, boosting efficiency and reproducibility. Working closely together with OpenAI is another point that makes Thermo Fisher interesting, as it gives indirect exposure to an AI giant while having the healthcare sector almost as an insurance for stability.

For investors, this presents multiple routes into AI exposure without assuming the speculative risk associated with pure AI firms. Large pharmaceutical companies are also embedding AI into R&D pipelines to de-risk drug development. According to the Novartis annual reports in 2023 they spent around 11 billion on R&D, whereas in 2024 they only spent around 10 billion. This reduction in spending might be because of financial cuts or general savings with the approaching patent cliff, could however also show how utilising AI can reduce costs by re-arranging resources to assets that are the most promising and dropping those less likely to become blockbuster drugs. Predictive models can now anticipate clinical trial outcomes, identify non-viable compounds earlier, and streamline regulatory documentation. These advantages not only improve efficiency but also financial resilience, two qualities that tend to correlate with sustainable stock performance.

More information on Novartis’ patent cliff: https://substack.com/home/post/p-177801468

Healthcare AI-focused exchange-traded funds (ETFs) and mutual funds are emerging to capture this cross-sector opportunity. They combine the innovation potential of technology with the defensive characteristics of healthcare, a blend well suited to investors with long-term horizons.

Challenges and Structural Barriers

Despite its promise, the path to widespread AI adoption in healthcare is not without obstacles. Regulatory approval processes have yet to fully adapt to the unique characteristics of adaptive algorithms. The U.S. Food and Drug Administration (FDA) and the European Medicines Agency (EMA) are both still defining how to assess continuously learning systems, a complexity that doesn’t exist in traditional medical devices or therapies.

Data privacy remains another major challenge. The sensitivity of patient records requires secure storage, ethical oversight, and transparent consent frameworks. Bias also poses an ongoing risk, if training datasets lack demographic diversity, AI systems can inadvertently reinforce existing health inequalities. Integration into existing hospital systems can be slow, given the prevalence of legacy infrastructure and the cost of retraining staff.

However, these growing pains are typical of any technological transformation within a regulated industry. With time, policy, and thoughtful implementation, they can be managed. The rewards, better diagnostics, faster research, and lower costs, make that adaptation worthwhile for patients, companies and investors.

Why Healthcare AI Holds Long-Term Value

Healthcare’s fundamentally stable demand differentiates it from other sectors prone to hype cycles. AI amplifies that stability by improving performance rather than reinventing the business model. For example, in oncology research, machine learning can rapidly analyse biomarker data to identify which compounds have the highest likelihood of success in clinical trials. This not only boosts the probability of scientific breakthroughs but reduces wasted investment along the way.

With ageing populations and a growing global burden of chronic disease, demand for healthcare innovation will remain high. AI is therefore not a passing trend in this industry, it is a structural enabler of long-term efficiency and resilience. As computational biology matures, the companies leading this integration could define the next generation of medical innovation.

Conclusion

The distinction between hype and progress often depends on where AI is applied. While technology firms continue to define the infrastructure and grab the headlines around artificial intelligence, sectors like healthcare quietly demonstrate how to convert computational advances into tangible value.

AI’s real potential emerges when it supports rather than supplants expertise, when it becomes a tool for scientists, clinicians, and researchers to do their work better. In that sense, healthcare AI may outlast the speculative excitement that currently dominates the market. It offers something far more enduring, a roadmap for how artificial intelligence can deliver measurable value to patients, professionals, and investors alike.

Using monthly data of Italy, France, and Germany’s yield curve, this article aims to predict the yield curve using the research of Diebold and Li (2006). Spreads between these countries are predicted using this method, a measure of uncertainty around these predictions is considered. The findings indicate that point predictions are not fully informative but the asymmetric risks shown by bootstrapped confidence intervals are insightful.

The Yield Data

The data consists of monthly bond yields for Italy, France, and Germany starting from 01-01-2005 up to 30-09-2025. The data shows the yields of bonds with maturities 2,3,4,5,6,7,8,9,10 and 30 years and the data is sourced from Bloomberg. Figure 1 contains a visual representation of the various bond yields of Germany over time. At the start of this timeline the bond yields were higher, around 4%. Halfway through the considered timeline bond yields were low. More recently the bond yields have been going up again, partly due to the halting of the Asset Purchasing Programme of the ECB, discussed in this earlier article.

Figure 1, source Bloomberg

The Nelson Siegel Model

Diebold and Li (2006) have used a modified Nelson Siegel model to predict bond yields and found it to be a superior model for predicting one year ahead. Modelling the yield curve is a complex task, the Nelson Siegel model tries to approximate the model of the yield curve by using a three-factor exponential model.

The parameter λ_t controls the exponential decay rate. Small values of this parameter imply a slow decay rate, while large values imply a fast decay rate. A small value of λ_t is better at modelling long term bond and a large value of λ_t fits better with a short term bond. Diebold and Li (2006) uses a value λ_t = 0.0609, their models view maturities in months. In my model we view maturities in years, so I use 12∗λ_t. The beta vectors β^(c)_t are estimated by (cross-sectional) ordinary least squares. This yields a time series of estimated vectors for each country:

A more detailed derivation for interested readers can be found in the Appendix A1. The yield curve is fitted at each data point available. For illustration, in Figure 2, four different dates in Germany’s yield curve history are presented with the fitted Nelson Siegel model.

Figure 2

Predictions

Now we use our fitted Nelson Siegel models to predict future bond yields. In particular, we predict the coefficients {βˆ_{1t}, βˆ_{2t}, βˆ_{3t}} using an AR(p) model. Diebold and Li (2006) use an AR(1) model to predict future bond yields. They found that predicting 12 months ahead gave the best results. For my prediction I use the Akaike Information Criterion (AIC) to select the amount of lags used in each model. The amount of lags p will differ per country and the AR(p) model is defined as:

The h-step-ahead factor forecasts βˆ_{i,t+h|t} are then inserted into the Nelson Siegel formula to obtain the forecasted yield curve:

We apply the above to obtain our prediction, Figure 3 presents the resulting prediction for Germany’s yield curve.

Figure 3

Investigating Spreads

The main interest for us is to inspect the spreads of countries in the EU. By forecasting the yield curve for Germany, Italy, and France, spreads can be constructed by subtracting these curves. In Figure 4 we see the forecasted spread of Italy-Germany 10YR. In Figure 5 the France-Germany 10YR maturity spread is forecasted. The model predicts a small decrease in the spread. Since the prediction is based on a AR(p) model it will not give big shocks as predictions.

In Figure 6 the German 10YR-2YR curve is forecasted. Once again the forecast is quite flat with a small predicted decrease.

Figure 4

Figure 5

Figure 6

Prediction Uncertainty

We have some predictions, but are they any good?

When forecasting it is always a good practice to include uncertainty in the predictions. Therefore, I will include bootstrapped confidence intervals that do not make distributional assumptions. This will cause the confidence intervals to be wider compared to standard normally distributed intervals, but yield more realistic confidence intervals. The method used specifically is called a ‘Sieve’ bootstrap. A more detailed explanation is in the appendix A2.

The Italy-Germany and France-Germany 10YR spread including bootstrapped 95% confidence intervals are given in Figure 7 and Figure 8. For both spread predictions the point forecast only predicts limited variation in the future, while the confidence interval is relatively wide. This implies there is significant uncertainty in the prediction. The confidence intervals are also asymmetric. For the Italy-Germany case there is more possibility to the upside of the prediction. For the France-Germany case it is quite the opposite, the risks are more skewed to the downside.

Overall, the bootstrap results suggest that while no strong deterministic changes in spreads are predicted, the distribution of future outcomes is skewed, highlighting directional risks that are not apparent from point forecasts.

Figure 7

Figure 8

Findings

In this article, a Nelson Siegel model is used to forecast government bond yield curves and yield spreads for Germany, France, and Italy. The model fits the historical yield curves reasonably well, but the point forecasts themselves are mostly flat over the next year and come with a lot of uncertainty. The more interesting results come from the confidence bands. These bands are wide and clearly asymmetric, meaning that the risks are not evenly balanced. For some spreads, there is more risk on the upside, while for others the downside risk dominates. This shows that even when the expected change in spreads is small, the direction of risk can still matter. Overall, this type of model is better suited for understanding uncertainty and risk around yield spreads than for making precise point predictions.

Appendix

A1 Nelson Siegel Model

We consider zero-coupon yields observed at discrete times t = 1, …, T, for multiple countries and maturities.

Let

denote the yield at time t for country c and maturity τ > 0 (in years), e.g. τ_j ∈ {2, 3, 4, 5, 6, 7, 8, 9, 10, 30} years. For each country c and time t, we assume a Nelson Siegel representation of the yield curve.

The parameter λ_t controls the exponential decay rate. Small values of this parameter imply a slow decay rate, while large values imply a fast decay rate. A small value of λ_t is better at modelling longer term bonds and a large value of λ_t fits better with a shorter term bonds. Diebold and Li (2006) uses a value λ_t = 0.0609, their model views maturities in months. In my model we view maturities in years, so I use 12 ∗ λ_t. Now all we have left are the betas. Using ordinary least squares the coefficients { βˆ_{1t}, βˆ_{2t}, βˆ_{3t} } are estimated for each point in time. At each observation we fit the Nelson Siegel model to fit the yield curve at that moment. For illustration, I show the fitted yield curve model at four specific data points in Figure 2. As you can see the fitted Nelson Siegel model fits reasonably well. It is also noteworthy that I do not posses the data of more short term maturity bonds. This creates a different yield curve image compared to yield curves that include maturities below two years. The model can be written as a vector model.

Define the factor loading vector as

and the time-t Nelson Siegel beta vector

Then for a given t and c, the model can be written compactly as

where

as mentioned before, given the observed yields y^{(c)}_t and fixed λ, the Nelson Siegel betas β^{(c)}_t are estimated by (cross-sectional) ordinary least squares. This yields a time series of estimated Nelson Siegel factor vectors for each country:

A2 Confidence intervals

We will perform a ’Sieve’ bootstrap that takes parameter uncertainty of the AR(p) model into account, as well as the prediction uncertainty. The main idea is to redraw the errors of the fitted AR(p) process with replacement. Using the errors new AR(p) models are estimated and with the bootstrapped models we obtain new predictions. The 95% confidence interval of the bootstrapped predictions will give us a quantified measure of the parameter and prediction uncertainty. Now we will use some mathematical definitions for the bootstrap method.

The stacked beta vector b_t ∈ R^6 (Two countries stacked) is modelled as

with residuals eˆ_t from the estimated model.

Let

and let

be the empirical residual pool.

For each bootstrap replication b = 1, . . . , B we:

1. Generate a pseudo sample

   recursively by

   where the residuals e^{∗(b)}_t are drawn with replacement from the residual pool, and the first p values b^{∗(b)}_1 , …, b^{∗(b)}_p are set equal to the observed b_1, …, b_p.

2. Re-estimate the VAR(p) on {b^{∗(b)}_t} to obtain bootstrap parameters μˆ^{∗(b)} and Aˆ^{∗(b)i} .

3. Starting from b^{∗(b)}_T, simulate a bootstrap forecast path b^{∗(b)}_{T+1}, …, b^{∗(b)}_{T+h} using

where each η^{∗(b)}_{T+k} is again drawn with replacement from the centred residuals {˜e_t}. For each b and k, the corresponding Nelson Siegel beta vectors are mapped back to yields via the Nelson Siegel formula and then to term spreads and cross-country spreads. The empirical quantiles of the bootstrap forecasts

provide bootstrap prediction intervals, for example the 2.5% and 97.5% quantiles yield a 95% prediction band.

References

Diebold, Francis X. and Canlin Li (2006). Forecasting the term structure of government bond yields. Journal of Econometrics 130 (2), 337–364.

Have you ever found yourself opening Duolingo at 11:59 pm, not because you’re passionate about learning French, but because you simply cannot bear to lose your 247-day streak? Millions of users worldwide engage with the app daily, yet if asked whether they’re actually progressing towards fluency, most would struggle to answer. The green owl has become less a teacher and more a digital taskmaster, gamifying language learning to the point where users often forget whether they’re studying Spanish or simply playing an elaborate mobile game.

Company Introduction

Duolingo sits at an intriguing intersection between educational technology and mobile gaming. It offers language courses in over 40 languages through bite-sized, gamified lessons. Users complete exercises in reading, writing, listening and speaking whilst earning points and maintaining daily streaks. The company operates on a freemium model: basic content is free and advertisement-supported, whilst Super Duolingo removes adverts and adds premium features. With colourful characters, levels, leaderboards and cartoon mascot Duo, the product resembles a mobile game rather than traditional educational software.

Duolingo’s mission is to “make the best education in the world available to everyone”. In practice, the app has dramatically lowered the barrier to language learning. Lessons take just five to ten minutes on a smartphone, making language study feel achievable even for those who cannot commit to intensive courses.

Founding Story & Business Model

Duolingo was co-founded in 2011 by Luis von Ahn and Severin Hacker, with von Ahn continuing as CEO. Von Ahn previously invented CAPTCHA and reCAPTCHA at Carnegie Mellon University. His reCAPTCHA innovation was particularly ingenious: it transformed online security verification into a crowdsourced digitisation project, with users unknowingly transcribing texts that computers couldn’t recognise. Google acquired reCAPTCHA in 2009.

Von Ahn’s motivation for Duolingo stems from personal experience. Growing up in Guatemala, he witnessed how mastering English could transform lives, yet prohibitive costs placed traditional language education beyond reach for most people in developing countries. So he came up with an idea of creating a free language learning app. To test the app, he and Hacker began teaching each other their native languages, Spanish and German. They identified the critical challenge: maintaining motivation in self-directed learning. This insight became Duolingo’s founding principle, shaping its game-like architecture from the start.

In the early days, the team was small and scrappy—essentially von Ahn and a few of his graduate students at Carnegie Mellon. Within six months, they secured $3 million in venture capital funding. Von Ahn’s pitch was audaciously simple: “I just sold a company to Google. I made a lot of money. I have this thing (Duolingo) for free. I don’t know how it’s going to make money. But trust me.” His track record carried the day.

Duolingo began monetising in 2017 through a freemium model where basic content remains free with advertisements, whilst Super Duolingo removes adverts and adds premium features. This model has proven effective: approximately 7-8% of monthly active users subscribe to premium, yet subscriptions generate roughly 80% of revenue. Advertisements contribute 10%, and the remaining 10% comes from the Duolingo English Test.

The English Test represents a strategic move towards educational credibility. Priced at only $49, a fraction of TOEFL or IELTS costs, the one-hour online exam uses AI-generated questions with adaptive difficulty. The algorithm adjusts in real time, quickly identifying each candidate’s proficiency level. Camera monitoring and human video review address concerns about cheating.

To persuade universities to accept their test, the company first conducted a scientific study proving high correlation between Duolingo English Test scores and TOEFL results. Von Ahn then leveraged his academic network to secure institutional acceptance, with Yale providing the crucial breakthrough. Yale initially treated the test as supplemental verification for strong candidates, but its endorsement signalled legitimacy to other institutions. By early 2020, approximately 800 programmes worldwide recognised the test.

Then COVID-19 closed testing centres globally, creating urgent demand for alternatives. Over three months in 2020, institutional acceptance exploded from 800 to over 3,000. Today, more than 5,000 programmes accept the Duolingo English Test. Beyond creating a meaningful revenue stream, this achievement signals that Duolingo can deliver serious educational products when necessary, not merely gamified entertainment.

Subscribe now

Expansion and Innovation Strategy

Management understands that gamification is Duolingo’s core competitive advantage. Rather than saturating the language market, the company has methodically expanded this model into adjacent learning categories. Mathematics (2023), music (2023), and chess (2025) now sit alongside its core offering. Von Ahn applies three criteria to evaluate new subjects: massive global demand, societal benefit, and suitability for mobile delivery. This disciplined approach has filtered out niche topics like programming and rejected frivolous subjects despite user interest.

