How Billionaires Die:
Methodology

Study Design

A retrospective observational analysis examined billionaire deaths between January 1, 2015, and December 31, 2025, using individual persons as the unit of analysis.

Data Collection and Verification

Case-finding: Candidates were identified from public billionaire directories (Forbes, Bloomberg, Hurun), yearly memoriam compilations, and targeted media searches in English and local languages.

Death confirmation: Deaths were classified as either High confidence (primary statement or two independent reputable sources matching identity and death year) or Medium confidence (one reputable source plus identity match on two or more attributes). Lower-confidence cases were excluded.

Billionaire status: Inclusion required net worth ≥$1 billion at or near death, verified by Forbes, Bloomberg, or Hurun estimates within 24 months before death. The hierarchy prioritized Forbes, then Bloomberg, then Hurun, using estimates closest to death. All figures were converted to constant 2025 US dollars.

Cause-of-Death Classification

Deaths were categorized into: aging-related disease, external cause, other medical (non-aging), or unknown/not disclosed.

“Aging-related disease” encompassed conditions with age-related incidence: cardiovascular/cerebrovascular disease, cancer, neurodegenerative disease, metabolic/renal failure, chronic respiratory disease, frailty/multi-organ failure, and terminal infections in advanced age or chronic illness. Natural-cause phrasing without competing indicators classified deaths as aging-related for those aged 70+. External causes overrode other classifications.

Sensitivity analyses tested stricter classification rules.

Mortality Analysis

Population data: Living billionaire counts came from the Forbes Billionaires Evolution dataset (2015–2024) and Forbes live website (January 2026 proxy for 2025).

Age banding: Standard 10-year bands (50–59, 60–69, 70–79, 80–89) plus a 90+ bucket were used.

Reference population: Hong Kong, with the world’s highest life expectancy (~85.5 years), provided the benchmark using sex-specific mortality rates from 2015–2022 life tables.

SMR calculation: Standardized Mortality Ratio = Observed Deaths / Expected Deaths, with confidence intervals computed using exact Poisson methods.

Results: Male billionaires (N = 336 deaths; 19,527 person-years, ages 50+) showed an overall SMR of 0.71 (95% CI: 0.63–0.79), indicating 29% lower mortality than Hong Kong’s general male population. The advantage concentrated at ages 60–89 (SMR 0.48–0.70, p < 0.001) and disappeared at 90+ (SMR 0.89, not significant). Female billionaires were excluded due to small sample size (28 deaths).

Projected Life Expectancy

Living billionaires’ projected life expectancy applied empirically derived, age-band-specific SMRs to a Gompertz mortality model fitted to Hong Kong rates:

m_HK(x) = α · exp(β · x)

Parameters (α = 1.41 × 10⁻&sup5;; β = 0.0992 per year) were estimated via OLS. The billionaire hazard was computed as m_bill(x) = m_HK(x) × SMR(age band), with remaining life expectancy obtained through numerical integration, truncated at age 110. These are period estimates assuming current mortality persists.

Biases and Limitations

Several mechanisms likely inflate the observed mortality advantage:

• Survivor bias: Only those surviving to $1 billion accumulation are included, excluding early deaths.
• Healthy-founder effect: Traits enabling wealth creation (discipline, education, stress tolerance) may independently promote longevity.
• Late entry: Median age at first Forbes appearance is 57; 18% enter at 70+.
• Pre-death wealth loss: Health deterioration may reduce billionaire status before death through asset sales, reduced business performance, or medical expenses.
• Unstable threshold: Net worths fluctuating near $1 billion may drop below during illness.

The SMR of 0.71 represents an upper bound on any causal wealth effect. Additional limitations include uneven public reporting across regions and languages, opacity of private wealth structures, imprecision in private asset estimates, and 30% absence of cause-of-death disclosure.