Applying the System Asset Pricing Model to Pharmacy Benefit Management: Measuring the System Welfare Cost of the U.S. Rebate Intermediary

PBM rebate system destroys 6.30 in welfare per 1 of profit

The Big Three PBMs earn 27.6B/yr in profit while imposing ≈174B/yr in system welfare costs. System beta β_W = 6.3 (90% CI: 4.8–8.1).

• β_W = ΔW / Π: welfare destroyed per dollar of industry revenue
• Monte Carlo (100,000 draws): 99.4% of specs show welfare-negative
• System-adjusted payoff Π = −66B × (1 − 6.3) = −146.3B/yr

System Asset Pricing Model maps CAPM to welfare

CAPM prices risk against market portfolio; SAPM prices payoff against welfare portfolio. β_W > 1 means welfare loss exceeds private gain.

• CAPM: E[R_i] = R_f + β_i(E[R_m] − R_f)
• SAPM: Π = Π − β_W · Π = Π(1 − β_W)
• PBM β_W = 6.3: each dollar of industry revenue associated with $6.30 welfare loss

Welfare cost flows through six distinct channels

Channel-weighted β_W = 6.3. Morbidity channel dominates (100–528 B/yr).

• 1. Rebate trap & formulary distortion: $36–60 B/yr deadweight loss
• 2. Patient out-of-pocket extraction: $2.5–3.2 B/yr via list-price coinsurance
• 3. Independent pharmacy destruction: 40M Americans in pharmacy deserts
• 4. Medication non-adherence: 100–528 B/yr avoidable morbidity
• 5. List-price inflation: $350B cumulative gross-to-net bubble
• 6. Governance capture: 17.5M/yr lobbying shields 62.7B DIR fees

Non-adherence costs exceed total PBM revenue by 3.6×

100,000 preventable deaths per year from cost-related non-adherence. Lower bound of $100B exceeds PBM industry revenue.

• Risk of dying from unaffordable Rx is 10× homicide risk
• Avoidable healthcare costs: $100–528 B/yr
• Channel 4 alone makes system deeply welfare-negative

99.4% of simulations confirm welfare negativity

100,000 Monte Carlo draws propagate uncertainty across all six channels. β_W never below 2.8 even under most charitable assumptions.

• Median β_W = 6.3; 90% CI [4.8, 8.1]
• P(β_W < 1) = 0.0000%
• Marginal β_W at current scale = 9.1 (exceeds average 6.3)

PBMs would need 86% efficiency to break even — they achieve 13%

Break-even system efficiency μ* = 0.86. Current estimated efficiency μ ≈ 0.13 — a factor of 6.6 gap.

• μ*: share of intermediation margin converted to welfare gains
• At μ = 0.86, β_W = 1.21 (welfare-neutral)
• Current μ = 0.13: PBMs retain 5 per 100 for genuine cost-containment

Standard correction mechanisms don't work for PBM intermediation

Pigouvian tax fails because externality is the business model itself. Coasean bargaining fails due to engineered opacity and undefined property rights.

• Externality is not a byproduct — it is the production itself
• Information asymmetry: PBM contracts shielded by NDAs
• Who 'owns' the rebate? Manufacturer, PBM, plan sponsor, or patient?

PBM rebate system ranks 5th highest among 10 SAPM domains

β_W = 6.3 places it above Bitcoin mining (5.0) and below monoculture agriculture (8.6).

• PFAS contamination: β_W = 35.2 (highest)
• Frontier AI deployment: β_W = 7.4
• Auto emissions: β_W = 1.21 (lowest among calibrated)

2026 FTC settlement and CAA confirm MST prediction

FTC's Express Scripts settlement mandates WAC-neutral formularies and delinked compensation. CAA 2026 requires compensation delinking by 2028.

• Missing System Theorem: β_W > 1 activities admit Pareto-improving reallocation
• FTC settlement: first institutional implementation of corrective reallocation
• CAA 2026: structural reform, not marginal tax

PBM rebate system is a regulatory artifact, not a force of nature

Created by 1987 Anti-Kickback safe harbor, sustained by ERISA preemption. High reversibility: welfare cost can be eliminated in 24–36 months.

• No impossibility theorem blocks reform — this is institutional PST
• Reform requires political will, not technological breakthrough
• β_W trajectory under reform: projected decline to <2 by 2032