Mathematics easily fits these criteria with enormous demand and clear educational value. Chess, as a game itself, fits Duolingo’s nature. Music remains more experimental, and whether gamified music education translates to genuine proficiency on a smartphone remains to be seen.

This expansion reflects a strategic ambition: transforming Duolingo from a language platform into a broad learning ecosystem whilst preserving its playful, mobile-first identity. By reusing its gamification toolkit of streaks, points, levels and mascots across new domains, Duolingo can attract new user segments and retain existing users with fresh content.

The 2025 LinkedIn partnership marks a significant shift. Users can now display Duolingo Skill Scores on professional profiles, positioning casual learning outcomes as quasi-professional credentials. Combined with the English Test serving as formal academic certification, these initiatives signal Duolingo’s ambition to gain legitimacy with employers and institutions.

Duolingo has also declared itself “AI-first”, integrating GPT-4 and similar technologies to personalise lessons and power conversational chatbots. Von Ahn notes that Duolingo is amongst the rare companies profiting from AI features, with the premium Duolingo Max tier charging users for AI-driven experiences including one-on-one tutoring sessions.

The strategic trajectory is clear: sustainable growth through disciplined product expansion, addictive user experience, and technology adoption that enhances both learning and monetisation. Yet a fundamental tension persists for investors: can Duolingo reconcile its gaming mechanics with educational credibility, or will this identity crisis ultimately constrain its market reach?

The Financials

Figure 1

An examination of Duolingo’s key performance indicators reveals the efficacy of its freemium business model at scale.

Monthly active users (MAUs) have expanded materially over the period, reflecting continued success in global user acquisition. More importantly, daily active users (DAUs) have grown faster than MAUs, indicating that new users are not merely sampling the product but integrating it into daily routines.

The user engagement rate, measured by the DAU/MAU ratio, has consistently climbed over the years. For investors, this metric is a critical barometer of product “stickiness,” as habitual engagement is the foundational asset upon which all monetisation is built.

Despite the explosion in user volume, paid subscribers have grown consistently, with the conversion rate remaining disciplined at 7–11% of MAUs, peaking at 11.5 million paid subscribers in Q3 2025.

Figure 2

Monetisation of this engaged user base continues upward. In Q3 2025, subscription bookings totalled $240.3 million, accounting for 85.2% of the total bookings of $281.9 million. Subscription bookings reflect cash received from all Duolingo subscription purchases, whilst total bookings also include advertising income, English Test revenues, and in-app purchases.

Although the subscription share fluctuates modestly from quarter to quarter, the long-term trend is clearly upward. This confirms that Duolingo is increasingly monetising its platform through paid users rather than peripheral revenue streams. Subscriptions have become a core and growing source of cash flow, demonstrating both users’ willingness to pay for Duolingo’s educational value and the company’s ability to convert engagement into sustainable revenue.

Figure 3

The profitability metrics highlight a decisive turning point in Duolingo’s financial profile, driven primarily by the increasing contribution of subscription revenue. Return on Equity (ROE), which remained between -32% and -10% throughout 2021 and 2022 as the company prioritised growth investment, began a steady recovery in early 2023. ROE turned positive by year-end 2023 and accelerated to 36.3% in Q3 2025. This improvement reflects operating leverage from a growing subscriber base, higher-margin recurring revenue, and disciplined cost scaling.

Return on Invested Capital (ROIC) has followed an even more dramatic trajectory. After reaching its weakest point at -82.7% in Q2 2022 during a period of heavy investment, ROIC recovered through 2023 and turned positive in Q1 2024, rising to 52.3% by Q3 2025. This trajectory highlights the scalability of Duolingo’s asset-light model: incremental subscription and AI-driven revenues require minimal additional capital, allowing profits to compound rapidly once fixed costs are absorbed.

Duolingo operates with more cash than debt, therefore Duolingo minimizes current solvency risks.

Figure 4

The company emphasises its performance using Adjusted EBITDA, a non-GAAP metric that excludes items such as stock-based compensation. This adjustment is common among technology companies and can be useful in assessing near-term operating cash flow, as stock-based compensation is not an immediate cash expense. However, for a business reliant on highly skilled talent, stock-based compensation is a core component of the cost structure.

The adjustment largely impacts reported financials. In Q3 2025, stock-based compensation amounted to approximately half of Adjusted EBITDA, suggesting that headline profitability substantially exceeds economic earnings. Whilst excluding stock-based compensation may improve period-to-period comparability, it does not eliminate the underlying cost. Rather, it defers the expense from the income statement to shareholders through equity dilution.

Duolingo’s compensation strategy illustrates this dynamic. The company privileges equity awards over cash salaries, aligning employee incentives with long-term value creation whilst preserving operating cash. When employees exercise options, the company issues new shares to fulfil awards. This equity issuance reduces existing shareholders’ ownership proportionally. Over the years, a shareholder’s stake may decline as the company issues additional shares for compensation purposes.

For investors, this does not imply aggressive accounting or misrepresentation, but it does require careful interpretation. A robust assessment of Duolingo’s profitability should consider Adjusted EBITDA alongside GAAP net income and share count dilution. Only by reconciling these measures can investors fully assess the true cost of growth and the sustainability of the company’s improving returns.

Subscribe now

Competitor Analysis

Duolingo occupies a relatively unique market position with few direct rivals at scale. The company has successfully created a category: gamified language learning for casual learners. This positioning has proven difficult to replicate. Inspired by Duolingo’s success, other language learning apps have introduced gamification features such as daily streaks and progress tracking, yet few have achieved similar user stickiness.

Busuu, Duolingo’s closest competitor, offers stronger pedagogical depth, including structured grammar explanations, CEFR-aligned courses, native-speaker community feedback, and free AI-based conversation practice. Yet despite these strengths, Busuu continues to lag Duolingo in engagement and retention. Duolingo’s advantage lies less in instructional quality and more in its intellectual property ecosystem. Through its mascot, streak mechanics, and social reinforcement loops, the company has built a platform difficult for users to leave. Switching costs are driven by habit rather than content superiority.

Other education technology platforms such as Memrise, Beelinguapp, and Rosetta Stone target more serious or professional learners and operate in a different segment. Structured course marketplaces like Coursera and Udemy prioritise depth and certification over daily engagement. These competitors represent fundamentally different learning philosophies and pose limited direct threat.

Artificial intelligence adds a further layer of disruption risk. Large language models can deliver personalised explanations, conversational practice, and adaptive learning at low cost. Yet AI alone struggles to sustain long-term motivation. Without habit-forming design, even the most advanced AI tools risk becoming episodic utilities rather than daily learning companions. For now, Duolingo’s strength lies not in teaching languages better than anyone else, but in making users come back tomorrow.

Subscribe now

Investment Implications

Figure 5

Despite delivering strong operational and financial results in Q3 2025, Duolingo’s share price declined sharply following the earnings release. The stock fell towards the $170 level, only modestly above its 2021 IPO range.

The market reaction appears driven less by current performance and more by forward-looking concerns. The sell-off is likely linked to the rising AI fear in general. Duolingo’s AI investment, estimated at roughly 12% of market capitalisation, has heightened near-term risk perception among investors, even as management frames this spending as long-term value creation.

At a deeper level, Duolingo faces a structural tension that financial metrics alone cannot resolve. Many users credit Duolingo for sparking their interest in a language, then attempt professional courses, find them too demanding, and return to Duolingo. This pattern reveals the company has carved out a sustainable niche as the “participation trophy” of language learning. Whilst this isn’t necessarily negative from an investment perspective, it constrains pricing power and institutional adoption.

Duolingo has made language learning fun, but gamification has its limits. Language acquisition isn’t about memorising several words daily. If you refer to the strategic learning style of Polyglots, they typically achieve fluency through intensive, systematic immersion over short, concentrated periods. Two months in an immersive environment typically yields better results than two years of daily streaks. This gap between engagement and genuine proficiency defines Duolingo’s market positioning and explains why users often plateau at casual competency.

Duolingo has built something genuinely difficult to replicate: a habit-forming learning platform with massive user engagement. The company demonstrates proven engagement mechanics in an era where user attention is the scarcest resource, benefits from low customer acquisition costs driven by viral growth and word-of-mouth, and, as a young company following its 2021 IPO, retains significant runway for expansion and business model optimisation.

In investing, it is often said that one should not catch a falling knife. This is definitely a falling knife at the moment. Yet the company’s fundamentals tell a different story. Whether this is a knife still falling, or an opportunity mispriced by short-term fear, is a question the market has yet to answer.

This analysis is for informational purposes only and should not be considered investment advice. Readers should conduct their own due diligence and consult with financial advisers before making investment decisions.

Introduction to Novartis

Novartis is one of the world’s largest pharmaceutical companies, driven by a mission to discover ways to improve and extend people’s lives. This mission is supported by a broad portfolio spanning multiple therapeutic areas, including oncology, immunology, and dermatology. The company develops drugs internally and strengthens its pipeline through acquisitions of intellectual property and patents from smaller biotechnology firms.

In recent years, Novartis has shifted its focus toward high-margin prescription medicines, reducing emphasis on generics and consumer health products. This strategic move emphasizes physician-prescribed therapies over over-the-counter products, positioning Novartis as a leader in innovative treatments. Understanding both the scientific and financial dimensions of the company is essential for evaluating its long-term stability and growth potential.

Historic influences on Novartis

Novartis’ legacy is deeply rooted in over a century of pharmaceutical innovation, originating from the merger of Ciba-Geigy and Sandoz in 1996. Historically, the company built its success on pioneering small-molecule and biologic therapies such as Gleevec for leukemia, Diovan for hypertension, and Lucentis for retinal disorders. These drugs not only defined Novartis’ scientific reputation but also generated substantial long-term revenue streams. However, as patents for these blockbusters expired throughout the 2010s, the company entered a period commonly referred to as a “patent cliff”, a stagnation or decline in revenue following the loss of market exclusivity. Once generic and biosimilar competitors entered the market, profit margins contracted rapidly, forcing Novartis to intensify its focus on new biologics and specialty drugs. This shift marks a transition from a historically chemistry-driven enterprise to a biotechnology-centered innovator. The company’s recent acquisitions, including Tourmaline Bio, can therefore be seen as part of a broader strategy to replenish its patent portfolio and mitigate recurring revenue erosion, while continuing to position itself at the forefront of high-value therapeutic innovation.

R&D Strategy and Investment

According to the 2024 financial report, Novartis invested approximately 10 billion USD in research and development. This substantial investment supports a pipeline of over 100 drug candidates across a wide range of therapeutic areas and development stages.

Figure 1

As shown in the company’s pipeline data, most candidates are already in late-stage clinical trials, where safety has been established and efficacy testing becomes the primary focus before market approval. The portfolio remains highly diversified, with oncology representing roughly one third of the total pipeline.

Key Drugs for future revenue

Cosentyx

Cosentyx functions by inhibiting interleukin-17A (IL-17A), a cytokine involved in driving inflammatory responses at mucosal surfaces. It is primarily prescribed for psoriasis, psoriatic arthritis, and other chronic inflammatory skin or joint diseases.

Figure 2

The global Cosentyx market represents a consistent and reliable revenue source for Novartis [1]. Potential off-label applications such as rheumatoid arthritis or systemic lupus erythematosus could further expand its market potential prior to patent expiry. However, administration requires medical supervision due to possible side effects or contraindications with other biologics. After the initial supervised dose, trained patients may self-administer the drug, enhancing convenience and adherence.

Kesimpta

Kesimpta directly competes with Roche’s Ocrevus in the treatment of multiple sclerosis (MS). Both drugs target the CD20 protein on B-cells, reducing B-cell activity which is a key factor in slowing disease progression and relapse rates.

Figure 3

Currently, Ocrevus dominates the MS market. While Kesimpta’s patent extends to 2035, Roche’s patent for Ocrevus expires between 2028 and 2029 in the US and Europe. This may invite biosimilars that could disrupt pricing and erode Kesimpta’s market potential, limiting Novartis’ ability to fully capitalize on its patent protection [2].

Kisqali

Kisqali is a major player in the global breast cancer market, valued at around $26 billion, with over 50 percent of new prescriptions attributed to it [3]. It is effective in both early-stage and ER-positive breast cancer when used in combination with hormone therapy. Since 70–80 percent of breast cancer cases are ER-positive [4], Kisqali has the potential to address roughly three-quarters of the market. If prescriptions were to reach full potential, this could translate into approximately $20 billion in annual revenue, capturing a significant portion of the global breast cancer market. The drug selectively inhibits cyclin-dependent kinases 4 and 6 (CDK4/6), thereby preventing cancer cell proliferation. With patent protection extending until 2036 [5] and increasing clinical adoption, Kisqali is well-positioned to become one of Novartis’ highest-revenue products and a key asset in bridging future patent cliffs.

Novartis acquires Tourmaline Bio for ~1.4 billion USD

The acquisition of Tourmaline Bio strengthens Novartis’ pipeline. Tourmaline Bio currently has three drugs in development, with two in phase II clinical trials (https://substack.com/home/post/p-176847210). Its technology focuses on interleukin-6 (IL-6) modulation, a key cytokine in T-cell differentiation into Th17 cells. Th17 cells release IL-17, a driver of inflammation central to Cosentyx’s mechanism of action.

This acquisition opens the potential for Novartis to develop a dual IL-6 and IL-17 therapeutic, capable of selectively reducing IL-17–mediated inflammation. If successful, this could enhance Cosentyx revenue and expand Novartis’ capabilities in inflammatory diseases.

However, this potential must be approached cautiously. Phase III trials for the IL-6 blocking therapy are pending, and approval processes may take significant time. Combining it with Cosentyx, while potentially transformative, would likely extend timelines further. The global autoimmune disease market is valued at approximately $7 billion [6], highlighting that immediate commercial impact is not guaranteed.

Figure 4, https://www.google.com/url?sa=i&url=https%3A%2F%2Fwww.mdpi.com%2F1422-0067%2F24%2F5%2F4865&psig=AOvVaw1fIJgp2dj6CDKJpSdav6ds&ust=1761837350291000&source=images&cd=vfe&opi=89978449&ved=0CBgQjhxqFwoTCKDklN_ZyZADFQAAAAAdAAAAABAE

Potential Impact of the Acquisition

Novartis’ acquisition of Tourmaline Bio presents a fascinating scientific opportunity, though its financial rationale appears less straightforward. The deal could help bridge Novartis’ upcoming patent cliff, yet the underlying economics of monoclonal antibody therapies make profitability uncertain without major advances in production efficiency.

Tourmaline’s IL-6–blocking monoclonal antibody represents a scientifically compelling approach to targeting key inflammatory pathways. However, with the average production cost of monoclonal antibodies around USD 100 per gram and treatment prices often reaching several thousand USD per gram, large-scale commercial viability remains challenging. In Tourmaline’s Phase 2b study, doses of 20 mg and 50 mg per patient per administration already translate into several hundred USD per patient raising concerns about commercial viability.

If Novartis were to successfully integrate Tourmaline’s IL-6 inhibitor with its existing IL-17 portfolio, it could theoretically command a significant portion of the inflammatory pathway landscape. Scientifically, that would be a breakthrough, financially however, it would depend heavily on pricing power and payer acceptance in an increasingly cost-sensitive market.

The global anti-inflammatory market, valued at approximately USD 133 billion in 2024 and projected to surpass USD 270 billion by 2034 [7], offers scale but also fierce competition and growing regulatory scrutiny over biologic drug pricing. Overall, this acquisition may prove to be a strategic research investment rather than an immediate financial win - scientifically ambitious, but economically uncertain.

The Patent Cliff and Financial Implications

At present, Entresto remains Novartis’ primary revenue driver. Its upcoming patent expiry poses a significant patent cliff, exposing the company to competition from generic alternatives. A transition towards Cosentyx as the main income source is expected; however, total revenue may decline in the short term as newer products like Kisqali and Kesimpta scale to Entresto’s revenue levels.

Figure 5, All products of Novartis - forecasted revenue (using historic growth rates)

This temporary income reduction could constrain R&D budgets, increasing the risk of future patent cliffs. While historical data suggest that Novartis’ diversified pipeline may mitigate this risk, such forecasts depend heavily on past performance. Therefore, scaling newer drugs like Kisqali rapidly is essential to maintaining both short-term financial stability and long-term innovation capacity.

