Health and medical costs

The LISF (Insurance and Surety Institutions Law) groups three lines within the accident and health operation: personal accident, medical expense and health (article 25). Medical expense is by far the largest. According to the AMIS 2024 Accident and Health Executive Report, direct premiums for medical expense insurance totaled 158,266 million pesos in 2024, 13.9% more than in 2023, split almost evenly between individual policies (79,279 million) and group policies (78,987 million). The line represents 93% of accident and health premiums.

Premium growth did not translate into breathing room. Medical expense claims cost reached 117.4 billion pesos in 2024, compared with 101.6 billion in 2023, and the line's combined ratio went from 99% to 101%: for every peso of premium, technical operations spent slightly more than a peso. The retained loss ratio was 76.5% of earned premium, and 81.5% for group business.

AMIS counts 13.95 million medical expense insureds at year end 2024, equivalent to 10.7% of the population, with an average age of 36. The risk premium, measured as claims divided by the number of insureds, was 8,470 pesos per insured for the line as a whole, 12,593 for individual and 6,587 for group. In 2020 the total was 5,172 pesos: an increase of 64% in four years.

Which conditions cost the most. By ICD-10 chapter, neoplasms led medical expense claims in 2024 with 24,992 million pesos, followed by diseases of the musculoskeletal system (21,134 million), the digestive system (12,920 million), injuries (12,288 million) and diseases of the circulatory system (10,470 million). At the other end of the distribution, the AMIS catastrophic claims study records individual cases exceeding 100 million pesos in cumulative cost; the largest, an endocrine disorder in a 19 year old man lasting 15 years, totals 190.3 million in December 2024 pesos. That tail explains why sum insured caps and excess of loss reinsurance matter so much in this line.

Where the money goes. Hospitalization, including physician fees, absorbed 84% of the spending covered by medical expense insurance in 2024; out-of-hospital drugs, 9%; and out-of-hospital laboratory and imaging tests, 1%. It is a hospital-driven line: the price trend of hospitals and physician fees carries more weight than that of outpatient drugs.

It is worth separating three things the press tends to mix up. General inflation is the change in the INPC. The INPC health inflation measures prices of a basket of drugs, consultations, surgeries, hospitalization and tests paid by households. The medical trend of an insured portfolio measures the growth of cost per insured, which includes price, utilization, intensity and service mix. All three grow at very different rates.

With INEGI (National Institute of Statistics and Geography) data compiled by AMIS, annual inflation of the health index was 5.02% in December 2025 and 4.77% in August 2026, compared with 3.69% and 3.26% for the general INPC on those dates. Within health, drugs rose 6.67% in the year to December 2025 and surgical procedures 5.35%. Over a long horizon the difference is small: from December 2015 to December 2024 the general index rose a cumulative 54.92% and the health index 54.94%.

The trend of insured portfolios is another story. The medical expense risk premium per insured grew 9.8% in 2024 and 64% between 2020 and 2024, far more than any price index. Consultancies that survey insurers reach similar figures: Aon's Global Medical Trend Rates Report 2027 projects for Mexico a gross medical trend of 13.7% in 2027 (14.8% in 2026), with a trend net of inflation of 10.3%, among the highest in Latin America, whose average is 10.4%. AMIS, in its reform proposal presented to the Chamber of Deputies in February 2026, spoke of private medical costs growing between 14% and 16% per year.

Medical trend can be broken down into multiplicative factors. Separating them helps with projection, because each responds to different causes and some can be managed through product design, provider networks or underwriting.

The long run: the Getzen model. No 13% rate can be sustained for decades, because medical spending would end up absorbing the entire economy. The Society of Actuaries maintains, as part of its research on health care cost trends, the Getzen model for projecting long-run trends, used mainly to value retiree health obligations. The December 2025 update estimates long-run increases of 5.0% per year that gradually converge to per capita income growth, raises the long-run share of health spending in GDP to 20.0% and the resistance point at which trend slows to 18.5%. The model was calibrated for the United States, but the idea is general: in a long-term valuation, for example of postretirement medical benefits under NIF D-3 or IAS 19 (see employee benefits), the trend must converge to a sustainable rate.

Expected medical cost grows with age, slowly at first and then very fast. AMIS data show it in frequency: in 2024 medical expense morbidity was 7.1% for ages 0 to 19, 8.5% for 20 to 39, 10.2% for 40 to 59 and 21.3% for 60 and over. Severity also rises with age, so cost per insured multiplies several times between ages 30 and 70.

The age-based rate is built with a relative cost curve: the cost at each age divided by the cost at a reference age. In practice it is smoothed with a piecewise exponential form and expressed as an annual aging factor, the percentage by which cost rises for each year of age. In individual policies with attained age premiums, the insured sees at renewal the combination of medical trend and their own aging.

There are two alternatives to the attained age approach. A level premium overcharges at younger ages and builds a reserve that is used in old age, as in whole life (see mortality tables); a community rated premium charges the same at all ages and depends on young people joining; if they do not, the average cost rises and antiselection spirals. The 2026 AMIS proposal for savings schemes to smooth premiums in old age points in the direction of the level premium.

