This article covers how Underwriting teams typically use GEO, both in response to live events and as an ongoing part of risk selection and portfolio management.

This article assumes your organisation is already sharing portfolio data with MIS and that Damage Factors are configured, either your own or the MIS baseline. If either of those isn't set up yet, start with [Sharing Your Data with MIS] and [Policy Insights] before working through the steps below.

Every figure referenced here, exposure, portfolio comparisons, moratorium timing, is only as current as the last time your portfolio was shared with MIS. Check the As At Date shown on the Insights tab before relying on a number for a live decision.

This document sits a level above the feature articles it links to. For field-by-field detail on how a table or layer works, see the linked article rather than expecting it to be repeated here.

For retail or major clients you have a direct relationship with (or brokers, for wholesale business), pre-event layers give you a window to reach out before impact, either to suggest preventative measures or to prepare the client for what the claims process will look like if they're affected.

Beyond the client relationship benefit, this also reduces uncertainty on reserving capital, since you're not waiting on a loss to materialise before you start thinking about it.

[SCREENSHOT: Wind Speed Probability layer ahead of hurricane landfall, showing likelihood of Cat 1 strength winds reaching an area]

See [Pre-Event Layers & Outlook].

Ahead of a hurricane, typhoon, or other windstorm, MIS may publish Wind Speed and Wind Speed Probability layers that extend out over water rather than stopping at the coastline. If you carry offshore risk, such as platforms or marine assets, these layers can be exported as shapefile or GeoJSON and run through your own modelling software to check exposure that a land-only view would miss.

[SCREENSHOT: Wind Strength layer extending offshore over the Gulf of Mexico, shown alongside the Layer Downloads panel]

See [Pre-Event Layers & Outlook] and [Downloading Layers].

Insights on GEO always run against your current in-force data. That means any historic event in the archive can be run against your portfolio as it stands today, not as it stood when the event actually happened.

This is useful for understanding how a past loss driver would affect your book now, and how your book has changed since. If there's a historic event that caused you major losses, ask MIS to run it against your current portfolio to see that impact today. Contact [email protected] to set this up.

See [Historic Events Archive].

Comparing your current book against historic events tells you how it has performed. Stress testing tells you what might happen if you grew it in a particular direction. MIS can load fabricated locations into a portfolio, assign them an aggregated exposure, and track what additional loss that value would have caused across a set of historic events. If you'd rather model a scenario that hasn't happened at all, rather than fabricated growth on top of real historic events, see [Simulated Events] instead.

For example, if you're deciding where within a state to grow your book, you could check which zip codes have been hit repeatedly over the last few years and what the damage impact would have been on new value placed there. There are a few ways to build this:

Any of these can then be run against several concerning historic events at once to build a fuller picture before you expand into an area. Reach out to Solutions to scope an exercise like this.

Risk Assessment Layers aren't tied to a live event, they're indicative of potential future occurrences, which makes them a binding-time tool rather than a response-time one.

[SCREENSHOT: Strike Risk layer showing minimal-to-severe risk zones across a region, with CEP-based risk banding]

See [Events We Respond To] for where Strike Risk sits alongside MIS's live event response.

With an Oblique Imagery archive going back over a decade for the US and Canada, and independent of whether an event has occurred, GEO can support underwriting decisions on individual properties outside a catastrophe response:

[SCREENSHOT: Oblique Imagery comparison showing a property before and after demolision, same north-facing angle]

See [Oblique Imagery].

Damage Factors let Underwriting set a defined, consistent view of risk across events, broken down however suits you, by line of business, construction type, method of attachment, or region. A common pattern is to run two factor sets side by side: a tighter, lower-range set for Binder books (where high accumulation of low-severity losses is the concern) and a wider, more reactive set for Open Market contracts (where the concern is flagging potential large individual losses).

You can switch between factor sets, or host separate portfolios, to get a view specific to each part of the book. These ranges are applied against the Assessment categories MIS publishes on each layer, see [Assessment Scales] for what each category means before setting your own ranges.
​

[SCREENSHOT: Damage Factors sets shown side by side, a tighter set and a wider set]

See [Policy Insights] for how Damage Factors are configured and how they drive the exposure calculation.

When multiple events occur close together in time and geography, for example two hurricanes landing in the same region within a few weeks, it can restrict your ability to write new business in the affected area for longer than necessary.

GEO helps in two ways: identifying the affected areas as early as possible so you know whether an incoming request for cover falls within them, and providing daily updates once an event has passed over a region, so you can confirm that cover written after that date won't be exposed to claims from that event and get back to underwriting there sooner.

Take a fictional portfolio, ahead of a hurricane tracking toward the Florida Gulf Coast, similar in shape to the Hurricane Milton event on GEO.

Can we run a historic event against our portfolio without waiting for MIS to do it?
Not currently, historic re-runs against your current in-force data go through the Client Success team. Reach out to Solutions to request one.

How far back does Oblique Imagery go?
Over a decade for the US and Canada, independent of whether a specific event occurred there.

Should Underwriting and Claims use the same Damage Factor set?
Not necessarily. Factors are configured per user but visible to your whole team once applied to an event, so different teams can run their own sets side by side without overwriting each other's view.

What's the difference between stress testing a portfolio and using Simulated Events?
Stress testing loads hypothetical new value onto your existing portfolio and runs it against real historic events. Simulated Events models a scenario that hasn't happened at all, such as an industry catastrophe scenario. Use stress testing for portfolio growth decisions and Simulated Events for broader scenario planning.