Stock Performance and Pipeline Alignment

Investors often perceive Novartis as a stable and secure investment due to its steady stock performance. However, this apparent stability is closely tied to the company’s ability to continually replenish its pipeline with new blockbuster drugs before older patents expire.

Figure 6

Novartis’ business model effectively mirrors venture capital principles, continuously investing in high-risk, high-reward assets such as drug candidates while occasionally acquiring entire companies to secure promising pipelines. This approach highlights the importance for investors to look beyond financial metrics and understand the scientific and clinical foundations of a company’s assets to anticipate both short-term volatility and long-term growth potential.

Conclusion

Novartis continues to exemplify the balance between scientific innovation and commercial strategy. Its success depends on sustaining a dynamic pipeline, managing patent transitions, and maintaining strong R&D investment despite revenue fluctuations. The company’s shift toward advanced prescription therapies reflects its commitment to high-value innovation and long-term competitiveness.

Figure 7

For both investors and analysts, Novartis serves as a case study in how a deep understanding of science is essential for evaluating financial stability and future market potential. It also illustrates how crucial the balance between exceptional science, commercialisation, and strategic acquisition can be for a pharmaceutical company. Maintaining that balance in the coming years will be vital for Novartis to preserve its global leadership and avoid the full impact of the approaching patent cliffs.

References:

1. https://dataintelo.com/report/cosentyx-drug-market

2. https://www.grandviewresearch.com/market-trends/top-20-multiple-sclerosis-drugs-market

3. https://firstwordpharma.com/story/5927904

4. https://www.breastcancer.org/pathology-report/hormone-receptor-status

5. https://pharsight.greyb.com/drug/kisqali-patent-expiration

6. https://www.databridgemarketresearch.com/reports/global-autoimmune-disease-treatment-market

7. https://market.us/report/anti-inflammatory-market/#:~:text=Key%20Takeaways,share%20of%2038.8%25%20in%202024

Have you ever wondered how stock markets actually function?

When you click “buy” or “sell” on a trading app, your order executes almost instantly. But have you ever stopped to think: who’s on the other side of that trade? Who ensures there’s always someone ready to buy when you want to sell, or sell when you want to buy?

The Second-Hand Book Problem

Imagine you want to sell a textbook for £10, but no buyers show up. You face two unappealing choices: wait indefinitely for someone willing to pay your price or lower it until a buyer finally agrees.

The next day, your classmate urgently needs that same book and would happily pay £10. However, there are no sellers available. They must either wait or pay £15 for a new copy at the bookshop.

This is inefficient. Both of you lost out because you couldn’t find each other at the right time.

Here’s where the market maker steps in. A market maker acts as a middleman who is always ready to buy and sell, even when no one else is available. They appear with a simple offer:

“I’ll buy the book for £9.90, and I’ll sell it for £10.10.”

Now you can instantly sell your book for £9.90 instead of waiting or dropping your price to £7. Your classmate can instantly buy the same book for £10.10 instead of paying £15 at the shop.

The market maker earns £0.20 profit per book, but in return provides immediacy, liquidity, and price stability for everyone.

(Figure 1)

In financial markets, the same principle applies. The book represents a share. The market maker continuously quotes two prices: an ask (buy) price at the market’s highest, and a bid (sell) price at the market’s lowest. The difference between the two is called the spread, which is how they make money.

This is precisely what Flow Traders does, except instead of textbooks, they’re trading trillions of euros in financial products each year.

Introduction: What is Flow Traders?

Flow Traders is a leading global market maker specialising in Exchange-Traded Products (ETPs). Founded in 2004 and headquartered in Amsterdam, the company provides liquidity across multiple asset classes and instruments, including ETPs, futures, spot products, fixed income, currencies, and crypto assets.

At its core, Flow Traders is a liquidity provider, continuously quoting bid and ask prices to ensure the smooth functioning of markets. This continuous presence ensures that investors can execute trades efficiently without causing dramatic price swings. The firm employs sophisticated high-frequency and quantitative trading strategies to manage risk while generating profit from the spread between buying and selling prices.

For a market-making firm such as Flow Traders, success depends on three key elements: trading volume, market volatility, and trading capital. In essence, higher trading volumes, greater market volatility, and wider bid–ask spreads all contribute to higher profitability.

Higher trading volumes create more opportunities to capture the bid–ask spread, meaning that the more trades are executed, the greater the potential for revenue generation.

During periods of heightened market volatility, bid–ask spreads tend to widen as uncertainty increases. This environment offers more profitable trading opportunities for market makers but also raises risk, requiring sophisticated risk management to maintain stability during market distress.

Finally, the amount of trading capital deployed directly determines a firm’s capacity to seize opportunities. Greater capital allows for larger position sizes, broader market coverage, and the ability to provide liquidity across a wider range of products and markets.

The Explosive Growth of the ETP Market

The Exchange-Traded Products (ETP) market has experienced remarkable growth over the past two decades, and this expansion directly benefits market makers like Flow Traders.

(Figure 2)

ETPs are investment vehicles that track an underlying index, basket of securities, or commodities. ETPs come in three main categories. Exchange-Traded Funds hold shares or bonds in proportion to an index. Exchange-Traded Commodities are designed to track single commodities or commodity indices. Exchange-Traded Notes are unsecured debt securities issued by banks with returns based on the performance of underlying securities. Despite their structural differences, all ETPs share similar advantages: they offer transparency, low management fees, minimal trading costs, and continuous intraday liquidity on secondary markets.

This sustained growth stems from fundamental shifts in investor preferences towards transparency, diversification, and efficiency. Perhaps most importantly, ETPs allow investors to gain exposure to markets that would otherwise be difficult or costly to access directly, such as emerging market bonds, commodity futures, or niche sector indices.

For Flow Traders, this structural growth translates directly into greater trading opportunities. The company now provides liquidity across more than 6,500 unique ETPs, representing over 40% of the global market.

The year 2020 was a defining example of this relationship. When the COVID-19 outbreak unleashed extreme volatility across global markets, investors rushed to rebalance their portfolios and manage risk through ETPs. While many financial institutions struggled with liquidity shortages, Flow Traders remained a reliable counterparty, ensuring that investors could buy and sell continuously, even as spreads widened. This period of intense trading activity and expanded spreads created ideal conditions for market makers.

The 2024 Capital Expansion Plan

To understand why Flow Traders made such a strategic shift in 2024, it is important to examine what happened in the years following its COVID-19 windfall.

(Figure 3)

At the time, the company maintained a shareholder-friendly dividend policy, distributing substantial portions of its annual profits to investors. In FY2020, it achieved a remarkable 164% return on average trading capital, generating exceptional profits. However, after paying out dividends, the trading capital base began to shrink. From €772 million in FY2020, trading capital declined to €611 million in FY2021.

This diminishing capital base created a vicious cycle. With less capital to deploy, Flow Traders was unable to capture as many opportunities, even when market conditions improved. The return on average trading capital fell to 56% in 2021. The pattern was clear: the company was distributing the very resource that generated its returns, effectively starving its own growth engine.

Management eventually realised that trading capital is the lifeblood of the firm, the fundamental constraint on profitability. The company had been in a self-defeating loop, distributing capital that could have compounded at attractive rates if retained and redeployed. Retaining capital would allow Flow Traders to take larger positions, provide liquidity across more products, absorb temporary hedging losses, and fully participate in volatile markets where opportunities multiply.

This insight became particularly pressing as management observed the ETP market’s continued structural growth. Global ETP AuM were climbing towards record levels, institutional trading was expanding, and new asset classes such as digital assets were creating fresh opportunities. Yet Flow Traders found itself capital-constrained, unable to fully capitalise on the growth before it.

Therefore, in July 2024, then-CEO Mike Kuehnel and the Board implemented a decisive trading capital expansion plan. Regular dividend payments were suspended, and the dividend policy was amended to prioritise capital accumulation. Profits realised during the financial year were to be fully appropriated to increase and strengthen reserves.

Though Kuehnel stepped down in August 2025 to pursue a new opportunity, his successor, Thomas Spitz, has reaffirmed his commitment to this capital expansion strategy. This continuity in strategic direction signals that the capital retention policy will remain firmly in place for the foreseeable future.

The results were immediate. Trading capital grew by 33% over 2024, rising from €584 million to a record €803 million by the end of the first quarter of 2025, representing the highest level in the company’s history.

Company Valuation

The valuation of Flow Traders focuses on trading capital rather than earnings. Traditional Discounted Cash Flow (DCF) models were considered but ultimately deemed unsuitable for this analysis. For a market-making firm such as Flow Traders, a DCF approach would require estimating numerous complex inputs, including future trading volumes, volatility, bid-ask spreads, capital efficiency, and the cost of equity. Each of these variables fluctuates with market conditions, making reliable cash flow forecasting highly uncertain.

Instead, this analysis adopts a valuation multiple approach, similar in principle to a price-to-earnings (P/E) ratio, but based on trading capital. In Flow Traders’ case, trading capital is the key driver of profitability, the fundamental input that generates returns. The question becomes: what multiple should the market pay for each euro of trading capital, given the returns that capital can generate? This approach offers clarity by focusing on the stable, measurable input (capital deployed) rather than the volatile output (periodic earnings).

Our model constructs future valuations based on two critical variables: projected trading capital growth and the market’s valuation multiple of that capital. The methodology proceeds as follows.

(Figure 4)

First, we establish a baseline growth rate for trading capital. Using data from the first half of 2024 through the first half of 2025, the period since implementing the capital expansion plan, we calculate the average half-yearly growth rate in trading capital. This period represents the most relevant baseline as it reflects the new dividend suspension policy and management’s commitment to capital accumulation. The observed growth rate is 12.79% per half-year during this period.

Applying the 12.79% half-yearly growth rate to Flow Traders’ current trading capital base generates a clear trajectory. From €831 million at the end of H1 2025, trading capital grows steadily each half-year, reaching almost €4,000 million by the end of 2031.

(Figure 5)

Second, we determine the appropriate valuation multiple for trading capital. By collecting stock price data corresponding to each financial reporting date that disclosed trading capital figures, we calculate the ratio of market capitalisation to trading capital. This metric represents the market capitalisation investors assign to each euro of trading capital.

(Figure 6)

The behaviour of the stock price to trading capital ratio under different strategic outcomes forms the foundation of our scenario analysis.

Under our base scenario, we assume that this multiple remains relatively constant at its post-expansion average of 0.03. For market-making firms, this ratio should theoretically remain stable if the return on trading capital stays consistent. Unlike operating companies, where competitive dynamics can dramatically shift returns on invested capital, market makers operate in a more structurally stable environment. Bid-ask spreads may widen or narrow with volatility, but the fundamental economics of providing liquidity do not transform overnight. If Flow Traders maintains its competitive position and continues generating similar returns on trading capital, the market should continue valuing that capital at comparable multiples. Applying the 12.79% half-yearly growth rate in trading capital, this base scenario projects the share price reaching approximately €125 per share by the end of 2031.

The bullish scenario models a case where the new strategy not only accumulates capital successfully but also demonstrates improving capital efficiency. We apply the average growth rate of the stock price to trading capital ratio observed during the post-expansion period. The strategic shift from dividend distribution to capital accumulation, if executed effectively, could fundamentally enhance the quality of earnings and justify higher valuations. Under this scenario, the share price could reach approximately €199 per share by the end of 2031.

The bearish scenario assumes that the market fails to recognize the capital expansion plans benefits and keeps pricing the stock along its historic stock / trading capital ratio’s trend. We apply the average historical growth rate of the stock price to trading capital ratio from the pre-expansion period before 2024 H1. This declining ratio reflects the market’s historical valuation pattern when the company prioritised returning capital to shareholders rather than reinvesting for growth. Without the benefits of scale and with capital growth limited by dividend payments, the ratio would likely revert to its historical trajectory. Under this scenario, the share price would grow more modestly to approximately €67 per share by the end of 2031.

Competition

When evaluating Flow Traders’ prospects, it is essential to consider the broader competitive landscape of the global market-making industry. Key competitors include Jane Street, Optiver, and IMC, all leading proprietary trading firms with significant exposure to Exchange-Traded Products (ETPs) and some big banks. These competitors operate with similar high-frequency and quantitative trading strategies, providing liquidity across global markets.

Jane Street is widely regarded as the dominant force in ETP trading, particularly in the United States, where it consistently ranks among the largest liquidity providers for ETFs. Optiver, headquartered in Amsterdam like Flow Traders, has built strong positions in derivatives and ETPs across both European and Asia-Pacific markets. Compared with these larger and more diversified rivals, Flow Traders maintains a sharper focus on ETPs as its core business segment. This specialisation enables operational efficiency and expertise but also concentrates its risk exposure in a single product category.

Acknowledging these concentration risks, Flow Traders is diversifying into adjacent areas including connectivity, platforms, data, and tokens. Whilst these remain related to ETP trading, they represent an effort to build new revenue streams beyond pure market-making, mirroring the diversification strategies of larger competitors.

The competitive dynamic in ETP market-making remains intense yet structurally supportive. The continuous expansion of ETP listings globally means that multiple firms can coexist profitably. However, maintaining technological superiority, effective risk management, and access to deep trading capital are critical to preserving market share. Flow Traders’ recent capital expansion strategy can therefore be seen as a defensive and strategic response to the scale advantages of its larger peers.

Investment Implications

This valuation framework suggests that Flow Traders’ decision to prioritise capital accumulation could drive substantial value creation if the company maintains reasonable returns on its growing capital base. Unlike mature financial firms that struggle to deploy incremental capital productively, Flow Traders operates in a structurally growing market with clear capacity to absorb much larger trading capital than it currently possesses.

However, the investment case rests entirely on two variables: continued capital growth and sustained returns on that capital. Investors backing this thesis are essentially betting that Flow Traders can compound capital at high rates for an extended period without suffering competitive or regulatory setbacks. The five-year projection to 2030 represents a long runway during which numerous factors could disrupt the trajectory.

For investors comfortable with this volatility and uncertainty, the current valuation may offer attractive entry points if one believes in the structural growth of ETPs and Flow Traders’ competitive positioning. For those seeking more predictable returns or uncomfortable with earnings volatility, the stock likely remains too risky. The model provides a framework for thinking about valuation, but ultimately each investor must assess whether Flow Traders’ unique business model and strategic direction aligns with their portfolio objectives and risk tolerance.

Figure 1. Overview of a standard R&D pipeline with different stages. The exact values for candidates moving from one stage to the next varies by company and research facility are however very low in pre-clinical testing and become more likely with progression towards commercialisation.

This article is written by

The research and development (R&D) pipeline forms the backbone of every pharmaceutical company, representing not only its current portfolio but also its prospects for future growth. Understanding the structure, stages, and dynamics of a company’s pipeline offers investors deep insight into its innovation capacity, risk exposure, and long-term valuation potential. Analysing how efficiently a company develops and progresses drug candidates through the R&D process can therefore be a powerful way to identify investment opportunities and evaluate corporate strategy.

Discovery and Preclinical Research

The earliest stages of drug discovery determine whether an idea has the potential to evolve into a viable therapeutic product. Companies at this stage must assess both the unmet medical needs in the market and their internal capabilities to develop, approve, and commercialise a new therapy. This phase can be divided into four sequential steps. It begins with the generation of an idea that addresses a market gap and fits the company’s scientific expertise. Once a potential drug target has been identified, in silico or computational modelling is employed to predict how that molecule might interact with biological systems. These “dry-lab” simulations, often supported by artificial intelligence, help researchers screen numerous candidates rapidly and cost-effectively before moving to laboratory experiments. Artificial intelligence is particularly valuable in identifying the most promising compounds early in the process and eliminating those unlikely to produce meaningful therapeutic results. Meaning that scientists can spend more of their time on targeted research.

https://www.assaygenie.com/in-vitro-vs-in-vivo-complete-comparison-selection-guide-research-methods/?srsltid=AfmBOop-r_mIMVJm6M6DxhXUnHY10qTRDhKc-saMDERXNCyQqYshz6rK

In vitro testing follows, where experiments on cell cultures explore the mechanism of action and toxicity of the compound. The resulting data are then refined using bioinformatic analysis to guide further optimisation. If these results are positive, and the proposal passes ethical review, the compound proceeds to in vivo testing in living organisms to evaluate safety before it can be considered for human trials. This preclinical phase is highly complex and scientifically demanding, with very few candidates ultimately advancing to clinical stages. For investors, observing how efficiently a company navigates these steps, particularly through external collaborations with universities or other biotechnology firms can indicate how effectively it allocates R&D resources and mitigates costs.