Suppose the gross cost of each claim rises 12% and the policy deductible stays fixed at $25,000. The portion of the deductible paid by the insured does not change, and the increase above it is absorbed by the insurer, except for coinsurance. Worse still: claims that previously did not exceed the deductible now do and start generating payments. The result is that the insurer's net cost grows more than the gross cost. This is deductible leveraging, a well known effect in excess of loss reinsurance that operates every year in medical expense insurance.

The effect is larger the higher the deductible relative to the typical claim, and it disappears if the deductible and the coinsurance cap are indexed to the same trend. That is why some insurers review their deductibles periodically, and why a rate that applies the gross trend to the net premium falls short.

Weeks or months pass between the date a condition occurs and the date the insurer finishes paying for it: the hospital bills, the insured submits reimbursements, physician fees and drugs are added. At each month end there are incurred claims the insurer does not yet know about or knows only in part. The reserve for incurred but not reported claims (IBNR) and the reserve for outstanding claims cover that lag. Their general framework is described in technical reserves.

In medical expense insurance development is short compared with liability, but very voluminous, so the work is done with monthly triangles by month of occurrence and month of payment. The most common method in health is completion factors: for each development age, the proportion of the ultimate claim already paid is estimated from the triangle's development factors. In the most recent months the factors are unreliable because very little has been paid, and they are blended with an expected loss ratio estimate, in the style of Bornhuetter and Ferguson.

A particular feature of health is long duration conditions. An oncology treatment or a chronic disease generates payments for years on the same claim, and policies with continuity coverage carry the obligation beyond the policy term. Reserves must recognize those cash flows, and the catastrophic claims reported by AMIS, with durations of 15 to 20 years, show that the tail exists.

The Mexican system is fragmented. Formal private sector workers and their families are entitled to IMSS (Mexican Social Security Institute), which reported a covered population of 79.2 million people at year end 2025, with 22.5 million registered jobs. State workers are served by ISSSTE (Social Security Institute for State Workers) and state institutes, and there are separate services for Pemex, the armed forces and other groups. The population without social security is served by state health services and, since federalization, by IMSS-Bienestar, a body created by decree published in the DOF (Official Gazette) on August 31, 2022; by the first two months of 2025, 23 states had signed coordination agreements to transfer their services to it.

Financing is what sets Mexico apart within the OECD. According to Health at a Glance 2025, the country spends 1,588 dollars per capita on health at purchasing power parity, compared with an OECD average of 5,967, equivalent to 5.9% of GDP versus 9.3%. With data from the WHO health expenditure database published by the World Bank, out-of-pocket spending represented 41.2% of current health expenditure in 2023 and domestic public spending 48.6%. More than 40 of every 100 pesos spent on health come directly out of households' pockets.

For private insurance, those figures mean two things. There is huge latent demand for protection against catastrophic expenses that are paid out of pocket today. And private insurance coexists with a public system that is the first payer for most people: an insured may use IMSS for a chronic condition and private insurance for a surgery, which affects the frequency and severity the insurer observes.

Long-term care (LTC) insurance pays an income or services when the insured loses the ability to perform basic activities of daily living without help, such as bathing, dressing or eating. Unlike medical expense insurance, it covers dependency rather than treatment, and its cost depends on the duration of dependency more than on medical intensity.

Actuarially, it is modeled as a multiple state system: independent, dependent to varying degrees and deceased, with transition probabilities by age. The risks are long term and hard to diversify: mortality improvement lengthens the years lived with dependency, lapse rates turn out lower than assumed and low interest rates make the reserve more expensive. These are risks the actuary must stress test with particular care, and they connect with longevity risk.

In Mexico aging is already a statistical reality. INEGI's 2024 Mexican Health and Aging Study (ENASEM) counted 17.7 million people aged 60 and over; 74.4% had access to medical services, of whom 64.4% through IMSS, 19.3% through ISSSTE or state institutes and 13.5% through private services. Long-term care insurance has not developed as a mass market product in Mexico, and care for dependent people is provided mostly within families. Designing dependency coverage, pension-linked annuities or hybrid products with life insurance is one of the open spaces for the profession.

The Mexican medical expense market combines sustained premium growth with ever thinner technical margins. The line's technical contribution was a mere 2.4% of direct premiums in 2024, and 1.6% for group business; according to AMIS, the line's 2024 profit came mainly from investment income on reserves. The industry's diagnosis is that private medical costs grow between 14% and 16% per year, with a loss ratio of 74% and almost 3 million claims handled per year.

In February 2026 AMIS presented to the Chamber of Deputies a reform proposal for medical expense insurance built on four pillars: greater transparency of information for consumers, products for older adults with clear coverage and rules, savings schemes to soften the impact of premiums in old age and measures against medical inflation through greater transparency and efficiency in the care chain. Each pillar has actuarial content: publishable morbidity tables by age, level premiums with reserves, and measurement of trend by component to distinguish price from utilization.

For the actuary who signs technical notes and reserve valuations for accident and health, the challenge is twofold. On one hand, justifying trend assumptions that depart sharply from the INPC with evidence from the portfolio itself and from market sources. On the other, explaining to policyholders and the regulator why their premium rises faster than inflation: individual aging and deductible leveraging are technical explanations, not commercial excuses.