Cigna: PBM Scale, Evernorth, and the AI Transformation of Pharmacy Benefit Management

Executive Summary

Cigna Group (~$235B FY2024 revenue) has repositioned itself as a health services company rather than a pure insurer, having divested most of its U.S. commercial insurance business to Health Care Service Corporation in 2024 while retaining Evernorth — the parent of Express Scripts, the largest PBM in the U.S. The strategic logic is that PBM scale is more defensible than insurance underwriting. AI is testing that logic directly: the opacity-dependent economics of PBM are precisely where AI-powered transparency is creating the most structural pressure. Cigna's concentrated bet on Evernorth makes it the most AI-exposed major health company among the large-cap healthcare services names.

Business Through an AI Lens

Following the HCSC transaction, Cigna's revenue is dominated by Evernorth Health Services, which generates approximately $175B in annual revenue from three businesses: Express Scripts (PBM, the largest by script volume), Cigna Healthcare (residual employer/international insurance), and Evernorth Behavioral Health.

Express Scripts processes approximately 1.5 billion adjusted scripts annually. The cognitive work in PBM operations — formulary design, prior authorization, drug utilization review, rebate contract analytics, specialty drug management — is extensive and highly structured. This is exactly the type of domain where AI models trained on pharmaceutical and clinical data can match or exceed human analyst performance at a fraction of the cost. The threat is not just that AI reduces Cigna's own operating costs; it is that AI enables competitors (both other PBMs and direct-to-employer AI-native players) to replicate historically complex PBM functions with far smaller teams.

Express Scripts' specialty pharmacy division, Accredo, is the higher-growth, higher-margin component of the PBM. Specialty drug management involves complex clinical protocols, patient support programs, and manufacturer relationships — areas where AI augments rather than replaces human clinical expertise, providing more durable margin.

Revenue Exposure

Express Scripts' revenue model has two primary value creation mechanisms that AI threatens differently. The first is formulary management and rebate administration — extracting manufacturer rebates through formulary placement leverage and retaining a portion as operating income. AI transparency tools that allow plan sponsors to see full rebate economics and compare alternatives in real time directly attack this mechanism.

The second is administrative spread — capturing the difference between what plan sponsors pay per claim and what Express Scripts pays the dispensing pharmacy. Multiple state Medicaid programs have banned spread pricing, and federal legislation targeting commercial spread pricing has bipartisan support. AI-powered PBM auditing tools, now offered by startups like Rx Savings Solutions and Capital Rx, allow employers to quantify their spread exposure with precision that was previously impossible.

Cigna's residual insurance segment (~$50B in premiums from employer and international markets) faces the same prior authorization AI scrutiny as UNH and Humana but with less intensity given the commercial-only focus and lower political salience relative to Medicare.

Cost Exposure

Cigna's cost structure at Express Scripts is predominantly pharmacy network payments, drug costs, and administrative overhead. AI-driven automation of prior authorization processing, drug utilization review, and member services represents a meaningful efficiency opportunity — the company has guided toward $400-600M in annual technology-driven efficiency gains through 2027.

On the investment side, Express Scripts must build or buy AI capabilities in formulary optimization to remain competitive against smaller, more agile competitors that are native to AI architectures. The company's legacy IT infrastructure — built on decades of claims processing systems — is a genuine liability when competing with cloud-native competitors. Migrating these systems while maintaining 1.5B scripts per year in operational continuity is a material execution risk.

Behavioral health is an interesting cost exposure angle. Evernorth Behavioral Health manages mental health benefits for large employers. AI-powered therapy apps (Spring Health, Lyra Health) are being embedded directly into employer benefit plans, potentially disintermediating Evernorth's behavioral benefit management role. The $5B behavioral segment is small but carries above-average margins that could compress as AI-native providers offer lower-cost alternatives.

Moat Test

Express Scripts' moats are real but under active assault. The genuine moat elements are: scale in pharmaceutical purchasing (negotiating leverage with manufacturers that smaller players cannot replicate), data from 1.5B annual scripts (clinical patterns, drug interactions, adherence data), and multi-year employer contract lock-in. A Fortune 500 employer switching its PBM is an 18-24 month process involving procurement, legal, and benefits administration — that friction is a meaningful switching cost.

The weakest moat element is the rebate and spread model itself. If AI-enabled transparency eliminates the information asymmetry that makes these mechanisms profitable, Express Scripts' revenue per script falls dramatically. The company has begun transitioning some clients to pass-through pricing models — but these generate 30-50% lower revenue per transaction. The transition dynamics are manageable if slow; a rapid regulatory mandate for industry-wide pass-through pricing would be an earnings shock.