Clinical Development

Once a candidate demonstrates sufficient promise in preclinical testing, it enters clinical development, where it is tested in humans to confirm safety and efficacy. This process unfolds across three main phases, each carrying distinct objectives, timelines, and financial implications.

Phase I trials involve small groups of healthy volunteers, typically between twenty and one hundred individuals. These studies primarily assess the safety and dosage range of the drug. Although failure rates remain high at this stage, the relatively low participant numbers make it a less expensive stage of development. Phase II trials expand testing to several hundred patients suffering from the target condition, focusing on whether the drug produces the desired therapeutic effect. Because this stage provides the first real evidence of efficacy, it is often regarded as the “go or no-go” decision point. Successful results here can significantly boost investor confidence and drive stock value, while failure can result in sunk costs and lost momentum.

Phase III represents the decisive stage of clinical development, testing the candidate on thousands of patients to confirm efficacy, monitor side effects, and collect comprehensive data for regulatory approval. This phase is the most costly and time-consuming. Companies frequently seek partnerships or licensing agreements to share the financial burden and operational risks associated with these large-scale trials. A notable example is the collaboration between BioNTech and Pfizer for the COVID-19 vaccine, where Pfizer’s global production and distribution infrastructure enabled BioNTech’s innovation to reach the market at scale. Such collaborations often determine whether a smaller biotech firm can translate promising science into a commercially viable product.

Drug Approval

Following successful Phase III results, a company can apply for regulatory approval. In Europe, this process is overseen by the European Medicines Agency (EMA), which reviews all safety and efficacy data before issuing a recommendation to the European Commission. The duration and outcome of this process depend on several factors, including the urgency of medical need, available treatment alternatives, and the quality of manufacturing facilities. For example, during the COVID-19 pandemic, vaccines underwent accelerated review due to the global health emergency. Conversely, drugs entering crowded therapeutic markets tend to face longer and more stringent evaluations.

Approval marks a pivotal transition point for any pharmaceutical company. It signals the beginning of potential revenue generation and the recovery of previous R&D investments. However, the process can still be derailed by concerns regarding safety, efficacy, or production standards. Companies must possess facilities that meet Good Manufacturing Practice (GMP) requirements before they can sell the drug commercially. For investors, examining a company’s regulatory track record, manufacturing readiness, and ability to navigate approval timelines offers crucial insight into its operational resilience and likelihood of successful market entry.

Post-Marketing and Lifecycle Management

After approval, a drug enters the post-marketing or commercialisation phase. This stage involves ongoing safety monitoring, studies into potential combination therapies, and efforts to expand the drug’s approved uses through new clinical trials. These “label expansions” can significantly increase a product’s market potential by allowing its use in additional therapeutic areas. At the same time, companies must prepare for eventual patent expiry. Once exclusivity lapses, generic or biosimilar competitors can enter the market, leading to price reductions and shrinking market share. Effective lifecycle management, through reformulations, new delivery systems, or next-generation products, is therefore essential to sustain long-term revenue. Investors should pay close attention to how well a company manages its intellectual property portfolio and prepares for these transitions.

The Strategic Value of the Pipeline

A pharmaceutical company’s R&D pipeline acts as both a risk map and a growth indicator. A diversified pipeline, with assets distributed across multiple stages and therapeutic areas, provides resilience against individual project failures. However, specialisation can also be advantageous. Companies that dominate a particular therapeutic domain can leverage their expertise and reputation to maintain market leadership. Novo Nordisk, for instance, controls roughly one-third of the global diabetes market, benefiting from a strong brand presence and deep technical knowledge. Analysing a company’s balance between diversification and specialisation offers valuable clues about its strategic orientation and stability.

Partnerships and Collaborations

Partnerships are a defining feature of modern pharmaceutical innovation. Collaborations with universities, research institutes, and biotech firms can accelerate discovery, reduce risk, and lend scientific credibility to projects. For large pharmaceutical corporations, acquiring or licensing from smaller biotechs provides access to cutting-edge research that might otherwise take years to develop internally. However, such arrangements often come at the cost of shared profits, as royalties or milestone payments are due to the partner upon commercial success. Investors should therefore assess not only the existence of partnerships but also their structure, financial implications, and strategic rationale. Well-designed collaborations can drive innovation and value creation, while poorly managed ones can erode margins.

Financial Implications for Investors

R&D should be viewed not merely as a cost but as a long-term investment in future earnings potential. The proportion of revenue allocated to R&D expenditure, the rate of candidate progression through the pipeline, and the likelihood of developing a blockbuster therapy all influence a company’s financial outlook. Understanding timelines and attrition rates across pipeline stages enables investors to forecast cash flow needs and future revenue streams. High R&D efficiency, defined by the ability to deliver successful approvals relative to investment, often correlates with superior long-term performance.

Conclusion

The R&D pipeline is the clearest window into a pharmaceutical company’s future. It reflects scientific capability, operational discipline, financial strategy, and the capacity for innovation. Each stage of the pipeline, from discovery to post-marketing, reveals different aspects of a company’s strengths and vulnerabilities. A well-balanced and efficiently managed pipeline demonstrates how effectively an organisation can transform research spending into sustainable long-term value.

For investors, analysing the composition and maturity of a pipeline provides a forward-looking view of potential earnings, risk diversification, and growth sustainability. It indicates whether a company’s strategy focuses on breakthrough innovation or incremental product development and whether it can remain competitive in the face of regulatory challenges and patent expirations. Strategic partnerships, the pace of clinical progression, and the success rate of drug approvals all offer insight into a company’s resilience and adaptability in an increasingly complex healthcare environment.

Ultimately, the R&D pipeline functions as both a scientific roadmap and a financial forecasting tool. It connects laboratory discoveries with market opportunities and shows how effectively a company can translate innovation into real-world impact. For investors seeking to identify pharmaceutical firms positioned for long-term growth, profitability, and societal relevance, few indicators are as revealing as the strength and dynamism of their research and development pipelines.

The ECB’s main goal is price stability, their aim for inflation is 2%. In recent years the ECB had a difficult task of maintaining these goals during the pandemic, and the period of low inflation in mid 2010’s. The ECB has introducted a new method of influence called the asset purchasing programme starting in 2014. How has this influenced the bond market?

Asset Purchase Programme ECB

These less-known policies of the ECB are called the asset purchase programmes (APP) and the pandemic emergency purchase programme (PEPP). This is a program of monetary policy of the ECB that aims at creating more price stability, i.e., inflation, by controlling the long term bond yield. These purchase programmes come down to the ECB buying bonds and securities from governments and corporations, exact specifications on this programme can be found on the ECB site European Central Bank (2025). The consequence of this action is the increased amount of money in the markets. The ECB can boost long term bond yield when the inflation rates are too low, which happened in 2013-2020. The APP program has started in 2014, this is a relatively new program and the longer term consequences are not deterministically known. Altavilla et al. (2019) has done research on the effect and has found a significant negative effect of the purchasing programme on the 10YR bond yield, which slowly decays years after the purchasing. The half life of the effect is estimated to be at 5 years according to the research paper. What happens after the purchasing programme is stopped is not known with certainty. The PEPP was introduced during COVID to counteract the risks of the pandemic in Europe’s economy. Below is a bar chart of the cumulative sum of the purchasing of the assets. The ECB has decided to stop this purchasing after July 2023, and let all the bought bonds and securities mature.

Figure 1: ECB APP + PEPP cumulative, stacked

Government bonds in Europe

The goal of the PPEP and APP is to control inflation by influencing the long term interest rates, which is the same as the long term bond yield. The research paper Altavilla et al. (2019) attributes the bond yield reduction to the lower amount of long-term bonds available, flattening the yield curve. I want to investigate with a simple model if we still see this effect with more recent data. I used the 10YR daily bond yield of five European countries and converted it to monthly yield changes. Below is a graph of the bond yield over time. You can see in this graph that when the ECB stopped purchasing new bonds in 2022/2023 the 10YR bond yields have started to rise again. This is in line with the findings of this earlier released research paper.

Figure 2: 10-Year Government Bond Yields (Daily)

We can also see that after the eurozone debt crisis in 2008-2010 the bond yield for governments diverged. The 10YR bond yield of Spain and Italy rose to a significantly higher level compared to the other countries. Spain and Italy has more debt, and were less stable economies compared to France, Germany, and the Netherlands. Since then the bond yield have slowly started converging again, and currently they are quite close.

The model

We estimate, for each countries 10 year bond yield (sector) y_{i,t}

y_{i,t} = α_i + β_{APP}*APP_t + β_{PEPP}*PEPP_t + u_{i,t}, i ∈ {NL, DE, FR, IT, ES}, t = 1, . . . , T,

With this simple model we can check if we can still clearly see the findings of the previous elaborate research paper on this topic. To estimate a model similar to the research paper would be to elaborate for a short article. We estimate if the coefficients of the changes in the APP and PPEP programme are significant on the 10YR bond yields of the selected countries. Using our found data we have the following results:

We see that for every governments 10YR bond yield the purchasing programmes have the negative effect that was found earlier. An increasing in asset purchasing in a given month tends to decrease the 10YR bond yield. This might also indicate we can inspect more rise in the 10YR bond yield in the near future, since the ECB has stopped with the purchasing programme. The estimated coefficient of Italy’s and Spain’s bond yield is lower compared to the others. This could be explained by their earlier said financial position after the eurozone debt crisis.

Conclusion

We have found evidence that with newer data the findings of Altavilla et al. (2019) still seem valid. The asset purchasing programme of the ECB significantly lowers the long term bond yield of European government bonds. Now the question is what will happen to the long term bond yield with the purchase programme maturing? It seems likely to me that the 10YR bond yield will increase slowly in the near future. As seen in the paper Altavilla et al. (2019), the effect of the purchasing programme is estimated to be years long. So the effect of the purchasing programme will likely not abruptly dissapear completely.

References

Altavilla, Carlo, Giacomo Carboni, and Roberto Motto (2019, July). The transmission of the ecb’s asset purchase programme: what do we know so far? ECB Working Paper Series 2293, European Central Bank. European Central Bank Working Paper No. 2293.

European Central Bank (2025). Asset purchase programmes — corporate sector purchase programme (cspp).

Viral Infections

Viral infections constitute the largest disease category by patient volume, with virtually every individual experiencing at least one viral infection annually. The viral pathogen landscape encompasses extraordinary diversity, from ubiquitous respiratory viruses causing common colds to pandemic threats like SARS-CoV-2, and chronic infections like HIV. With over 300k distinct viral species identified that infect mammals (Virology Blog), comprehensive scientific understanding remains impossible, focusing research efforts on clinically significant pathogens with the greatest public health impact.

Viral evolution through mutation represents a fundamental challenge for therapeutic development. During cellular infection, viruses commandeer host cellular machinery to replicate their genetic material. This replication process lacks the proofreading mechanisms of cellular DNA replication, resulting in frequent random mutations that can alter viral properties including transmissibility, pathogenicity, and drug resistance. RNA viruses like influenza and coronaviruses demonstrate particularly high mutation rates, necessitating continuous surveillance and therapeutic adaptation.

The $240 billion market valuation (Grand View Research) approaches oncology in magnitude despite serving exponentially more patients, reflecting the generally self-limiting nature of most viral infections. This dynamic creates a volume-dependent economic model exemplified by seasonal influenza vaccines, priced at €20-60 but requiring mass administration for profitability. Current vaccine technologies predominantly utilize inactivated viral particles or recombinant proteins to elicit protective immunity against specific viral antigens, though newer platforms like mRNA vaccines demonstrate enhanced adaptability to emerging variants.

Pandemic events introduce unique market dynamics that disrupt traditional growth patterns. The COVID-19 pandemic illustrated how global health emergencies can simultaneously devastate broader markets while creating extraordinary opportunities for companies developing relevant countermeasures. BioNTech’s rapid ascent exemplifies this potential, generating billions in revenue through mRNA vaccine development. Increasing globalization and human-environment interactions heighten pandemic risk, suggesting periodic market disruptions will continue.

More information on BioNTech: https://futurefunds.substack.com/p/was-the-biontech-pfizer-shot-in-your

While routine market growth remains constrained by saturation, breakthrough opportunities persist in addressing chronic viral infections. Curative therapies for conditions like HIV could establish monopolistic market positions analogous to successful oncology drugs, fundamentally altering competitive dynamics within specific therapeutic niches. Investment strategies must balance steady returns from established vaccines and antivirals against the transformative potential of novel therapeutic modalities targeting previously intractable viral diseases.

Bacterial Infections

Bacterial infections present unique classification challenges in medical markets, as bacteria exist along a spectrum from essential microbiome constituents to dangerous pathogens. While all humans harbor bacterial communities, approximately 300 million individuals currently suffer from pathogenic bacterial infections, with an estimated 800 million new cases arising annually (calculated an estimate from Wikipedia data). Though this incidence appears lower than viral infections, the fundamental differences in bacterial biology necessitate entirely distinct therapeutic approaches and market dynamics.

Unlike viral infections, which often rely on preventive vaccination, most bacterial infections require active treatment with antibiotics, compounds that either eliminate bacteria directly (bactericidal) or inhibit their growth sufficiently for immune clearance (bacteriostatic). However, the therapeutic landscape faces an escalating crisis: antimicrobial resistance. Through natural selection, bacteria harboring resistance-conferring mutations survive antibiotic exposure and proliferate, eventually dominating bacterial populations. This evolutionary pressure is compounded by horizontal gene transfer via conjugation, a process enabling bacteria to share genetic material, including resistance genes, across species boundaries. The resulting multi-resistant organisms pose increasingly severe clinical challenges as sequential antibiotics fail.

https://www.thelancet.com/journals/lancet/article/PIIS0140-6736%2821%2902724-0/fulltext

Current strategies involve reserving certain antibiotic classes exclusively for multi-resistant infections, attempting to preserve their efficacy. However, this reserve arsenal continues shrinking while antibiotic discovery has stagnated throughout the 21st century. The convergence of rising resistance and diminishing innovation creates an approaching therapeutic cliff with profound market implications.

At $57 billion (BioSpace), the antibacterial market appears critically undervalued given both current patient burden and the looming resistance crisis. This valuation disconnect represents both risk and opportunity. Novel therapeutic modalities, particularly non-antibiotic approaches, could fundamentally restructure market dynamics by providing alternatives to traditional antibiotics while preserving remaining reserves. Additionally, precision therapeutics targeting specific pathogens while sparing beneficial microbiome bacteria would address a major limitation of current broad-spectrum antibiotics. Microbiome disruption from conventional antibiotics increasingly links to secondary health complications including depression, dermatological conditions, metabolic disorders, and inflammatory diseases, creating additional market opportunities for microbiome-sparing alternatives. Companies successfully developing next-generation antibacterial therapies could capture premium pricing in a market facing existential therapeutic challenges, making this sector particularly attractive for breakthrough innovation investment despite current modest valuations. Many big companies like GSK, Pfizer or Novartis are researching drugs against resistant bacteria, but are not specialized on those. Many of the small companies did not IPO yet or those that have in the past have mostly failed and have essentially lost 95% or more of their market cap. So need to wait for IPO for proper investment

Fungal and Invasive Fungal Infections

https://www.australiawidefirstaid.com.au/resources/common-types-of-fungal-infection

Fungal infections affect an extraordinary 1.6 billion individuals globally (Grand View Research), representing one of the most prevalent yet underrecognized disease categories. The critical clinical and market distinction lies between invasive and non-invasive fungal infections. Non-invasive fungi remain localized to specific anatomical sites, typically skin, nails, or mucous membranes, causing mild symptoms amenable to topical antifungal therapy. In contrast, invasive fungal infections disseminate systemically, creating life-threatening conditions requiring aggressive systemic treatment. Despite the massive infected population, only 7 million of the 1.8 billion annual cases represent invasive, clinically significant infections.

