Climate and sustainability for actuaries

Four international initiatives serve as references for this page. They are not the only ones, but they show the range: a practical knowledge hub, a regulatory review of what actuaries are doing, and two monitoring indices built on the same logic we propose for Mexico further on.

The Servicio Meteorológico Nacional (SMN, National Weather Service) of CONAGUA (National Water Commission) publishes an annual Reporte del Clima en México (Climate Report for Mexico) with the national mean temperature, accumulated rainfall and the evolution of drought. The figures below are taken directly from those reports. They should be read with an actuarial caveat: the SMN changed its reference normal from 1981 to 2010 to 1991 to 2020 between its 2020 and 2022 reports, so anomalies for different years are not always comparable with each other. Absolute temperature is.

A scenario is not a forecast. It is a coherent trajectory of emissions, policies and economics that lets us ask "what would happen to my portfolio if". Actuaries use two families. The IPCC scenarios (Sixth Assessment Report, AR6) combine shared socioeconomic pathways (SSPs) with radiative forcing levels and are the basis for physical risk. Those of the NGFS, the network of central banks and supervisors, translate those trajectories into macrofinancial variables and combine physical and transition risk; they are the ones supervisors ask for in stress tests.

For the actuary, the most useful part of AR6 is not the mean temperature but what happens to the extremes. Figure SPM.6 quantifies how many times more often an event that was rare in the 1850 to 1900 climate occurs today, and would occur with more warming. These are averages over the planet's land areas, not values for Mexico, but they show the order of magnitude of the frequency problem.

In November 2024 the NGFS published version 5.0 of its long-term scenarios, organized into four categories: orderly (Net Zero 2050, Low Demand, Below 2 °C), disorderly (Delayed Transition), hot house world (NDCs, Current Policies) and too little, too late (Fragmented World). Version 5.0 introduced a new damage function that markedly raises physical risk losses. In May 2025 the NGFS also published its first generation of short-term scenarios, with a three to five year horizon, closer to an insurer's planning horizon.

Mexico does not yet have an actuarial index of climate extremes equivalent to the ACI or the AACI. This section describes precisely how one could be built, which we provisionally call the ICAM (Índice Climático Actuarial México), using public Mexican data and the same architecture as the North American and Australian indices. We do not publish index values: calculating them requires processing, quality-controlling and homogenizing daily series from thousands of stations, work we have not done. What follows is the specification, designed so that a professional working group, a university or a graduate student can replicate it.

We propose 1961 to 1990 as the main base period, for three reasons: it is the period used by the ACI and the ETCCDI, which allows Mexico to be compared with the United States and Canada; it predates most of the recent warming, so the index measures the change rather than absorbing it; and the SMN keeps national series back to 1953. As a secondary series, the index would be recalculated against 1991 to 2020, the normal the SMN uses today, to align with its reports. Publishing both avoids the confusion seen in the table in section 03, where the same country appears to warm less or more depending on the normal used.

The question a board of directors asks most often about climate change is simple: if the event that used to be 1 in 100 years is now 1 in 50, how likely is it to occur during the life of a mortgage or a concession, and how much should it cost to cover? The calculator answers with a Poisson model. The presets use the IPCC frequency multipliers from section 04; applying them to a specific Mexican risk is an assumption the user must justify.

Between 2023 and 2025 Mexico went from having no specific rules to a disclosure and ESG risk management framework with several pieces. For the insurance actuary, what matters is that climate risks are no longer only a catastrophe underwriting issue: they come in through investment policy, the enterprise risk management system, the RCS (solvency capital requirement) and public reporting.

The International Sustainability Standards Board (ISSB) issued its first two standards on June 26, 2023. IFRS S1 sets general requirements for disclosing sustainability-related risks and opportunities that could affect the company's prospects in the short, medium and long term. IFRS S2 develops climate-related disclosure and is applied together with S1. Both fully incorporate the TCFD recommendations (governance, strategy, risk management, metrics and targets) and apply to annual periods beginning on or after January 1, 2024, subject to adoption by each jurisdiction; in the first year, climate-only disclosure is permitted.

S2 requires, among other things, assessing the resilience of strategy to climate change through scenario analysis, and disclosing scope 1, 2 and 3 greenhouse gas emissions. For an insurer, scope 3 emissions include those "financed" by its investments, and scenario analysis of the insurance portfolio is, in essence, an actuarial exercise: projecting claims, premiums and capital under alternative trajectories.

The link to IFRS 17 is direct. The same fulfillment cash flows projected to measure insurance contracts are the raw material for showing how they would change under a climate scenario. Consistency between the two disclosures will be scrutinized by auditors and supervisors.

According to the World Bank, in 2006 Mexico was the first government to use the catastrophe bond market to finance its risk, and by 2024 it had sponsored 20 bonds. In April 2024 the World Bank placed bonds for Mexico totaling 420 million dollars (earthquake and Atlantic storms), and in May another 175 million for Pacific hurricanes: 595 million in force until April 2028. The bonds are parametric: they pay if the event meets parameters of magnitude, location or central pressure, not according to measured loss.

The federal government complements the bonds with parametric insurance placed through Agroasemex. At the June 2026 renewal the parametric cover doubled to 10,000 million pesos (about 575 million dollars) for earthquake, volcanic eruption, hurricane and flood of medium to high severity, in force from June 5, 2026 to May 5, 2027. Together with the outstanding bonds, total parametric protection is around 1,170 million dollars.

The most important institutional change was the abolition of the Fondo de Desastres Naturales (FONDEN, Natural Disaster Fund). The decree published in the DOF on November 6, 2020 abolished 109 public trusts, including FONDEN's, which according to IMCO held 35,140 million pesos at the end of 2020; the SHCP formalized its dissolution in 2021. Since then, disaster funding has been allocated in the annual budget, with no ability to accumulate from one year to the next. For an actuary, this amounts to moving from a reserve fund to a pay-as-you-go scheme: risk transfer, which used to complement the fund, now carries a larger share of the protection.

The General Directorate of Epidemiology of the Ministry of Health monitors health effects of extreme natural temperatures every year. At the preliminary close of the 2023 heat season (epidemiological weeks 12 to 40), 4,306 cases and 421 deaths were reported, concentrated in Sonora (117), Nuevo León (102), Baja California (54), Tamaulipas (32) and Veracruz (31); 165 were attributed to the third heat wave, from June 1 to 22.

Those figures count confirmed heatstroke and dehydration, not total excess deaths. Much heat-related mortality is recorded as cardiovascular, renal or respiratory, so epidemiological surveillance underestimates the effect. For a life and health actuary the right tool is excess mortality: comparing observed deaths with those expected by age, sex, region and week, a technique the pandemic made familiar.

There are three practical implications. In medical expense insurance, more frequent heat waves raise hospitalization rates among older adults and outdoor workers. In life insurance, heat adds seasonal volatility to mortality that an annual table does not see. And in pensions and life annuities the net effect is uncertain: heat raises mortality among the elderly, but long-term longevity improvements depend much more on other factors. Without supporting data, longevity assumptions should not be adjusted for climate; their sensitivity should be tested.