---
title: "diversification lowers your share, not the loss"
canonical_url: https://ensurance.app/guide/diversification-lowers-your-share-not-the-loss
markdown_url: https://ensurance.app/guide/diversification-lowers-your-share-not-the-loss.md
subtitle: "spreading across regions shrinks your slice of each bad year. it does not shrink the year, and it stops working when one footprint covers several of your regions"
category: nature-finance
---

# diversification lowers your share, not the loss

*spreading across regions shrinks your slice of each bad year. it does not shrink the year, and it stops working when one footprint covers several of your regions*

**What is geographic diversification?** It is spreading exposure across places that are unlikely to have their bad year at the same time, so no single flood, fire, drought or storm takes more than a small share of a book. Reinsurers, property and infrastructure funds, lenders and allocators all lean on it, and it works. Its limit, stated plainly: geographic diversification lowers your share of a loss. It does not lower the loss.

The flood is still the size the floodplain and the weather let it be, and the fire runs as hard as the forest and the fuels above the town allow. That ground sets part of how big the event gets inside its footprint, whether or not anyone holds, spreads or funds the exposure on top of it. [**Ensurance**](/manual/ensurance?from=guide) is how money reaches that ground. It is not the ground, and it is not insurance.

## why geographic diversification works

Start with the strongest case, because it is a strong one.

A Florida hurricane, a Japanese typhoon, a European windstorm and a California earthquake do not share a footprint. A reinsurer that writes all four can hold less capital than four specialists would hold between them, because the chance that all four peaks arrive in one year is small. That diversification credit is much of why a cedent can buy cover at a price it can pay ([what reinsurance cannot cede](/guide/what-reinsurance-cannot-cede?from=guide) has the treaty desk's side).

Portfolio construction runs on the same logic. A property fund spreads across metros, an infrastructure fund across corridors, a farm lender across counties. Each buys independence between places, so one bad season is a dent and not a wound. [The fine print of the free lunch](/guide/the-fine-print-of-the-free-lunch?from=guide) covers diversification in general; this post stays with the map.

None of that is wrong. If you hold exposure to places, you should spread it.

## your share, not the loss

Here is the turn. Diversification is a statement about the holder, not about the place.

An illustration, with no real figures. Ten holders each own all ten buildings in one valley. In the year one valley floods, one holder loses ten buildings and nine lose nothing. Now each holder owns one building in each of ten valleys. In the same year, every holder loses one building. Each holder's bad year got smaller and easier to fund. The valley lost ten buildings either way.

Spreading turns one ruinous loss into many survivable ones. It sorts who holds the loss. It does not reach what made the loss that size.

:::johnson
**diversification lowers your share of the loss, not the loss.**

Spreading across regions makes each holder a smaller owner of every bad year. The footprint, and the ground that sets the hazard inside it, stay exactly where they were.

[investors: map the footprints in your book →](/solutions/investors?from=guide&topic=spatial-risk)
:::

## it works only while places are independent

The credit depends on one assumption: that the places you spread across have their bad years at different times. Two things break it.

### footprints are bigger than the regions on the map

A book's regions follow administrative lines: a state, a country, a metro. A hazard's footprint ignores them.

| footprint | what it crosses | evidence |
|---|---|---|
| A drought region | Most of a country's farmland at once | In 2012, about 80% of US agricultural land saw low rainfall and high temperatures, the most extensive drought since 1956 (Council of Economic Advisers, 2013) |
| A stalled jet stream | Several breadbaskets on different continents in the same summer | In summers when amplified jet-stream waves held for two weeks or more, cereal output fell 4% across the affected regions and up to 11% in one; simultaneous heat waves in major crop regions were 20 times more likely when the pattern held (Kornhuber et al., 2019) |
| An ocean cycle | Weather on several continents in one season | El Niño–Southern Oscillation is the only climate mode the authors found capable of forcing globally synchronous crop failures; in 1983 it caused the largest one in the modern record (Anderson et al., 2019) |
| A river basin | Seven states, many cities, farms and plants on one supply | Nearly 40 million people in seven US states rely on the Colorado River system for some or all of their municipal water; it irrigates nearly 4.5 million acres (Bureau of Reclamation, 2021) |

The first three rows measure crops, not insured property. The point carries: when weather moves together across regions, everything exposed to that weather moves with it. A book spread across ten farm states can be one position in a drought year.

