---
title: the score is the easy half
canonical_url: https://ensurance.app/guide/the-score-is-the-easy-half
markdown_url: https://ensurance.app/guide/the-score-is-the-easy-half.md
subtitle: "every building can now carry a climate risk score. what to fund, where, when, and who pays is still undecided"
category: nature-finance
---

# the score is the easy half

*every building can now carry a climate risk score. what to fund, where, when, and who pays is still undecided*

**What is climate risk decision making?** It is the step after the score: turning a hazard rating for a building or a town into a choice about what to fund, where, when, and who pays. The score says where it will hurt. Climate risk decision making commits money, land and years to moving one loss, and it has to survive a budget meeting.

The score is the easy half now. Almost any building can carry a flood, fire, wind and heat rating, and a free public index will give you an expected loss for every census tract in the United States. What the rating cannot tell you is which dollar to spend first, and that is where most of the market still stops.

Much of the hazard in that score is set by ground the building does not own. The watershed holds the snow and the slope above the town. The floodplain spreads the flood before it reaches the street, and the marsh takes the edge off the surge. That ground was setting the hazard long before anyone scored it, and it keeps doing so whether or not anyone pays for it. [**Ensurance**](/manual/ensurance?from=guide) is how the people who depend on it can pay for its keeping. It is not the watershed, and it does not stop a storm.

## the score got cheap

The score got cheap from both ends: a free public baseline and commercial scores for single buildings.

The public baseline costs nothing. The Federal Emergency Management Agency's (FEMA) National Risk Index gives **expected annual loss** (EAL, the average yearly loss you would see over many years) for every US census tract across 18 hazards, split by buildings, people and agriculture. It is not per building and it does not look forward, but it puts a dollar figure on nearly every place in the country.

On the commercial side, MSCI, the index and analytics firm, completed its purchase of the climate risk modeler First Street on 3 August 2026. What that deal means is covered in [the $120 million diagnosis](/guide/the-120-million-diagnosis?from=guide).

When a number gets cheap, the scarce thing moves. The scarce thing now is a decision someone is willing to sign.

## what the score answers and what it leaves open

Hazard scores, catastrophe models and climate analytics answer real questions. They answer them about the building. The questions a budget has to answer are different.

| the score answers | the decision still has to answer |
|---|---|
| Where it will hurt | What to fund there |
| How bad, on average (the EAL) | Which loss you are trying to move: the average year or the year that breaks you |
| How the hazard changes under the climate paths you chose | Which bet you left out: the water, slope or forest the site depends on |
| Which buildings and plants are exposed | Which ground sets the hazard, and who owns it |
| How the hazard trends over the coming decades | When to act, and what waiting costs |
| Nothing | Who pays, from which budget line, and for how long |

The left column is a description. The right column is a set of commitments. Turning one into the other is the work, and no score does it for you.

:::johnson
**the score tells you where it will hurt. it does not tell you what to fund.**

Risk data is cheap and everywhere. The decision after it (which loss, which bet, which ground, when, and who pays) is still the hard half.

[read next: the average year is not the year that breaks you →](/guide/the-average-year-is-not-the-year-that-breaks-you?from=guide)
:::

## the usual decision lands on the building

When a decision does get made, it usually lands inside the fence: raise the building, harden it, move the equipment up a floor, renew the policy for another year. These are not bad choices. Hardening pays: the National Institute of Building Sciences found in 2019 that federal mitigation grants save about $6 for every $1 spent, counting benefits to society as a whole, not insurer losses. Insurance does what nothing else does, which is pay after the loss.

Two things limit that default. The first is money. Insurers in some states now give discounts for hardened homes, but for wildfire retrofits Resources for the Future found the costs are "orders of magnitude greater than the insurance savings." The premium alone rarely pays for the work.

The second is geography. The hazard in your score often starts outside the fence: in the burned slope above the water intake, the forest above the reservoir, the floodplain and the wetland between the river and the street. Researchers have run coastal wetlands and forest treatment through industry-grade models, with the ground and without it. The figures are mostly modeled, not counted, and [the watershed isn't in the asset register](/guide/the-watershed-isnt-in-the-asset-register?from=guide) sets them out, but the mechanism is plain: ground nobody in those buildings owned changed what the storm or fire did to them.

A building-only decision can only spend money on the building. If the hazard is set upstream, part of the answer is upstream too, and that ground usually belongs to someone else. A [risk assessment that ends on a list of parcels](/guide/a-risk-assessment-is-a-list-of-parcels?from=guide) is one way a town gets there. Knowing where a site touches living systems is not the same as knowing which failure would cost it money; [exposure is not risk](/guide/exposure-is-not-risk?from=guide) works through that gap.

## five questions after the score

Climate risk decision making comes down to five questions. Each gets its own post in this series.

