Beyond The Headlines

As uncomfortable as it is, we’re frequently reminded that the stock market is noisy and often volatile, swayed by a constant stream of opinions, data and events. Unexpected shocks – a sudden geopolitical tremor, a surprising earnings’ miss or a swift policy change – are an inevitable part of market dynamics. The recent news surrounding UnitedHealth is a case in point.

Starting in December 2024, UnitedHealth’s share price faced significant pressure. The tragic murder of Brian Thomson, CEO of UnitedHealth’s insurance subsidiary (United Healthcare) triggered intense media scrutiny and public backlash against health insurance companies. This was compounded by a year-end stop-gap funding bill containing potential Pharmacy Benefit Management (PBM) reform and a subsequent February 2025 Wall Street Journal (WSJ) report detailing a Department of Justice (DOJ) investigation into UnitedHealth’s Medicare Advantage (MA) billing practices. These sequential events culminated in a substantial share price drop – an approximately 24 per cent drawdown since December.

Navigating such volatility requires a disciplined approach. In this newsletter, we outline how we generally think and decide the best course of action when faced with negative surprises.

Everything we do begins and ends with our core principles. At the forefront is long-term investing – viewing stocks as ownership in businesses, not as trading tickets in a stock market. When unexpected events arise we return to our roots, asking what it means to the fundamental economics of the business, by specifically addressing two key questions:

1. Does this event signal a potential erosion of the company’s competitive advantage?
2. Does it alter the company’s underlying growth trajectory?

These questions are our compass. If our research confirms the company’s competitive moat remains strong and its long-term growth potential unchanged, we generally maintain our position. Conversely, if the event reveals a fundamental shift in either of these key areas, we reassess our investment thesis. Depending on the magnitude of the change, we consider the likelihood of selling. This isn’t about short-term price swings; it’s about recognizing when the core assumptions that support our investment thesis are invalidated. We believe this is a sensible and effective approach.

Applying this framework to the most recent UnitedHealth situation, the central allegation made by the WSJ against the company is that it may have obtained overpayments from the CMS (Centers for Medicare & Medicaid Services, on behalf of the federal government) by manipulating the reimbursement system, specifically through inflating patient risk scores. To understand this claim, it’s important to clarify the link between risk scores and premium payments. The health insurance premium is essentially determined using a cost-plus model, where the “cost” represents projected future medical expense and the “plus” is a profit margin due to the insurance company. Logically, these projections are based on an assessment of patient health status, which is the key. In practice, this assessment is called “risk coding”, by which healthcare professionals assign diagnosis codes which then result in a “risk score”. In short, diagnosis codes drive risk scores. Higher risk scores lead to higher risk-adjusted premiums. The WSJ report attacked the company right on this link – overbilling by engaging in improper or aggressive diagnosis and risk coding practices. It also reported that the DOJ has launched a civil investigation into the company’s MA risk coding practices.

The publication of the report triggered the stock price to open more than 12 per cent lower. However, it is important to acknowledge that the report contained allegations, not established facts. The nature, the scope, and even the existence of any alleged improper billing practices at UnitedHealth remains unclear, to the public at least. Applying Shannon’s information theory,¹ which defines information as the reduction of uncertainty, the quality of this information is relatively low. This suggested the sharp market reaction was potentially driven more by emotional or other short-term oriented trading, rather than a rational evaluation of the company’s underlying value.

While we have conducted thorough internal research and discussions on the WSJ articles, predicting a conclusive outcome of a “reported” civil investigation is impossible. Both the WSJ’s reporting and our analysis are subjected to limitations, including incomplete information, inaccurate context, potential misinterpretation and biases. However, it is part of our job to form a view amid incomplete information and uncertainty. We therefore prioritize focusing on key information we deem important and knowable and apply our best judgement to fill the knowledge gaps when necessary.

For instance, our evaluation is grounded in an understanding of the health insurance industry’s broader context. This sector is characterized by an inherent tension between imperative cost containment and the potentially limitless demand for healthcare services, leading to conflicting interests among stakeholders. As a result, a perfect pricing mechanism or system in which everyone is happy is practically impossible. Consequently, MA payment has been a subject of ongoing debate and scrutiny ever since its inception. The current controversy surrounding MA overpayments is not a novel development, but rather a recurring feature of the industry. Particularly, MA coding and risk adjustment, which are foundational to payment determination, have faced persistent scrutiny across generation of administrations. While issues raised in the recent WSJ report may have been new to the general public, they have long been familiar with regulators.