### many places converge on one source

The second break is quieter. Places that never share a storm can share a supply. An illustration: a book spread across Arizona, Nevada and Southern California reads as three markets on a risk report, yet all three are among the seven states whose cities and farms rely on one river system for some or all of their water (Bureau of Reclamation, 2021). How much varies by city, but a utility, a homebuilder and a farm lender in those markets can each look independent and still depend on the same upstream watershed. [The correlation outside the matrix](/guide/the-correlation-outside-the-matrix?from=guide) shows why a price matrix misses that; a map catches it, if you draw it by source instead of by state.

At the largest scale every position stands on one biosphere; [you cannot diversify out of the plate](/guide/you-cannot-diversify-out-of-the-plate?from=guide) makes that argument. This post stays at the scale of a basin or a drought region, where the dependence is specific enough to name and fund.

## what spreading cannot do

**It cannot shrink the event.** The loss in a footprint is set by what is exposed there and by the hazard, which the ground partly sets. Spreading moves the loss between holders.

**It cannot hold when the footprint grows.** When one event touches more of a book than the model assumed, the credit shrinks and the capital needed rises. Raising price, capping exposure or leaving is then rational, and the risk lands on whoever cannot leave; see [a footprint turns a thousand policies into one event](/guide/a-footprint-turns-a-thousand-policies-into-one-event?from=guide) and [when the insurer leaves, the risk stays](/guide/when-the-insurer-leaves-the-risk-stays?from=guide).

**It cannot fund the place.** Every holder in the footprint benefits when the ground is kept, and spreading pays for none of it. The wider a loss is spread, the smaller each holder's reason to pay looks, though the reason added up across holders is unchanged.

## what is left when spreading runs out

Keep diversifying. Then add the one move it cannot make: hold the place itself as a position.

That means funding the ground that sets the hazard in a footprint you are already exposed to: the soil and aquifer recharge in a drought region, the headwater forest and floodplain above a river city, the forest fuels above a corridor. The honest limit belongs here. No forest changes an El Niño, and no floodplain stops a heat dome. Living systems change how some hazards land in some places (how far water spreads, how hard a fire runs), and the evidence is strongest for frequent and mid-size losses, not the extreme year ([the average year is not the year that breaks you](/guide/the-average-year-is-not-the-year-that-breaks-you?from=guide)). They move part of the loss curve, not all of it.

[Long or short the source?](/guide/long-the-source?from=guide) is the allocator's test of whether you hold any of what you depend on. The short version: spreading makes you a small holder of many footprints. Funding the source makes you a holder of one place, a position that works on part of the loss rather than only on your share of it.

## what we do and do not do

We do not build portfolios, run catastrophe or climate models, or hold risk. Ensurance is not insurance: it pays no claims, and no carrier is known to credit ensurance funding in price, so it lowers no one's premium today.

What is live today: accounts, called [agents](/agents?from=guide), at named places, such as [colorado-river.basin](/colorado-river.basin?from=guide); [certificates](/specific?from=guide) and coins on Base that fund a place; and proceeds routed to the people who keep the ground. Instruments that pay on a measured trigger are designed, not live ([what parametric ensurance actually does](/guide/what-parametric-ensurance-actually-does?from=guide)). [What ensurance cannot do](/guide/what-ensurance-cannot-do?from=guide) lists the rest.

## what an allocator can do this quarter

1. **Redraw your regions as footprints.** Group holdings by drought region, river basin and fire corridor, not only by state.
2. **Count shared sources.** Ask where each holding's water comes from and what ground sets its hazard. Where the answer repeats, you hold one position.
3. **Ask what the credit assumes.** Whoever gives you a diversification benefit, ask which places it treats as independent, and on what evidence.
4. **Name one place and fund it.** Pick the footprint where most of your book overlaps and fund its ground alongside the others who share it.