The antifungal market, valued at $17 billion globally (Grand View Research), projects modest 4% annual growth through 2030. However, this growth trajectory faces significant uncertainty from environmental factors. Climate change and extreme weather events create unpredictable effects on fungal ecology, potentially expanding geographical ranges of pathogenic species while simultaneously altering optimal growth conditions. These environmental shifts could either accelerate market growth through increased incidence or constrain it through changing disease patterns.

Market dynamics are further complicated by awareness paradoxes. Increased public health awareness could drive diagnosis and treatment rates, expanding the market. Conversely, enhanced awareness might improve preventive behaviors, reducing transmission and market size. Current therapeutic options, while adequate for most non-invasive infections, offer limited breakthrough potential, constraining innovation-driven growth. The established nature of topical and systemic antifungals creates a mature market with limited differentiation opportunities, suggesting steady but unremarkable returns for investors seeking stable revenue streams rather than explosive growth potential.

Parasitic Infections Including Protozoa

Parasitic infections represent one of the largest infectious disease categories by patient volume, affecting over 5 billion individuals annually. Current prevalence estimates indicate 1.5 billion people harbor active parasitic infections, with 3.5 billion new or recurrent cases yearly (Food Control Vol 145 March 2023, Springernature Gut pathogens 16 article 7 in 2024). This disease category encompasses diverse organisms: parasites that obligately live in or on hosts, free-living protozoa that can cause opportunistic infections, helminths (parasitic worms), and ectoparasites like lice that inhabit host surfaces.

Despite this enormous disease burden, the antiparasitic market remains surprisingly modest at $18 billion, projected to grow at 5% annually through 2030 (Grand View Research). This apparent market-disease mismatch reflects the predominant geographical distribution of parasitic diseases in resource-limited settings where purchasing power remains constrained.

https://www.nature.com/articles/s41598-025-07653-8#Fig1

Market evolution faces competing forces. Improving sanitation and hygiene standards globally should reduce transmission of many parasitic diseases, particularly soil-transmitted helminths and water-borne protozoa. However, climate change creates countervailing pressures, potentially expanding vector ranges for diseases like malaria. The Anopheles mosquito’s geographical expansion directly correlates with malaria transmission zones, suggesting climate-driven market shifts rather than pure growth.

Investment strategies must account for this geographical and epidemiological repositioning. Companies with diverse antiparasitic portfolios can capitalize on shifting disease patterns, capturing growth from expanding vector-borne diseases while managing decline in sanitation-sensitive infections. This market rewards adaptability over single-product focus, as the antiparasitic landscape undergoes fundamental restructuring driven by environmental and developmental changes rather than therapeutic innovation. Success requires anticipating disease migration patterns and positioning products accordingly, making this a uniquely dynamic market despite modest overall growth projections.

Conclusion

Non-communicable diseases continue to serve as the structural backbone of healthcare investment portfolios, offering durability through stable patient populations and long-term therapeutic adherence. They anchor predictable revenue streams and provide insulation from short-term volatility, ensuring reliable cash flows for companies with established franchises in cardiovascular care, diabetes management, respiratory therapeutics, and chronic neurological conditions. Yet stability alone does not deliver competitive differentiation. The most successful strategies balance this foundation with exposure to areas where clinical unmet need intersects with financial undervaluation, diabetes and neurology being two of the most prominent examples.

On the other side of the spectrum, communicable diseases inject dynamism into the market. They remain susceptible to external shocks, whether from emerging pandemics, climate-driven shifts in pathogen distribution, or the accelerating crisis of antimicrobial resistance. While such volatility carries risk, it simultaneously presents asymmetric upside. Companies with scalable platforms, such as mRNA vaccine technologies, broad-spectrum antivirals, or novel antibacterial modalities, are uniquely positioned to translate scientific breakthroughs into market monopolies within compressed timeframes. The COVID-19 experience demonstrated that global health emergencies can rapidly redirect capital flows and reset competitive hierarchies, creating unprecedented opportunities for those positioned at the intersection of innovation and preparedness.

Taken together, the global life science market represents not simply a defensive growth sector but a structurally transformative industry where demographic trends, scientific discovery, and financial engineering converge. Optimal portfolio construction therefore requires a dual mindset: conservative allocation into mature chronic disease markets that underpin resilience, and forward-leaning investment into high-margin or platform-driven innovations that can redefine entire therapeutic areas. The ability to identify undervalued markets with expanding patient populations, diabetes, neurology, or antimicrobial resistance, will likely be a key determinant of superior returns over the coming decade due to the nature of their growing patient base or pathogenic situation.

With projected annual growth exceeding 10% through 2033, the life sciences are more than a healthcare necessity, they are a financial imperative. Investors who can integrate epidemiological foresight with capital discipline stand to capture not only reliable compounding returns, but also the outsized rewards that accrue to those anticipating the next breakthrough in global health.

Non-communicable diseases make up nearly two-thirds of the global healthcare market, despite accounting for only about one-quarter of patients. Defined by their non-transmissible nature and long duration - often lifelong - non-communicable diseases provide a stable, predictable patient base. This translates into steady growth opportunities, with less exposure to short-term shocks such as pandemics, but significant upside potential when breakthrough therapies, particularly in oncology, have an impact on the market.

Cardiovascular Diseases

https://www.bhf.org.uk/-/media/files/for-professionals/research/heart-statistics/bhf-cvd-statistics-global-factsheet.pdf

Cardiovascular diseases remain the leading cause of global mortality. Coronary heart disease alone claims over 9 million lives annually (British Heart Foundation), with total cardiovascular disease affecting over 600 million patients globally and adding approximately 70 million new cases yearly (Journal of the American College of Cardiology Vol. 80 Issue 25 pages 2372-2425 20-27 December 2022).

The cardiovascular drugs market, valued at $150 billion (Precedence Research), encompasses pharmaceutical interventions and does not include medical devices such as stents. The market composition varies significantly between acute and chronic care segments. Emergency interventions combine surgical procedures with pharmaceutical support, while chronic management relies on established drug classes, from affordable options like aspirin for thrombosis prevention to specialized medications commanding premium prices.

https://www.precedenceresearch.com/cardiovascular-drugs-market

Market dynamics differ markedly between acute and chronic cardiovascular care. Emergency treatment involves immediate interventions, while chronic management represents a mature market featuring beta-blockers, anticoagulants, and other maintenance therapies. Patient adherence to prescribed regimens creates market stability with limited switching between medications. However, aging demographics ensure continued market growth as cardiovascular disease prevalence increases with age in an overall fairly developed market.

Cancer

Cancer encompasses diverse cellular dysfunctions leading to uncontrolled growth, representing another major cause of global mortality. Despite affecting only 53 million individuals currently (WHO), with 20 million new diagnoses annually, the high mortality rate and rapid disease progression mask the true market size and therefore its potential. Improved survival rates could substantially expand the patient base requiring long-term management and follow-up care.

The oncology market’s $320 billion valuation (Global Market Insights) reflects exceptional revenue per patient, creating a “winner-takes-all” dynamic. Companies developing successful treatments can establish effective monopolies for specific cancer types, given the extensive R&D costs and lengthy clinical trials required. Rising life expectancy correlates with increased cancer incidence, as extended lifespans provide more opportunities for cellular damage to evade immune detection.

https://www.nature.com/articles/s41392-020-00213-8

The cancer market presents a paradox: rapid financial movements contrast with slow clinical development. While trials require extended periods to demonstrate efficacy, investment speculation drives volatile market valuations based on patents, pipeline potential, and breakthrough hopes. This uncertainty encourages portfolio diversification strategies, as successful developments can generate returns that dwarf losses from unsuccessful ventures.

More Information: https://futurefunds.substack.com/p/investing-in-the-cure-the-financial

Chronic Respiratory Diseases

Chronic respiratory diseases (CRDs) represent long-term conditions affecting airways and lung structure. Asthma exemplifies this category, a chronic inflammatory condition causing reversible airway constriction that affects 350 million people globally (PubMed 38060225). Management typically requires dual inhaler therapy: preventive maintenance and acute rescue treatment.

Environmental factors increasingly drive CRD prevalence. Pollution, climate change, and extreme weather patterns contribute to rising patient numbers. The structural impact of CRDs on respiratory systems typically results in lifelong conditions requiring continuous management.

https://www.researchgate.net/figure/Expected-air-pollution-increase-in-without-applying-the-policy_fig6_321698370

The $80 billion (Zion Market Research) CRD market appears modest relative to the extensive patient base. Similar to diabetes, patients rarely change established medication regimens after initial adaptation. Future growth potential lies in both expanding prevalence due to environmental factors and technological advances enabling permanent cures rather than symptom management. This dynamic could disrupt current treatment paradigms, creating monopolistic opportunities for curative therapies while potentially diminishing markets for maintenance treatments. Investment strategies must consider patient preferences, as most would likely choose expensive one-time cures over cheaper but continuous treatments.

Diabetes

https://en.wikipedia.org/wiki/Epidemiology_of_diabetes Prevalence (per 100 inhabitants) of diabetes worldwide in 2021 among population ages 20 to 79

Diabetes represents one of the most prevalent chronic diseases globally, affecting over 800 million patients according to the WHO. The condition manifests in two primary forms: Type 1 diabetes, characterized by pancreatic beta-cell destruction resulting in insulin deficiency, and Type 2 diabetes, marked by progressive insulin resistance and relative insulin deficiency. Both variants necessitate lifelong dietary management and often require exogenous insulin administration, imposing substantial lifestyle modifications on patients. The global diabetes burden continues expanding, with approximately 10 million new diagnoses annually based on Diabetes Atlas projections. As seen in the prevalence map above some other factors also drive diabetes prevalence, for example in Egypt around 90% of the population was diagnosed with Hepatitis C. Insulin pathways within the body can be affected by a Hepatitis C infection therefore being a driving factor in the Diabetes prevalence in Egypt.

The chronic nature of diabetes creates a uniquely stable market environment. Patients typically establish treatment regimens early in their disease course and demonstrate remarkable therapeutic loyalty, maintaining consistent medication patterns throughout their lives. This market remained relatively static for decades until Novo Nordisk’s recent introduction of GLP-1 receptor agonists, representing the first major therapeutic innovation in diabetes management. These agents serve dual metabolic functions, glycemic control and weight reduction, creating significant market overlap between diabetes and obesity therapeutics.

https://www.novonordisk.com/content/dam/nncorp/global/en/investors/pdfs/financial-results/2025/Q1-2025-investor-presentation.pdf Breakdown of Q1 Novo Nordisk sales showing growing share of GLP-1 Drugs

The current $90 billion market valuation (Grand View Research) appears modest relative to the extensive patient population, reflecting a substantial gap between diagnosed and undiagnosed cases. Estimates suggest hundreds of millions of individuals remain undiagnosed globally, representing untapped market potential. New therapeutic entrants face the challenge of overcoming established patient loyalty to traditional insulin therapy, requiring compelling advantages in efficacy, convenience, or cost-effectiveness. However, GLP-1 agonists demonstrate strong adoption among newly diagnosed patients while also capturing the lucrative weight management market.

Market dynamics favor steady, sustainable growth driven by multiple factors. Improved diagnostic capabilities in developing nations will identify previously undiagnosed patients, while next-generation therapeutics attract new patient starts. Companies specializing in GLP-1 receptor agonists are particularly well-positioned, benefiting from dual market exposure and expanding indications. This convergence of diabetes and weight management markets, combined with growing disease prevalence and improving global healthcare access, creates a robust investment opportunity with predictable growth trajectories.

More Information: https://futurefunds.substack.com/p/novo-nordisk-and-eli-lilly-the-glp

Neurological Conditions

Neurological disorders affected nearly half the global population in 2021 (The Lancet Neurology), establishing neurology as one of medicine’s most significant therapeutic areas. This diverse category encompasses conditions ranging from acute cerebrovascular events to chronic neurodegenerative diseases like Alzheimer’s, and from episodic conditions like migraines to progressive disorders affecting motor function. These conditions share the common feature of impacting the nervous system, whether central (brain and spinal cord) or peripheral, resulting in symptoms spanning from paralysis to cognitive decline.

https://www.grandviewresearch.com/industry-analysis/neurology-devices-market#:~:text=The%20plan%20aims%20to%20reduce,Mental%20Health%20and%20Substance%20Use.

The heterogeneous etiology of neurological disorders adds complexity to market analysis. Genetic mutations drive conditions like Huntington’s disease, while protein aggregation underlies Parkinson’s and Alzheimer’s pathology. Infectious agents cause meningitis and encephalitis, neoplastic processes create brain tumors, and vascular insufficiency leads to stroke. This etiological diversity translates to varied therapeutic approaches and market segments, each with distinct growth trajectories and investment profiles.

Despite affecting billions globally, the neurological therapeutics market remains relatively modest at $90 billion (databridgemarketresearch), suggesting significant expansion potential. This apparent undervaluation reflects both the complexity of drug development for neurological conditions and the current treatment landscape dominated by symptomatic rather than disease-modifying therapies. The intricate nature of the nervous system creates substantial R&D challenges, extending development timelines and increasing costs while complicating regulatory approval processes.

Investment opportunities in neurology benefit from powerful demographic and economic trends. Aging populations worldwide drive increased prevalence of neurodegenerative conditions, while improving healthcare infrastructure in emerging markets expands the addressable patient base. The market’s internal diversification - spanning acute interventions to chronic management across multiple disease states - provides portfolio stability while maintaining growth potential. Companies successfully navigating the complex development landscape can establish strong market positions protected by high barriers to entry, creating sustainable competitive advantages in this expanding therapeutic area.

Conclusion

Non-communicable diseases represent the stable core of global healthcare, offering predictable long-term revenue while still leaving room for breakthrough-driven upside in areas like oncology and neurology. For investors, these markets provide resilience and steady growth, with demographic and environmental trends ensuring continued expansion. The next part of this series will turn to communicable diseases, where higher volatility creates both risks and opportunities.

The drama of LVMH’s hostile takeover bid for Gucci at the close of the last century remains vivid, yet Bernard Arnault’s acquisition ambitions never ceased. This time, his sights turned to Hermès, one of the crown jewels of the luxury world.

At the beginning of the new century, Arnault acquired an initial stake in Hermès through his subsidiaries in the secondary market. Under French financial market law, the Autorité des Marchés Financiers (AMF) requires disclosure for any holdings exceeding 5%. Arnault cleverly kept his stake at 4.92%, avoiding mandatory disclosure.

The Equity Swap Strategy

When the 2008 financial crisis struck, Hermès shares plunged. Arnault pounced on the opportunity, quietly orchestrating a series of equity swaps with three investment banks: an Irish subsidiary of Natixis (4.7%), Société Générale (4.5%), and Crédit Agricole (3%).

The mechanism worked as follows: for example, Arnault would exchange a fixed 5% return for the uncertain future returns on Hermès shares purchased by the banks. If future returns reached 8%, the banks would pay Arnault the additional 3%; conversely, if returns fell to -3%, Arnault would compensate the banks for both his promised 5% and the extra 3% loss.

These agreements included call option clauses that granted Arnault the right to purchase these shares at predetermined prices once Hermès reached certain price levels. This strategy avoided triggering the 5% disclosure threshold, keeping Hermès unaware of the impending hostile takeover whilst securing positions at favourable prices.

By October 2010, as global markets were basking in bull market euphoria, LVMH’s swap agreements with the banks had expired. Arnault directly purchased the banks’ holdings, emerging from the shadows to reveal a 17.12% stake. As always, he claimed this was merely a legitimate financial investment in Hermès, with no intention of launching a tender offer or seizing control.