Timeline Scenarios

1-3 Years (Near Term)

The PBM transparency regulatory timeline is the most consequential near-term variable. The FTC's 2024 PBM report laid the groundwork for federal legislation; if the current Congress passes a PBM reform bill targeting spread pricing and rebate transparency, Express Scripts faces $1.5-3B in operating income reduction. Simultaneously, the integration of HCSC's commercial book into a competing PBM (HCSC retained its own PBM relationships) removes a natural channel for Express Scripts cross-selling.

3-7 Years (Medium Term)

The specialty pharmacy market is the medium-term battleground. GLP-1 drugs (Ozempic, Wegovy, Mounjaro) represent one of the largest specialty drug categories in history — Express Scripts' management of the GLP-1 formulary and prior authorization process for its 25M+ covered commercial lives is a significant revenue and influence opportunity. AI-driven GLP-1 adherence management and outcomes tracking could make Accredo the most valuable component of Evernorth.

7+ Years (Long Term)

The long-term question is whether PBM functions become embedded within AI-powered direct-to-employer health platforms (Amazon Clinic, Mark Cuban's Cost Plus Drugs at scale, or a tech-native benefits administrator). If large employers can access drug pricing, clinical oversight, and benefits administration through a single AI platform that eliminates the PBM intermediary, Express Scripts' entire model is at risk. This is a 7-10 year scenario but not implausible.

Bull Case

GLP-1 specialty drug management becomes a major revenue driver — with 50M+ potential U.S. patients and $15,000-25,000 annual drug costs, Express Scripts' formulary leverage with Novo Nordisk and Eli Lilly generates extraordinary rebate economics at scale. AI-powered clinical decision support embedded in Accredo's specialty programs improves patient outcomes and creates measurable value that justifies above-market pricing. Cigna's employer-focused residual insurance book benefits from AI-driven population health management that reduces commercial claim costs, supporting MA and employer plan renewals. The HCSC transaction proceeds fund a transformative AI acquisition in benefits administration technology, repositioning Evernorth as a platform rather than a processor.

Bear Case

Federal PBM legislation mandating pass-through pricing reduces Express Scripts' net revenue per script by 30-40%, creating a $3-5B operating income gap that requires structural cost reduction. AI-native competitors (Capital Rx, Rightway, Amazon) capture 10-15% of the commercial PBM market by 2028 using transparent pricing models that AI-sophisticated HR buyers increasingly demand. The HCSC transaction, intended to simplify Cigna's story, creates execution risk as Express Scripts loses cross-selling opportunities from the divested commercial book. Evernorth Behavioral Health is partially displaced by AI-powered mental health apps embedded directly in employer benefit portals, eliminating $500-800M in high-margin revenue.

Verdict: AI Margin Pressure Score 7/10

Cigna scores a 7/10 — the highest among the major health insurers — because its concentrated bet on the PBM model is the S&P 500's clearest example of an information asymmetry business being actively dissolved by AI-powered transparency. Unlike UNH and Humana, which have diversified AI tailwinds in their care delivery assets, Cigna's post-HCSC portfolio is heavily tilted toward the most vulnerable portion of the healthcare services stack. The GLP-1 thesis is real but does not offset systematic PBM margin compression.

Takeaways for Investors

The HCSC transaction changed the risk profile more than consensus recognizes. Post-divestiture, Cigna is effectively an Evernorth bet. Investors who modeled Cigna as a diversified health conglomerate need to update their framework — the company now has far less insurance business to buffer PBM disruption.

GLP-1 formulary management is the most important 2026-2028 earnings variable. Express Scripts' ability to negotiate meaningful GLP-1 rebates and manage adherence at scale will define whether Accredo drives earnings growth or becomes a low-margin dispensing operation as manufacturers shift to direct-to-patient models.

Monitor the AI-native PBM cohort data. Capital Rx and Rightway are publishing multi-year employer client data showing savings versus incumbents. When that data achieves statistical significance across large employer groups, it becomes a direct recruiting tool for Express Scripts defections.

PBM legislation probability is systematically underweighted. Analyst consensus models PBM reform at 20-30% probability. Given bipartisan agreement on the issue (unusual in U.S. healthcare policy) and active FTC, CMS, and Congressional pressure, the true probability is likely 40-60% of meaningful reform by 2027.

Evernorth behavioral health deserves separate monitoring. The $5B behavioral segment's margin profile could compress significantly as AI mental health platforms scale. Tracking Spring Health, Lyra, and employer benefit adoption rates is a leading indicator of this disruption.