Allocators and family offices: [investors](/solutions/investors?from=guide&topic=spatial-risk), for mapping exposure in sensitive locations and holding certificates on the places it depends on. Grids, roads and corridors: [infrastructure investors](/solutions/infrastructure-investors?from=guide&topic=spatial-risk). Lenders and real-asset capital: [capital providers](/solutions/capital-providers?from=guide&topic=spatial-risk). Or [explore the places already on the map](/explore?from=guide).

This is research, not investment, insurance, legal or tax advice.

## frequently asked questions

### what is geographic diversification?

Geographic diversification is spreading exposure across places that are unlikely to suffer losses at the same time, so that one event takes only a small share of a book. It lowers each holder's share of a loss. It does not lower the loss itself, and it works only while the places are truly independent.

### does geographic diversification reduce climate risk?

It reduces your concentration in any one place's bad year, which is real and worth having. It does not reduce the hazard, and it weakens when one footprint (a drought region, a stalled jet stream, an El Niño year, a shared river) covers several of the regions you spread across.

### what can't diversification fix?

The size of the event, a footprint bigger than your regions, places that share one source, and the condition of the ground that sets the hazard. Spreading moves a loss between holders. Funding the floodplain, forest, soil or aquifer works on part of the loss itself, for some hazards in some places.

## sources

[Council of Economic Advisers, *Economic Report of the President* (2013), chapter 8](https://www.govinfo.gov/content/pkg/ERP-2013/html/ERP-2013-chapter8.htm) — "About 80 percent of agricultural land experienced low rainfall and high temperatures, making the 2012 drought the most extensive since 1956"

[Kornhuber, K. et al. (2019), "Amplified Rossby waves enhance risk of concurrent heatwaves in major breadbasket regions," *Nature Climate Change* 10: 48–53](https://doi.org/10.1038/s41558-019-0637-z) — With persistent wave patterns for two or more summer weeks, cereal production down 4% averaged across affected regions and up to 11% in one; 20-fold higher risk of simultaneous heat waves in major crop regions; figures as reported in [Columbia Climate School, 9 December 2019](https://news.climate.columbia.edu/2019/12/09/newly-identified-jet-stream-pattern-could-imperil-global-food-supply-says-study/)

[Anderson, W. B. et al. (2019), "Synchronous crop failures and climate-forced production variability," *Science Advances* 5(7): eaaw1976](https://pmc.ncbi.nlm.nih.gov/articles/PMC6609162/) — Abstract: ENSO, "the only mode capable of forcing globally synchronous crop failures," caused the largest synchronous crop failure in the modern record in 1983

[US Bureau of Reclamation, SECURE Water Act report (2021), Colorado River Basin fact sheet](https://www.usbr.gov/climate/SECURE/docs/2021secure/factsheets/Colorado.pdf) — Basin overview: nearly 40 million people in seven states rely on the Colorado River and its tributaries for some or all municipal needs; nearly 4.5 million acres irrigated

## the series

1. [risk has an address](/guide/risk-has-an-address?from=guide) — **Read first:** what spatial risk is, the pin, the footprint and the source, and the one move finance does not make
2. [a footprint turns a thousand policies into one event](/guide/a-footprint-turns-a-thousand-policies-into-one-event?from=guide) — Accumulation risk: why insurers count roofs by the footprint
3. [a loan book is a map](/guide/a-loan-book-is-a-map?from=guide) — Concentration risk: what lenders and bondholders share with the ground under their borrowers
4. [diversification lowers your share, not the loss](/guide/diversification-lowers-your-share-not-the-loss?from=guide) — Where geographic diversification works, and where the footprints outgrow it
5. [when the insurer leaves, the risk stays](/guide/when-the-insurer-leaves-the-risk-stays?from=guide) — Uninsurable areas, and who ends up holding the risk
6. [how to manage spatial risk at the source](/guide/how-to-manage-spatial-risk-at-the-source?from=guide) — Map the footprint, find the ground that sets the hazard, fund it together

**Companion series:** [what multi-peril insurance covers](/guide/what-multi-peril-insurance-covers?from=guide) — the several perils one place carries, and the ground as the layer under the policy.