At the time, French law had not incorporated equity swaps into the definition of direct shareholding under the monetary and financial code, creating a regulatory grey area. Arnault exploited this ambiguity in French financial market law. The three investment banks, fully aware that these equity swaps were aimed at control, chose to cooperate. The cash-settled swaps represented a major financial commitment, yet LVMH failed to disclose the associated risks in its reports, skirting the edge of legality.

The Hermès Family’s Counter-Strike

Hermès was shaken. Although the family controlled about 70% of the company, these shares were dispersed across more than 200 relatives, with none holding more than 5%. Arnault had thus become the single largest shareholder.

On 3rd December 2010, behind closed doors in Paris, 51 Hermès family members gathered for a secret meeting. The family unanimously agreed to consolidate 50.2% of shares into the unlisted H51 family holding company, signing a freeze agreement preventing sales for 20 years. Voting rights were unified under a “persons acting in concert” arrangement, legally constituting a family trust. In legal terms, this was a concerted action: shares were transferred to H51, which held legal ownership, whilst family members retained beneficial rights.

The manoeuvre achieved triple protection: asset isolation from external threats, concentrated voting power, and an impenetrable shield against ownership dilution. As insurance against internal defection, the family retained pre-emptive rights over any sales of unlocked shares by members.

Despite publicly denying takeover intentions, Arnault continued buying, reaching 22.02% by late 2011. The Hermès family responded by taking LVMH to the AMF, alleging insider trading, information concealment, market manipulation, and fraudulent acquisition. LVMH countered that its swap arrangements were legitimate financial leverage, arguing they represented implicit consideration for future acquisition rights and remained within legal boundaries.

Finally, in 2013, the AMF ruled there was no insider trading but criticised LVMH for violating transparency rules. They condemned LVMH’s use of equity swaps to conceal actual control relationships and its failure to fulfil disclosure obligations, violating French corporate law’s fair dealing principles and essentially harming other shareholders’ rights. LVMH was fined €8 million, ordered to reduce its 22.02% stake to 8%, and barred from acquiring additional Hermès shares for five years.

A Landmark Precedent

The case set an important precedent: for the first time, French financial regulators explicitly included actual control through equity swaps within disclosure obligations. This marked a shift from the narrow, formalistic rule of direct shareholding disclosure to a “substance over form” principle. The ruling was later incorporated into the EU Financial Market Transparency Directive, establishing universal standards for the disclosure of derivative-based equity positions.

However, during Arnault’s hostile takeover period from 2010 to 2013, Hermès’ share price soared dramatically. When the dust settled and Arnault finally liquidated his 22% stake, he walked away with a staggering €3.8 billion profit.

Arnault’s raid had paradoxically enhanced Hermès’ valuation, as his aggressive acquisition strategy inadvertently drew market attention to the brand’s intrinsic worth. The market had indeed tracked every twist and turn of his acquisition dance, betting on the outcome with ruthless precision. In attempting to devour Hermès, Arnault had instead fattened it for his own spectacular exit.

The takeover attempt also catalysed Hermès to restructure its internal governance, prompting the family to refine its dual-tier management structure and consolidate its ownership. This defensive reorganisation ultimately strengthened the company’s strategic positioning, with Hermès’ market capitalisation eventually surpassing even that of LVMH. Perhaps the family enterprise model, with its focused brand stewardship and generational continuity, may indeed prove more valuable than absorption into a diversified luxury conglomerate.

This article is written by

The global life sciences market is projected to exceed 10% annual growth through 2033 (BioSpace), representing one of the most secure long-term investment opportunities across any sector. Its universal customer base effectively encompasses every human being, given that all individuals are potential patients at some stage of life. This unparalleled breadth makes life sciences both resilient and indispensable. The following series of comprehensive analysis examines market dynamics across eight major disease categories to identify optimal investment strategies for this rapidly expanding sector. The first article will be an overview of the market split between communicable and non communicable diseases with the following articles offering a deeper dive into the different markets.

Executive Summary

Strategic investment requires careful evaluation of market segments, growth potential, and profit margins through the lens of underlying disease characteristics, patient demographics, and therapeutic landscapes. Non-communicable diseases demonstrate stable, long-term revenue potential through chronic patient populations requiring lifelong treatment, with cardiovascular diseases leading at 150B market valuation despite serving 600 million patients. In contrast, the diabetes market shows lots of opportunity being valued at 90B relative to its 800 million patient base. Oncology markets exhibit exceptional per-patient revenue ($320B market/53M patients) with “winner-takes-all” dynamics favoring breakthrough innovators. Communicable diseases introduce market volatility but offer substantial upside through pandemic preparedness platforms and antimicrobial resistance solutions. Analysis reveals critical market inefficiencies where disease burden substantially exceeds current valuations, particularly in diabetes and neurological conditions affecting nearly half the global population. Climate change and demographic shifts are fundamentally reshaping disease distribution and market opportunities. Optimal portfolio construction balances stable chronic disease exposure with selective growth positions in high-margin therapeutic areas and opportunistic infectious disease platforms, ensuring steady returns while maintaining capacity to capture breakthrough innovations in one of the world’s most resilient growth sectors.

Communicable vs. Non-Communicable Diseases

The fundamental distinction between communicable and non-communicable diseases lies in their transmission patterns and etiology. Communicable diseases spread between individuals through pathogenic agents such as viruses or bacteria, as exemplified by the COVID-19 pandemic. Non-communicable diseases cannot be transmitted person-to-person (though genetic predisposition may be inherited) and include conditions such as cancer, cardiovascular diseases, and mental health disorders.

Disease duration significantly impacts market dynamics. Communicable diseases typically resolve within days to weeks, while non-communicable conditions often persist throughout a patient’s lifetime. This temporal difference creates distinct market characteristics: communicable disease markets experience higher patient turnover and treatment volume, whereas non-communicable disease markets rely on stable, long-term patient populations with minimal medication switching. As illustrated in the figure above, the relationship between patient share and market share requires careful interpretation. When evaluating the total number of patients treated annually, it is essential to account not only for treatment duration but also for the proportion of patients who require active intervention versus those who do not need any or minimal treatment.

Key Disease Examples and Market Implications

Non-Communicable Diseases: Diabetes exemplifies the chronic disease market profile. As a lifelong condition with manageable symptoms under proper treatment, diabetic patients typically maintain consistent medication regimens throughout their lives. Patient loyalty to established treatments remains high unless breakthrough innovations or substantial cost advantages emerge. This pattern extends across the non-communicable disease spectrum, from mild conditions to severe illnesses like cancer.

Communicable Diseases: Influenza represents the typical acute disease pattern, causing moderate to severe symptoms for brief periods, often seasonally. Prevention through vaccination and rapid recovery characterize this market segment. While individual impact remains limited, communicable diseases pose unique risks through pandemic potential. As COVID-19 demonstrated, rapid disease spread can paralyze populations when too many individuals become simultaneously infected, creating cascading economic and social disruptions.

The distinct etiologies of these disease categories shape their market dynamics. Non-communicable diseases arise from genetic factors, lifestyle choices, and environmental exposures. Diet significantly influences disease risk, excessive glucose intake drives diabetes, while saturated fats contribute to cardiovascular complications through thrombus formation by raising low density lipoprotein and potential embolism (dislocation). These conditions develop independently of infection rates or interpersonal contact, representing internal physiological dysfunction.

Communicable diseases encompass viral, bacterial, parasitic, and fungal infections. Their market dynamics depend on pathogen transmissibility and population immunity. In our globalized world, airborne and droplet-transmitted diseases can rapidly achieve pandemic scale, creating volatile but potentially lucrative market opportunities.

Market Overview

The life sciences market can broadly be divided into non-communicable diseases, which provide stable long-term revenue through chronic treatment, and communicable diseases, which are more volatile but offer high-growth potential during outbreaks.

Cardiovascular disease affects 600 million people with 70 million new cases each year and generates $150 billion in market value, while diabetes, which is impacting 800 million patients and growing by 10 million annually, remains significantly undervalued at $90 billion. Cancer, with 50 million patients and 20 million new cases annually, commands $320 billion and offers the highest revenue per patient, while neurological conditions affect nearly half of the global population yet represent only a $90 billion market. Chronic respiratory diseases add another 350 million patients and 50 million new cases annually, with an $80 billion valuation. On the communicable side, viral infections affect nearly the entire global population each year and drive a $250 billion market, bacterial infections reach 300 million patients with 800 million new cases but just $57 billion in value, fungal infections strike 1.6 billion people annually, including 7 million invasive cases, for $17 billion, and parasitic infections affect 1.5 billion and add 3.5 billion new cases yearly, generating $18 billion. For a breakdown of the different categories explaining the numbers and growth potential as well as possible complications or market driving factors refer to the following articles.

Conclusion

The global healthcare market represents a unique blend of stability and volatility, with opportunities spanning chronic disease management, breakthrough oncology therapies and pandemic preparedness. The sector’s growth is not only resilient but also structurally underpinned by demographic and environmental megatrends.

This article serves as the starting point of a three-part series. The next two pieces will take readers deeper into the data, first exploring the market characteristics of non-communicable diseases, then examining the dynamic but volatile world of communicable diseases. Together, the series will reveal useful insights into the different healthcare market segments for investors to understand the background behind different diseases from understanding the science to what affects the market.

This article is written by Menno Westenbrink, founder of Future Funds

Everyone uses money, you pay your rent with it, buy food with it and go on holidays with it. But have you ever wondered about the forces that determine money's value and availability? Behind every transaction lies a complex system of institutions and mechanisms working to manage economic stability. We used to live in a system where money was tied to gold and silver, creating automatic but rigid constraints. Today, powerful institutions actively shape monetary conditions through sophisticated tools most people never see. In today's article we are going to explore these forces behind money and how they evolved to understand the invisible architecture that influences your economic life every day.

The Evolution: From Gold Standard to Modern Banking

For most of human history, money was backed by something tangible, usually gold or silver. This system, known as the gold standard, seemed simple and trustworthy. If you had a dollar, you could theoretically exchange it for a specific amount of gold. This backing gave people confidence in their currency and provided automatic limits on how much money governments could create.

The Gold Standard Era

Under the gold standard, money supply was directly tied to gold reserves. If a country wanted to print more money, it needed more gold. This created a natural brake on inflation and government spending, as there was only so much gold available. Countries couldn't simply create money out of thin air to solve their economic problems.

However, this apparent stability came with a severe cost: inflexibility. When economic crises hit, governments had limited tools to respond. They couldn't increase money supply to stimulate the economy or lower interest rates beyond what gold reserves allowed. The economy was essentially on autopilot, with no one at the controls during turbulence.

The Great Transition

The journey away from gold was long and complex, beginning when most countries suspended gold convertibility during World War I to finance their war efforts. Though initially intended as temporary wartime measures, many nations found it difficult to return to the strict discipline of gold after experiencing the flexibility of managing their own money supply. The Great Depression of the 1930s provided another severe test, with countries like Britain in 1931 and the United States in 1933 abandoning gold to combat deflation. However, the final break didn't come until 1971, when President Nixon ended the last remaining link between the dollar and gold, primarily due to the mounting costs of the Vietnam War and unsustainable international monetary pressures rather than domestic depression concerns.

The 1929 Crisis: When Rigid Systems Failed

While the gold standard didn't finally end until 1971, the Great Depression provides a compelling example of why fixed monetary systems proved inadequate for modern economic management. When the stock market crashed in October 1929, the economic downturn that followed wasn't just severe, it was prolonged and devastating precisely because the monetary system couldn't adapt.

The Deflationary Spiral

Under the gold standard, as the economy contracted, so did the money supply. Banks failed, taking depositors' money with them. People lost confidence and hoarded what money they had, further reducing the money in circulation. Prices fell, but this wasn't good news because it meant businesses couldn't earn enough to pay their debts, leading to more bankruptcies and unemployment.

The central banks of the time were powerless to stop this deflationary spiral. They couldn't inject money into the system because they were constrained by their gold reserves. They couldn't lower interest rates significantly because that might trigger gold outflows as investors sought better returns elsewhere.

The Human Cost

The economic data paint a stark picture of the severity. According to FRED (Federal Reserve Economic Data) and Federal Reserve Bank of St. Louis analysis, unemployment rose from about 4% in 1929 to a peak of 25% in 1933, while real GDP shrank 29% from 1929 to 1933. But behind these numbers were real people, families losing their homes, workers standing in bread lines, young people with no prospect of employment.

Crucially, countries that abandoned the gold standard earliest recovered fastest. Britain left gold in 1931 and began recovering by 1932. The United States effectively left gold in 1933 with Roosevelt's policies and saw improvement shortly after. France, which clung to gold until 1936, suffered the longest and most severe depression.

Meet the Key Players: Central Banks as Economic Architects

Today's financial system operates very differently, and at its heart sits the central bank, an institution that would seem almost magical to someone from the 1920s.

The Central Bank's Mission

Modern central banks have varying goals but generally focus on two key objectives: maintaining price stability and supporting economic growth and employment. The Federal Reserve (Fed) explicitly operates under a dual mandate of price stability and full employment. The European Central Bank (ECB) has price stability as its primary mandate, though it also considers broader economic conditions. The Bank of England similarly focuses primarily on price stability while supporting government economic policy. Unlike the automatic constraints of the gold standard, these institutions have both the power and responsibility to actively manage the economy through discretionary policy decisions. These institutions maintain credibility through independence from political interference and transparent communication about their objectives and methods.

Think of central banks as the economic equivalent of a skilled driver. Where the gold standard was like being in a car with no steering wheel, you might travel in a straight line, but you couldn't avoid obstacles. Central banks can steer, brake, and accelerate as conditions require.

The Modern Toolkit: How Central Banks Shape Your World

Central banks wield two primary tools that directly affect your daily life: interest rates and asset purchasing programs. While most people have heard of interest rates, asset purchasing programs remain mysterious to many, yet they're arguably more important in today's economy.

Interest Rates: The Economy's Steering Wheel

When a central bank changes interest rates, it's like adjusting the economy's accelerator or brake pedal. Lower rates encourage borrowing and spending, as businesses invest, people buy homes, and economic activity increases. Higher rates do the opposite, cooling down an overheated economy.

This affects you directly. If you have a mortgage, car loan, or credit card debt, interest rate changes impact your monthly payments. If you're saving money, rate changes affect how much interest you earn. If you're looking for a job, rate changes can influence whether companies are hiring or laying off workers due to the company's financial situation.

Asset Purchasing Programs

Here's where things get really interesting, and where our modern system shows its true power compared to the rigid gold standard era. Asset purchasing programs, also known as Quantitative Easing (QE), represent one of the most important financial innovations of the past decades. In simple terms, asset purchasing programs allow central banks to buy large quantities of bonds and other securities. When the central bank buys these assets, it pays for them by creating new money electronically, essentially creating money out of thin air.

This might sound dangerous or even fraudulent, but it's actually a sophisticated tool that serves several crucial purposes:

Injecting Money When Banks Won't Lend

When lending conditions tighten, banks become cautious and stop lending to businesses and individuals, even if interest rates are low. Asset purchasing programs solve this by buying bonds directly from banks, giving them cash that they can then lend out. The central bank essentially forces money into the financial system when normal channels are blocked.

Lowering Long-Term Interest Rates

This is like having a thermostat that only controls the temperature in one room of your house (short term interest rate). Asset purchasing programs give central banks a way to influence the temperature in other rooms too (longer term interest rates). Research demonstrates these effects are substantial: the European Central Bank's Asset Purchase Programme has been shown to reduce 10-year government bond yields by approximately 95 basis points, nearly a full percentage point (Eser et al., 2019).

Figure 1 illustrates this effect clearly: while short-term rates remain relatively unchanged, longer-term rates experience significant declines when central banks implement asset purchasing programs.

Figure 1

Providing Market Confidence

Perhaps most importantly, these programs signal to markets that the central bank will do whatever it takes to prevent economic collapse by providing liquidity. This confidence itself helps prevent the kind of panic that made the Great Depression so severe.