1. **Which loss?** Expected annual loss is an average; the bad year sets the capital, and living systems move some losses and not others. [the average year is not the year that breaks you](/guide/the-average-year-is-not-the-year-that-breaks-you?from=guide)
2. **Which bet?** A stress test reports the scenarios you ran, and the source your site depends on is usually the one nobody ran. [the bet that isn't in your scenarios](/guide/the-bet-that-isnt-in-your-scenarios?from=guide)
3. **Which ground?** A risk model lists hazards, assets and damage curves, and the watershed that shapes the hazard has no row. [the watershed isn't in the asset register](/guide/the-watershed-isnt-in-the-asset-register?from=guide)
4. **When?** Every year you defer has a cost, and it rises as bad years worsen, the fix gets dearer and cover gets harder to buy. [waiting is a position](/guide/waiting-is-a-position?from=guide)
5. **Who pays, and on what case?** A budget says yes to a named loss, a no-action baseline, a cost and a payment it can book. [the cost-benefit case for the ground upstream](/guide/the-cost-benefit-case-for-the-ground-upstream?from=guide)

The order matters. Skip the first question and the cost-benefit case gets built on the average year, which undercounts what the bad years cost. Skip the third and every option on the table is a building option.

## what to do with a climate risk score

You do not need a new model to start. You need to read the one you have differently.

- **Pull the dollar figure, not just the grade.** For a US place, [read FEMA's expected annual loss by peril](/place?from=guide). A score of 7 out of 10 is a mood; a loss in dollars per year can go in a budget.
- **Split the loss.** Ask what share comes from frequent, mid-size events and what share from the rare extreme year. They call for different tools.
- **List what the site depends on.** The water, the slope, the road, the forest, the floodplain. [Map the exposure](/exposure?from=guide) and the ground behind it.
- **Name who owns that ground.** A farm, a forest agency, a land trust, a county. That name is the start of a payment.
- **Write down what you are not covering.** Work on the ground does not remove the extreme year, and insurance may not be on offer for all of it. Put that in the record even when nothing in the budget touches it. For the investor's version of that gap, see [the tail risk you can model but can't buy](/guide/the-tail-risk-you-can-model-but-cant-buy?from=guide).

That is a decision memo, not a dashboard. It also makes the next conversation with an insurer, a lender or a council shorter.

## what we do and do not do

We do not sell climate risk scores, run catastrophe models, compute tail losses or pick the right year to act. We do not hold risk or pay claims. Hazard scores and loss models from specialist suppliers are inputs to the decision, and a real case still needs a hazard specialist and a loss model of its own.

What **ensurance** does is the part after that: name the ground that sets the hazard, give it an account and a [certificate](/specific?from=guide) people can fund, and send the proceeds to the people who keep it. [Risk Resilience](/risk-resilience.ensurance?from=guide) is the flow the protocol tracks for that work. A certificate starts as ensured, meaning its keeping is funded, and moves toward entrust, meaning permanent protection. Funding alone does not finish that path.

## frequently asked questions

### what is climate risk decision making?

It is turning a climate hazard score into a choice about what to fund, where, when, and who pays. The score describes the hazard at a place. The decision picks which loss to move, which ground to work on, when to act and how to pay for it, and writes that down so a budget can approve it.

### what do you do with a climate risk score?

Turn it into dollars, split the frequent losses from the rare extreme ones, list what the site depends on outside its fence and who owns that ground, then decide what to fund and when. A score alone does not tell you any of that.

### is a climate risk score the same as an expected loss?

No. A score ranks hazard, often on a scale of 1 to 10. Expected annual loss is the average yearly loss in money over many years. Both describe the building or the place, and neither says what to fund or how bad the worst year could be.

### can nature lower climate risk?

For some losses, yes. Wetlands, forests and floodplains can lower frequent and mid-size losses where the mechanism fits the place, though much of the evidence is modeled rather than observed. They do not remove the extreme tail, and a record storm or fire can exceed any buffer.

## sources

[FEMA, National Risk Index](https://hazards.fema.gov/nri/) — Census-tract expected annual loss for 18 natural hazards, split by building value, population and agriculture

[MSCI, completion of the First Street acquisition](https://ir.msci.com/node/22881) — Purchase completed 3 August 2026

[National Institute of Building Sciences, *Natural Hazard Mitigation Saves* (2019), federal grants fact sheet](https://nibs.org/wp-content/uploads/2025/04/ms_v3_federalgrants.pdf) — Federal mitigation grants (FEMA, EDA, HUD) save $6 per $1; societal benefits, not insurer losses

[Resources for the Future, Working Paper 25-30, *From Risk to Reward: Insurance Discounts for Wildfire Mitigation* (2025)](https://www.rff.org/publications/working-papers/from-risk-to-reward-insurance-discounts-for-wildfire-mitigation/) — Retrofit costs are "orders of magnitude greater than the insurance savings"

## the series

1. [the score is the easy half](/guide/the-score-is-the-easy-half?from=guide) — **Read first:** What a climate risk score leaves undecided, and the five questions after it
2. [the average year is not the year that breaks you](/guide/the-average-year-is-not-the-year-that-breaks-you?from=guide) — Expected annual loss is an average; the bad year sets the capital, and living systems move some of it, not all
3. [the bet that isn't in your scenarios](/guide/the-bet-that-isnt-in-your-scenarios?from=guide) — A stress test reports the risk you modeled; the source your site depends on is usually the scenario nobody ran
4. [the watershed isn't in the asset register](/guide/the-watershed-isnt-in-the-asset-register?from=guide) — A climate risk model needs hazards, assets and damage curves; the ground that sets the hazard has no row
5. [waiting is a position](/guide/waiting-is-a-position?from=guide) — When the bad years get worse, the protection is worth more and each year of waiting costs more
6. [the cost-benefit case for the ground upstream](/guide/the-cost-benefit-case-for-the-ground-upstream?from=guide) — A worked example on a hypothetical water utility, from the loss to a payment a budget can book