To ensure MA payment accuracy, the CMS has, over the years, implemented several key mechanisms. Beyond general audits, Risk Adjustment Data Validation (RADV) Audits scrutinize the validity of diagnosis codes submitted by MA plans. Industry experts, including a former director² at Elevance Health, noted that numerous insurers like Centene, ChenMed and Optum³ have faced up to six CMS audits in a single year. These audits involve detailed reviews of sampled beneficiary medical records to substantiate reported diagnoses.⁴ Based on identified discrepancies in the sample, a payment error rate is calculated and then extrapolated across the entire MA contract. In case of identified overpayments, the CMS claws back funds from the plan. Furthermore, the risk adjustment formula undergoes periodic revisions to determine how to best weight diagnoses to accurately predict cost mitigating any emerging issue such as unnecessary diagnoses.

In the past four years, the Biden Administration has placed significant emphasis on addressing perceived MA overpayment (of around 5 per cent), making it a key objective to reduce MA funding. Several policy changes have been enacted since 2023. These include: 1) reduced annual rate increases; 2) a significant change of the risk adjustment model, from version 24 (V24) to version 28 (V28),⁵ to reduce the impact of risk coding; 3) formalized RADV process to improve detection of inappropriate diagnosis coding; and 4) more stringent quality bonuses (star rating) criteria. Notably, both RADV and V28 directly address the major concerns raised by the WSJ, seeking to align MA payments with predicted costs based on risk scores. It’s important to note that WSJ articles also citied data up to 2021, which predates the recent regulatory changes.

Given the level of historical and ongoing scrutiny, the prospect of systematic and intentional inappropriate billing by UnitedHealth, while theoretically possible, appears unlikely. With nearly 10 million MA beneficiaries, representing the largest national market share at 29 per cent, UnitedHealth already operates under the regulator’s microscope with extensive legal compliance and ethical frameworks in place. Any discrepancies, should they arise, are more likely attributable to the technical complexity of the MA payment system and limited in scope rather than sustained, deliberate misconduct. Furthermore, the sector’s premium and margins are ultimately managed by the annual competitive bid process. The transparency of the process effectively mitigates the risk of large-scale, sustained inappropriate billing, because any substantial deviations from industry norms are readily identifiable.

However, as Professor Elroy Dimson aptly noted, risk means more things could happen than will happen. A proper risk assessment acknowledges the worst outcome. If UnitedHealth were engaging in significantly more aggressive risk adjustment practices than its peers, the margin outperformance driven by this factor has been estimated to be up to 0.5 percentage points – approximately 2 per cent of UnitedHealth’s group earnings. Given the company’s scale and diversification, the overall negative impact appears limited.

We believe a proper assessment of any material event or information requires a comprehensive understanding of a company’s fundamentals and its broader ecosystem. UnitedHealth has consistently delivered profitable growth over many decades, navigating a variety of economic cycles and political shifts. We attribute this sustained success to its superior strategic foresight, a network of interconnected competitive advantages, inherent resilience and alignment with secular growth trends. These strengths have demonstrably compounded over time. We highly doubt systematic inappropriate billing practices are a significant driver of UnitedHealth’s sustained success. Based on our evaluation, the likelihood of this event materially impacting the company’s moat or long-term growth prospects appears quite low. Therefore, the event, while warranting diligent monitoring, doesn’t alter our view on UnitedHealth’s underlying value and long-term investment thesis. We believe that the market negativity is overdone. 

Ultimately, investment decisions involve subjective judgement, which is susceptible to error. We remain committed to updating our views as new information emerges. However, the main purpose in this newsletter is to offer transparency into our thought process – how we respond to, and think beyond immediate headlines, with the mindset of the long-term investor and an unwavering focus of underlying business value, which, we believe, is what drives enduring investment success.

¹Claude Shannon was an American mathematician, electrical engineer and computer scientist, known as the “father of information theory” and credited with laying the foundations of the information age. In Shannon’s theory, the simplest definition of information is that information is a decrease in uncertainty.

²Tegus interview with a former director of provider economics at Elevance Health on Mar 19^th, 2025.

³Optum is a subsidiary of UnitedHealth.

⁴The agency uses a stratified sampling approach to select contracts at higher risk for improper payments with a focus on historical track record and unusual coding patterns.

⁵The v28 risk model, which is designed to reduce upcoding by reducing the comorbidity multipliers utilized by providers in the MA program, is introduced in 2023, and is in the process of being phased-in (between 2024-2026). It will see an incremental 33 per cent phase-in each year until fully incorporated in 2026.