Why Today's System Works Differently

The 2008 financial crisis and COVID-19 pandemic demonstrated how fundamentally different our monetary system has become. Rather than being constrained by automatic mechanisms like gold reserves, central banks could deploy sophisticated tools quickly and effectively. These weren't just emergency measures, but applications of the same systematic approach central banks use regularly to manage economic conditions.

This represents the core difference between past and present: where the gold standard operated automatically regardless of economic circumstances, today's system operates through deliberate institutional judgment guided by economic expertise.

How This Affects Your Daily Life

Understanding these tools helps explain many aspects of your economic experience that might otherwise seem mysterious.

The Homeowner/Renter

When central banks implement asset purchasing programs, one of their primary targets is mortgage bonds. By buying these securities, they directly lower mortgage rates, making homes more affordable for buyers. This increased demand typically drives up home prices, benefiting existing homeowners but potentially making it harder for first-time buyers.

If you're a renter, these programs have mixed effects. Lower mortgage rates might encourage your landlord to refinance and potentially lower rent, but higher property values might also lead to higher rents in the long term.

The Worker

Asset purchasing programs aim to prevent the kind of mass unemployment that characterized the Great Depression. By keeping credit flowing and preventing business failures, these programs help maintain employment during economic downturns.

However, the effects aren't uniform. Industries that rely heavily on borrowing (like construction and manufacturing) tend to benefit more than service industries. Geographic effects also vary, with areas that have more financial activity often seeing greater benefits.

The Saver/Investor

For savers, asset purchasing programs create a complex situation. These programs are explicitly designed to lower interest rates, which reduces returns on savings accounts. This is intentional, central banks want to encourage people to spend or invest rather than save when stimulus is needed.

For investors, asset purchasing programs often drive up prices of stocks, bonds, and real estate. The central bank's bond purchases create a "portfolio effect", when investors can't earn attractive returns on bonds, they move money into riskier assets, driving up those prices.

The Debtor

If you have variable-rate debt, asset purchasing programs generally work in your favor by reducing interest rates. Even fixed-rate borrowers benefit if they can refinance at lower rates.

However, if these programs are too successful at preventing deflation, the resulting inflation can erode the real value of wages, making it harder to service debt even if nominal interest rates are low.

The modern financial system, with central banks using tools like asset purchasing programs, has fundamentally changed how our economic systems work. Instead of the deflationary spirals and bank runs that characterized the 1929 crash and other stress periods, we now have institutions with both the power and knowledge to intervene decisively when markets themselves can’t act.

This doesn't mean our system is perfect, it creates its own challenges and trade-offs. But it does mean we live in a world where economic catastrophes like the Great Depression are far less likely, and where the "forces behind money" are guided by human judgment rather than the rigid constraints of gold reserves.

Understanding these forces helps explain not just how money works today, but why your economic experience is shaped by decisions made in central bank boardrooms around the world. The next time you hear about interest rate changes or asset purchasing programs, you'll know you're witnessing the modern equivalent of economic engineering.

References

Eser, F., Lemke, W., Nyholm, K., Radde, S., & Vladu, A. L. (2019, July 8). Tracing the impact of the ECB's asset purchase programme on the yield curve.

The customers of ASML

Who are the customers of ASML? There is no public documentation on the exact specification on all ASML's clients. TSMC, Intel Corp, and Samsung Electronics are the biggest clients of ASML Fitch Ratings (2024), and SMIC and SK Hynix have been found to be significant customers of ASML Reuters (2025).

TSMC (Taiwan Semiconductor Manufacturing Company) is one of the largest semiconductors and circuits producers in the world, it focuses solely on producing for large customers. SMIC (Semiconductor Manufacturing International Corporation) is a chinese based semiconductor foundry whose revenue comes from producing chips. SK Hynix Inc is a South Korean based chip producer with a main focus on memory chips. Intel is a semiconductor manufacturing corporation that designs, manufactures, and sells mainly computer components. Samsung electronics is a broader organization that develops, produces and sells many different products in the technology sector.

There is some difference in the type of clients ASML has, some are only focusing on producing chips on a large scale, while other organisations are big tech companies. Below I included a correlation bar chart based on the daily returns of the assets above with ASML returns since 2021. Most assets are positively correlated, this result alone does not tell us a lot. TSM has the highest overall correlation. This is not suprising since it is one of the largest clients of ASML, and their whole business model is producing chips. But do returns of TSM have any predictive value for the returns of ASML?

Figure 1

Can TSM and other clients returns predict ASML prices?

Short term

For the short term test, I downloaded daily return data of ASML and its customers from the first of april 2021 up until now. I use a Granger Causality test that tests if the past returns of TSM hold predictive power over the return of ASML. A Granger causality test assumes stationary time series. I used a Dickey-Fuller test to check if the returns are a stationary time series. The Granger Causality is as follows.

yt = α0 + Σᵖᵢ₌₁ αᵢyt₋ᵢ + Σᵖᵢ₌₁ βᵢxt₋ᵢ + εt. (1)

Where yᵢ is the current observed return of ASML and the first sum contains all the lagged previous returns of ASML with coefficients αᵢ. The second sum contains previous observed returns of TSM with coefficients βᵢ. The test determines if the coefficients of TSM are significantly different from zero or not, i.e., if they have any predictive value.

According to the Granger causality test the returns of TMS hold significant predictive value over the return of ASML on a given day. The specific strength of predictability is not determined with this model. We only know the historic returns of TMS are an indicator of the ASML returns. In the graph below we see the lags (amount of previous TMS returns used to estimate the ASML return on a day). The y-axis contains the p-value for different amount of historical returns of TSM as predictors. The y-axis contains the p-value, a lower p-value means we are more certain there is an effect. The best predictor is using the previous day's return of TMS. That is in a sense intuitive, older returns carry less information compared to yesterday's return. Since the model is based on ASML's past returns and TSM past returns, it indicates the added value of using TSM returns besides only using ASML returns. Important to note is that we have not found a direction of the effect, the direction of the effect can be different for various lags.

Figure 2

Apparently the biggest client of ASML which is focused purely on chip producing has predictive value over the future returns of the ASML stock, will this hold for other clients of ASML?

I performed the same test with the other clients of ASML to check for similarities. For SK Hynix there was no predictive power according to the test and given the data. For the other clients SMIC, Intel, and Samsung there was significant predictive power. For intel we have similar results to TSM, historical prices have predictive prices for the returns of ASML, and yesterday's price is the best predictor. For Samsung Electronics only the past two days of returns have a significant predictive effect. And suprisingly, for SMIC only higher lags have significant predictive power. Namely the return of the past 30 days seem to have a significant effect on the return of ASML.

Figure 3

Figure 4

Figure 5

Long term

Testing for more long-term effect is very interesting, in a longer time frame there can be trends. And if a client is doing well, there is likely a significant lag until ASML potentially delivers new advanced machinery for the client. For illustration, ASML's most advanced machine the 'EUV' can take one and half year to produce.

To test for long-term effects, I use the same test but download monthly return data instead of daily return data. I downloaded the monthly returns starting from January 2015 up until now. The total observations will be less than the test before. However, we still have enough to perform the Granger Causality test. Below are the results for each client, SK Hynix, SMIC, and Samsung's return do not hold any significant predictive power over ASML according to the test. So we will leave those graphs out.

Intel returns seem to have significant predictive capabilities with 2-3 months of lagged returns, so still not quite long term effects. However, for TSM we see many lags of historical prices that seem to have significant predictive power. The most significant seems to be 6 months of lag, so the returns of the past 6 months of TSM are the best indicator of ASML's share return according to the Granger test. Most lagged monthly retuns of TSM are significant predictors of ASML's return according to the Granger test. It seems TSM is a strongly intertwined client of ASML whose share price are more connected compared to the other clients.

Figure 6

Figure 7

The tests performed above are all based on the share price. This price is ofcourse subject to many factors like speculation. As a quick check I performed the same test to the year-on-year quarterly revenue growth rates, and the quarter-on-quarter revenue growth rates of tsm and asml. I chose to use them both to see if the revenue growth rate has some predictive effect or exist due to seasonality. I found that using the past four quarterly growth rates of tsm has significant preditive value over the next quarterly growth rate of asml. Investigating the coefficients of the lagged model we can see the fourth lag of the quarterly growth has the highest coefficient. This is true for both versions of the quarterly revenue. This is in line with our thoughts: It takes a while for increasing demand in chips to reach ASML. The model finds that an increase in quarterly year-on-year revenue will influence ASML the most exactly one year in the future. On shorter term there is also effect, but the effect is strongest one year into the future.

Figure 8

Conclusion

We have seen results of predictive power in the short- and long-term. These results do not tell us how to perfectly predict ASML returns based on clients historical stock returns. But it provides evidence of a predictive relationship, which can be used as an indicator for example.

We have seen that the larger clients TSM and Intel have significant predictive power. TSM which focuses only on producing chips not only has the highest correlation with ASML, but also the most significant predictive power short- and long-term. Apparently, there is some significant lag between the organizations, which is quite intuitive given their business relationship. We see the same pattern in the revenue relationship between these assets. An year-on-year quarterly increase of revenue from TSM has a positive effect in general on the quarterly revenue of ASML, mainly exactly one year later. It is likely TSM will have an increase in revenue this year and the coming years due to the high demand for chips. Will the revenue and stock price of ASML follow? The analysis says it is likely to follow

Appendix

References

Fitch Ratings (2024, May). Fitch upgrades asml to "a+"; outlook stable. https://www.fitchratings.com/research/corporate-finance/fitch-upgrades-asml-to-a-outlook-stable-17-05-2024. Accessed: 2025-08-21.

Reuters (2025, March). Asml's annual report says export curb worries hit customer spending in 2024. Reuters. Accessed: 2025-08-21.

Figure 1, IQVIA weekly NPA total prescriptions, weekly data June 27, 2025; Incretin analogs market includes injectable GLP-1s, oral GLP-1s, and GLP-1/GIP dual agonists.

This article is written by

GLP-1 receptor agonists such as Ozempic have emerged as one of the most disruptive innovations in healthcare, with the potential to reshape both diabetes and obesity treatment markets. This article examines the scientific foundation of these therapies, the competitive positioning of Novo Nordisk and Eli Lilly, and the investment implications arising from their diverging strategies in this far from saturated market.

The Growing Demand for GLP-1 Therapies

Global demand for GLP-1 drugs is accelerating rapidly. According to the World Health Organization, more than 830 million people worldwide live with diabetes, roughly 90% of whom have type 2 diabetes, which can be treated with GLP-1 drugs. Beyond diabetes, these therapies are increasingly prescribed for weight management, opening a significantly larger addressable market.

The market is effectively a duopoly. Novo Nordisk and Eli Lilly dominate, though their strategies and corporate structures differ markedly. Understanding these distinctions is essential for evaluating long-term investment opportunities.

Current Marketed Therapies

Novo Nordisk’s portfolio is anchored by semaglutide (Ozempic, Wegovy, Rybelsus), while Eli Lilly’s leading product is tirzepatide (Mounjaro, Zepbound). Structurally, tirzepatide is more complex, with 225 carbon atoms compared to semaglutide’s 187 (Figure 2). Complexity influences manufacturing scalability, and while the FDA has noted compounding challenges for semaglutide, tirzepatide’s dual mechanism introduces its own production risks.

Figure 2, https://www.biochempeg.com/article/355.html

Mechanistically, semaglutide mimics endogenous GLP-1, whereas tirzepatide functions as a dual agonist, targeting both GLP-1 and GIP receptors. Clinical trial data show superior weight-loss efficacy for tirzepatide, particularly in higher weight-loss categories (>15%), where it outperforms semaglutide by approximately 20% (Figure 3).

Figure 3, Rodriguez PJ, Goodwin Cartwright BM, Gratzl S, Brar R, Baker C, Gluckman TJ, Stucky NL. Semaglutide vs Tirzepatide for Weight Loss in Adults With Overweight or Obesity. JAMA Intern Med. 2024 Sep 1;184(9):1056-1064. doi: 10.1001/jamainternmed.2024.2525. PMID: 38976257; PMCID: PMC11231910.

Pipeline Strategies

The two companies’ R&D approaches diverge significantly. Eli Lilly maintains a diversified pipeline across multiple therapeutic areas, reducing concentration risk and enabling cross-segment growth. Novo Nordisk is far more specialized, with roughly half of its pipeline dedicated to diabetes and obesity. Specialization offers the potential for market dominance and pricing power but reduces diversification.

Figure 4

In marketed products, ELi Lilly offers approximately 52 therapies compared with Novo Nordisk’s 27. However, Eli Lilly derives nearly two-thirds of its revenue from a relatively small number of blockbuster drugs, increasing dependence on their success. Novo Nordisk’s sales are more evenly distributed across GLP-1s, other diabetes therapies, and rare-disease treatments. In 2024, Novo generated roughly half its revenue from GLP-1s, while Ely Lilly earned $10.4 billion of its $15.6 billion total revenue from key products, illustrating that despite its broad pipeline, Eli Lilly is becoming more concentrated in its commercial performance of these key drugs. These Key drugs mainly consist of GLP-1 drugs that only account for around a third of their R&D pipeline, meaning a deviation from internal focus to revenue volume.

Market Dynamics and Economics

Prescription data from the Peterson-KFF Health System Tracker indicate Eli Lilly currently leads Novo Nordisk in U.S. GLP-1 initiations. Yet overall penetration remains low: only ~6% of Americans under 65 are using GLP-1 drugs, despite ~40% being eligible through private insurance. This implies a potential pool of ~40 million additional U.S. patients as illustrated in Figure 5.

Figure 5

At current list prices (~$1,000/month) and an average treatment duration of nine months, per-patient revenue approximates $9,000. Estimated production costs remain minimal, around $5 per month, according to a JAMA Network Open study on Ozempic. The resulting gross margin profile is extraordinary. Applied to the eligible U.S. population, the theoretical market exceeds $350 billion in eligible first time users, constrained primarily by manufacturing capacity. Globally, the GLP-1 market is projected to double by 2030, with North America retaining ~78% of market share as of 2024 (Grand View Research in Figure 6.). As GLP-1 drugs are also utilized for diabetes patients that have to continuously inject the drug the market of the over 700 million type 2 Diabetes patients globally will maintain even after the weight loss hype is over.

Figure 6, https://www.grandviewresearch.com/industry-analysis/glp-1-receptor-agonist-market

Both companies are expanding manufacturing aggressively while pursuing oral formulations that could broaden the patient pool by reducing reliance on injections. Potential disruption may emerge from companies like Roche, which is developing an oral GLP-1/GIP dual agonist and could enter the market in the post-hype cycle with a lower-cost alternative.

Governance and Strategic Outlook

Corporate leadership further differentiates the two players. Novo Nordisk recently appointed Maziar Mike Doustdar as CEO. With three decades at the company and deep knowledge of its foundation-controlled ownership structure, Doustdar is well positioned to drive scaling efficiency while reinforcing Novo’s long-term strategic stability. The foundation structure, though complex, shields Novo from short-term market pressures and supports a multi-decade investment horizon.

At Eli Lilly, David A. Ricks has served as CEO since 2017, emphasizing the expansion of GLP-1 production capacity and a research-driven pipeline. Eli Lilly’s strategy reflects a focus on leveraging the massive opportunity in diabetes and obesity while maintaining breadth in other therapeutic areas.

Valuation and Investor Perception

As the GLP-1 market transitions from a valuation expansion phase into a more earnings-driven cycle, investor expectations for both firms have recalibrated. Eli Lilly’s forward P/E multiple declined from ~110 in June 2024 to ~51 by mid-2025, reflecting a move from hyper-growth pricing toward sustainable expansion. Novo Nordisk’s multiple fell from ~47 to ~19 in the same period, signaling a transition from a growth story to a cash-generating incumbent.

Despite Lilly’s longstanding positioning as a diversified pharmaceutical company, its valuation is increasingly tied to GLP-1 revenues, suggesting a strategic pivot toward concentration in this therapeutic class. Novo Nordisk, by contrast, continues to pursue specialization, aligning investor expectations with margin strength and market leadership in diabetes and obesity.

Over the long term, both firms must balance growth ambitions with earnings delivery. Lilly faces the challenge of maintaining credibility as a diversified innovator while relying more heavily on GLP-1s, whereas Novo Nordisk will need to sustain dominance in a narrower therapeutic focus. Their ability to align strategy with investor expectations will shape the durability of this global duopoly.

Investor Outlook

In the near term, Eli Lilly appears better positioned due to greater production capacity and the superior clinical efficacy of tirzepatide. Novo Nordisk, however, is likely to sustain profitability through specialization and its ability to capitalize on an aging European patient base. The competitive field may broaden as late entrants like Roche pursue lower-cost, more convenient formulations, but for now, the GLP-1 market remains firmly in the hands of Novo Nordisk and Eli Lilly.

The next two to three years will be decisive in determining whether these companies can capture near-term demand while preparing for the second wave of GLP-1 innovation. For investors, tracking R&D progress, manufacturing expansion, and regulatory milestones will be critical to assessing the evolving risk-reward profile in what remains one of the most significant therapeutic markets globally.

In the recently concluded earnings season, the entire luxury industry appeared to be entering a period of deep recession, yet Hermès' performance remained remarkably resilient. But as a standout performer in the sector, are its financial fundamentals truly as robust as they seem?

Hermès International SCA shares fell 3% following the release of its financial report on July 30, despite the brand posting accelerated Q2 growth, from 7.2% to 9.0%. However, beneath this headline improvement lies a concerning trend: whilst leather goods accelerated, other divisions deteriorated.

Figure 1

The financial results reveal that the revenue share of leather goods continues to climb. This suggests Hermès may be indirectly managing rare handbag sales through its quota system, potentially implementing covert price reductions by lowering quota requirements for handbags.

The Hidden Dynamics of Leather Goods Revenue

Hermès refuses to acknowledge its quota marketing strategy, yet this has become an unspoken understanding between consumers and sales associates. To purchase iconic bags like the Birkin or Kelly, customers must first buy additional items such as ready-to-wear, watches, furniture or tableware. If a Birkin costs €10,000, customers previously needed to purchase €20,000 worth of quota items to receive an offer from their sales associate; now they need only €10,000 worth. Isn't this effectively a disguised price reduction amid economic pressures?

Whilst it's difficult to establish a direct link between this season's non-leather goods sales and the quota system, the revenue ratio between leather goods and other categories serves as a telling indicator. This metric offers a more intuitive reflection of Hermès' actual sales performance, beyond its polished financial figures.

Figure 2

Therefore, we construct the Leather Goods Revenue Ratio as the share of trailing twelve months (TTM) revenue from the Leather Goods and Saddlery segment relative to the combined TTM revenue of all other business segments. This ratio is computed quarterly, in alignment with Hermès’ financial reporting schedule. The TTM approach, rather than single-quarter values, helps mitigate the seasonal volatility in luxury goods consumption, providing a more stable and representative indicator of changes in the company’s revenue composition over time.

For stock market performance, we employ the natural logarithm of Hermès’ closing stock price (Log Stock Price) on the date of each quarterly financial disclosure. This variable captures the market’s immediate valuation following the disclosure of new financial information. The logarithmic transformation normalises stock price variation, reduces skewness and facilitates statistical analysis.

As shown in Figure 2, the Log Stock Price (red line) demonstrates a persistent upward trajectory, indicating sustained growth in Hermès’ market valuation. Meanwhile, the Leather Goods Revenue Ratio (blue line) rose steadily from 2015, stabilising around 1.0 before declining sharply between 2021 and 2023, coinciding with the post-pandemic spending boom. From 2024 onwards, the ratio has begun to regain ground, signalling a shift back toward greater dependence on leather goods.

How Leather Dependence Shapes Stock Returns

Building on these baseline variables, we construct two derived measures for econometric analysis. The independent variable ∆LGRRt​ captures the proportional change in the Leather Goods Revenue Ratio between consecutive reporting periods. An increase in this metric indicates higher dependence on leather goods revenue. The dependent variable, ∆ log Pt , represents the change in Log Stock Price, defined as the difference between the natural logarithm of the closing price on the current release date and its value on the previous release date.

∆ log Pt =Beta_0+Beta_1 ∆LGRR_t+Epsilon_t

Our hypothesis posits that an increase in ∆LGRRt, indicating greater reliance on leather goods revenue, is associated with a decrease in ∆ log Pt , reflecting negative stock performance. This relationship would be consistent with potential market concerns regarding revenue concentration risk.

Figure 3

The initial regression yielded a slope coefficient (β₁) of -0.3408, consistent with our hypothesised negative relationship. However, the estimate was not statistically significant (p = 0.3554), limiting the robustness of this result. The scatter plot suggests a weak negative association, with few observations falling in the lower-left quadrant. Overall, the model’s explanatory power was low, partly due to the influence of a single outlier.

Spotting Market Inefficiencies

Figure 4

This outlier, marked on the graph, actually represents a potential buying opportunity. The leather goods revenue ratio declines, signalling a more balanced and resilient revenue mix, yet the stock price falls. This suggests that the market has failed to recognise this positive development and temporarily undervalued the shares.

Figure 5

After excluding this market mispricing, the p-value improved to 0.195. Whilst the result still does not reach statistical significance, the relationship becomes more justifiable. Nevertheless, the negative association between leather goods revenue dependence and stock performance persisted, providing continued support for our theoretical framework.

Overall, the analysis indicates a meaningful, though not statistically definitive, inverse relationship between the leather goods revenue ratio and Hermès’ stock performance. This outlier illustrates how market inefficiencies can create compelling investment opportunities when actual performance diverges from market expectations.

Practical Takeaway for Your Portfolio

Think of Hermès like the world's most exclusive nightclub – the harder it is to get in, the more everyone wants to be there. But here's what most investors miss: even the most exclusive clubs have their quiet nights.

Our research reveals something Wall Street analysts rarely discuss in their glossy reports. When Hermès becomes too dependent on selling those coveted Birkin and Kelly bags, it's actually a warning sign disguised as success.

The next time you see Hermès' quarterly results, don't just look at the headline growth numbers that financial media celebrates. Instead, dig into the ratio we've identified. When leather goods revenue starts dominating the mix, it's time to pause and ask: "Is this sustainable growth, or is Hermès painting itself into a corner?"

More importantly, watch for those moments when the market gets it wrong: when leather goods dependence decreases while the stock price drops anyway, these are potential goldmine moments for savvy investors who can see the bigger picture while others panic.

This article has been written by

What if you could stop working years, or even decades, before most people retire? That’s the promise behind the FIRE movement (Financial Independence, Retire Early), a philosophy that challenges the traditional 9-to-5 grind. Instead of waiting until your mid-60s to finally enjoy freedom, FIRE encourages disciplined saving, smart investing, and intentional lifestyle choices so you can live life on your own terms. And the best part? There’s no single way to do it. Whether you’re drawn to a minimalist approach, a luxurious retirement, or a balance somewhere in between, FIRE offers different paths to financial independence tailored to your goals.

Not everyone chasing FIRE envisions the same future. In fact, the movement has splintered into different paths, each reflecting how people imagine their ideal lives once work is optional.

Take Lean FIRE, for example. Picture someone who thrives on minimalism, living in a small, cozy home, biking instead of driving, and finding joy in experiences that cost little. By saving more than half of their income and keeping expenses lean, they fast-track their independence and gain freedom much earlier than most.

On the opposite end is Fat FIRE. This is the dream for those who don’t want to cut back on life’s luxuries, imagine retiring while still enjoying fine dining, frequent travel, or even living in a high-cost city without compromise. It requires building a much bigger financial cushion, but the reward is comfort without sacrifice.

Then there’s Barista FIRE, which appeals to people who don’t necessarily want to stop working altogether but crave flexibility. Think of someone who saves enough to cover most of their lifestyle from investments, then picks up part-time or passion projects, teaching yoga, running a small business, or yes, even working at a coffee shop, without relying on the paycheck for survival.

Finally, Coast FIRE is like planting seeds early and letting them grow. Imagine someone who invests aggressively in their 20s and 30s, then lets compound interest do the heavy lifting while they “coast” through their career. They still work to cover everyday expenses, but they don’t need to stress about saving for retirement anymore, the future is already taken care of.

The Realities and Trade-Offs of the FIRE Movement

Although the FIRE movement has captured media attention in recent years, the reality is that only a very small fraction of people actually reach early retirement. Data from Motley Fool (2016–2022) shows just 1% of Americans retire between ages 40 and 44, 2% between 45 and 49, and only 6% by their early 50s, a clear indication of how rare FIRE truly is.

The biggest barrier is income: even with strict budgeting, most people need a six-figure salary to accumulate enough wealth to exit the workforce in their 30s or 40s. That said, wealth-building is not exclusive to high earners; in fact, one-third of U.S. millionaires never earned six figures in a single year, proving that disciplined saving and investing can still create financial independence over time.

The benefits of FIRE are appealing, greater freedom, flexibility, and the possibility of pursuing passion-driven work, but the downsides are significant. Extreme frugality can reduce quality of life, and retiring without a clear sense of purpose may lead to mental health challenges. There are also financial trade-offs: stepping away too early means losing out on prime earning years, employer-sponsored benefits like health insurance and retirement contributions, and potentially lowering future Social Security payouts.

At a broader level, some economists even question the impact widespread early retirement could have on workforce participation and consumer spending. In short, FIRE offers the dream of independence, but it comes with steep requirements and trade-offs that only a minority of people achieve.

Case Study: Corporate Salary Progression in a U.S. City

To understand how FIRE works in practice, let’s look at a simplified example of a corporate employee in a major U.S. city. Imagine starting with a salary of $60,000 and receiving a modest 2% annual raise. By adopting a very high 70% savings rate, this individual commits to living on just 30% of their income while saving the rest. In the first year, that translates to $42,000 saved and $18,000 spent.

These savings are invested with an assumed 8% annual return, which is in line with long-term stock market averages. Thanks to the power of compounding, the portfolio grows rapidly. After just four years, the account balance surpasses $150,000; by year seven, it’s more than $348,000; and after 12 years, the total investment portfolio has climbed to $822,728. During that same period, salary increased to just over $74,600, meaning annual savings also rose to more than $52,000.

What’s striking here is how quickly wealth accelerates. By the 12-year mark, the portfolio alone could sustainably generate around $32,900 a year using the 4% withdrawal rule, or even produce nearly $66,000 annually if markets delivered average returns. In other words, this person could cover a lean lifestyle entirely from investments or significantly supplement part-time income, making options like Barista FIRE or Coast FIRE very realistic.

This case highlights a core principle of FIRE: it isn’t just about earning a high salary, but about how much you save and invest consistently. Even with a relatively average starting income, discipline and compounding can dramatically shift financial possibilities within a decade.

The FIRE movement is as much a mindset as it is a financial strategy. Inspired by Vicki Robin’s philosophy in "Your Money or Your Life”, it encourages individuals to examine what “enough” truly means for their own lives, rather than blindly chasing extreme wealth. While the 25-times-expenses rule and 4% withdrawal rate provide a practical roadmap for early retirement, the core of FIRE lies in using frugality not as a limitation, but as a tool to reclaim time, focus on purpose, and pursue what genuinely matters. As Robin puts it, financial independence is less about hoarding money and more about liberating one’s life, buying freedom one thoughtful choice at a time.

In the end, I believe the FIRE movement sparks a valuable conversation about how we view work and money. While the idea of retiring in your 40s is appealing, the reality is that it often requires very high earnings and sacrifices that not everyone can make. Personally, I think working every day has value, not just financially, but in giving structure, purpose, and the chance to keep improving ourselves. FIRE may not be realistic for the majority, but its core lesson is useful: to be intentional with our time and money so that work becomes a choice, not just a necessity.

Sources

https://www.nerdwallet.com/article/investing/financial-independence-retire-early

https://www.investopedia.com/terms/f/financial-independence-retire-early-fire.asp

https://www.cbsnews.com/news/how-the-fire-movement-is-inspiring-early-retirees/

https://www.ramseysolutions.com/retirement/what-is-the-fire-movement

https://blog.thecalculatorking.com/what-is-the-fire-movement/

Most assets seem to be correlated to each other. So how does the correlation look between Europe's and United States's financial markets? In this article, we look at the correlation between the S&P500 and the STOXX600 as proxies for those financial markets. Do deviations from the normal correlations go accompanied with certain market movements?

ETF prices of S&P500 and STOXX600

I downloaded the historical adjusted closing prices of the ETFs S&P500 (SPX) and STOXX600 (STOXX) representing the United States and European financial markets, respectively, on Yahoo Finance. These ETFs are chosen as proxies for the financial markets in those regions since these are well-known, highly liquid ETFs, which give an accurate depiction of the market. Below you can see a plot of the ETF prices over time. The time series look correlated, as do many assets. We will dive a bit deeper into the correlation of these markets and inspect how this relates to the returns of these assets.

Figure 1

Correlation

I calculated the rolling correlation of the returns of the STOXX600 and S&P500 using a window of 30 days. So, on each given day we have a correlation based on the past 30 days of returns. Included in the plot is a rolling percentile band containing the last 252 observed correlations (amount of trading days in a year typically). This is chosen since we do not know the underlying distribution of the rolling correlation, so I do not want to make any assumptions about that. Besides that, fundamental changes over longer time periods could change the correlation between these markets, therefore we do not make a confidence bands based on the whole time series. You can see that the correlation mostly stays within these bounds, but sometimes breaks out of it. But why is this interesting at all? Is the correlation actually a long-term stable phenomenon? The result of a Dickey-Fuller test is included in the table below. This tests if the long-run correlation is stationary, in other words, reverting to its long-term average. I performed the test both for a longer time period (10 years) and the 3 years of data seen in the graph. We can see that in both cases we have evidence that the rolling correlation is stationary. This implies that correlation tends to move around its long-term rolling correlation and on average will return to its mean. Therefore, when the correlation is outside its normal 95% confidence bound, it has significantly diverged from the average and will likely return to the average. But does the diverging of the correlation go paired with any changes in the ETF returns?

Figure 2

When is the correlation abnormal?

To visualize when correlations are diverging from the norm I made two graphs containing indicators when the correlation is lower than the lower confidence bound and higher than the upper bound. The graph can be found below. The yellow vertical lines imply that the rolling correlation was below the 95% confidence interval, and the brown lines imply it was above. When looking at the yellow lines indicating low correlation, we see that one of both assets will outperform the other in returns shortly after the indicator. In December 2023 the correlation of the assets got below the bounds. After which the STOXX600 dipped while the S&P500 continued to rise at the start of January. We see a similar phenomenon in January 2025 where the STOXX600 rises strongly while the S&P500 goes down after low correlation has been signaled. And after may/june of 2025 the S&P500 outpermors the STOXX600 after low correlation indicators were triggered. If a low correlation shock always leads to a significant difference in short term returns is too difficult to say. But it is certainly interesting to observe that in this case.

There some moments where the correlation is very high. Just before September 2024 and in April 2025 the high correlation is accompanied with both assets going down. When both markets move down the rolling correlation suddenly becomes very high since both markets are dropping. During times of crisis the correlation between indices also goes up Sandoval and Franca (2012). This is logical, when both assets go down, they are by definition more correlated. Therefore when the whole market is trending downwards these market ETFs are higher correlated.

Figure 3

Conclusion

We found that the rolling correlation is stationary, so it revolves around its long term mean. And in our data low correlation shocks lead to either one of the assets outperforming the other in the short term after the shock. For investors holding STOXX600 and S&P500, it could be worthwhile to incorporate some indicators of the correlation between these assets. To investigate if this is a recent phenomenon due to chance or a general trend, more research is needed. Maybe an interesting short deep dive for investors with both ETFs. Potential trading strategies incorporating a correlation indicator could be interesting for this correlated financial markets.

References

Sandoval, Leonidas and Italo De Paula Franca (2012). Correlation of financial markets in times of crisis. Physica A: Statistical Mechanics and its Applications 391 (1), 187–208.