In the September release of SpatialKey, several new features have been introduced that enhance product usability, analytic value, and accessibility to data. Additionally, a new application has been delivered to help you estimate loss potential to severe storms in the U.S. These features will help you unlock extended business value and potential from the SpatialKey product suite.

What’s New at a Glance

Comparison Maps

You can now create comparison map layers that visualize changes or relationships between two sets of data. These can be two instances of the same dataset (like snapshots of your portfolio from two points in time) or two entirely different datasets (like your portfolio and market data). Here are a few business applications where this capability would be valuable:

  • Accumulation/Portfolio Growth. Given two snapshots of your portfolio, you can easily visualize and evaluate change over time thematically by region (e.g. state, postcode) or a heat map depicting concentrations, as measured by any metric in your data (e.g. TIV, AAL). You can easily filter to see the trend for a specific line of business, underwriter, business organization, or any other dimension you choose.
  • Premium adequacy. Given a portfolio containing AAL and premium information, calculate loss ratios by geography. This highlights areas where your cat premium may not sufficiently cover estimated annualized losses. Visualizing this information will help you identify patterns or concentrations of risks where you assume a disproportionate amount of risk relative to the premium collected. You can further filter to see how this differs for any dimension of your data, like occupancy, construction, or distance to coast, to identify opportunities to refine pricing and risk selection at the underwriting level.
  • Model comparison. Model your data with two different models (e.g. between model vendors or different model versions) and see areas where the models differ most. Again, filter your data to explore more complex relationships – maybe one model predicts higher AAL in the Florida panhandle for wood construction. This can be used to help you formulate your view of risk and built intuition around the methodological differences and views of risk presented between modelers or model versions.
  • Market share. Given your portfolio and industry/market data (or even basic demographics, like household income or population) explore your market share by geographic area. Use this insight to identify growth opportunities and to understand where you have a disproportionate concentration of exposure in high risk areas relative to industry.

To illustrate, here’s a working example of model comparison using a sample residential insurance portfolio with modeled results from two modelers (Model Q and Model Z). Each model calculates the expected average annual loss (AAL) by location. This is the average dollar amount of expected financial loss that will occur in any given year.

In the past, we’ve been able to visualize a thematic map showing a single calculated value. In this case, we can look at the premium we collect for all risks:

Alternatively, we could also look at the total AAL values from each of our models, which would show us the distribution of expected loss. However, both deliver an incomplete picture. What we really want to understand is the adequacy of the premium relative to assumed risk. As a rough metric, we’ll look at the overall ratio of the AAL to premium, also known as the loss ratio. We’ll create a comparison map layer to visualize this metric.

To go about setting up the comparison, we can select the new Comparison Map option from the list of available visualization pods. We’ll be comparing two fields (premium and AAL) within this dataset. We can choose different ways of calculating the comparison. In this case, we will divide AAL by premium, which will give us the loss ratio. Any ratio greater than 1 suggests that we may lose money on an annualized basis relative to the premium we charged. First, we’ll compare premium with the AAL produced by Model Q. Our completed comparison setup reads: “Divide the Sum of Model Q’s Gross AAL by the Sum of Premium”.

Comparison map layers can be represented as heat maps or thematic maps. We’ll be using a thematic layer with a diverging color scheme. The diverging scheme we’ll use will use red for counties where AAL is greater than premium and green for counties where AAL is less than premium.

Once we add that layer to the map we see a thematic map of counties that visualizes premium adequacy.

We can see that the counties further inland are capturing more attractive premiums relative to the risk. The overall picture looks good; the risk level seems appropriately balanced by the premiums being collected. There is one county where the loss ratio is greater than 1. You can use this information to further evaluate risk composition in this county and refine your underwriting guidelines.

Now we’ll show a comparison map layer using the AAL from Model Z instead:

We can quickly see a big difference. There are quite a few counties along the western coast of Florida that have higher loss ratios. There is a clear difference between Model Q’s and Model Z’s view of risk, especially along the Gulf coast. Those red counties indicate that we may be over-exposed and could be in trouble if Model Z is accurate.

Comparison map layers can also be heat map layers. Here’s one final example of a heat map that shows the least profitable hotspots:

The new comparison map layers in SpatialKey open up new analytic possibilities. We’re excited to get this advanced analysis into our customers’ hands, with the same ease-of-use that you’ve come to expect from SpatialKey.

Policy Exposed Limit

To assist you in developing a more accurate view of your exposure that contemplates your policy terms and conditions, we have added a pod tailored to helping you examine your Exposed Limit. This feature comes with additional import configurations for policy data, a financial engine, and a new analytic pod. It calculates Exposed Limit in aggregate, based on a risk zone, and by policy for a specific peril. You can now visualize risk zone accumulations of Exposed Limit.

In its initial release, the financial engine contemplates the following terms and conditions for single and multi-­‐peril policies:

  • Policy Blanket Limits and Deductibles
  • Policy Minimum & Maximum Deductibles
  • Policy Layer/Line
  • Policy Attachment Point
  • Site Limits and Deductibles

All terms and conditions must be provided in a single currency and a single TIV value must be present for each site/location.

Users link policy data to any location dataset via a common field or ID. The user simply enables this feature for a location dataset in Manage Data, following the steps to map the relevant fields.

Once the data is linked, open any application and select the Policy Exposed Limit from the Pod menu for the desired dataset. Since Policy Exposed Limit is calculated by peril, select a peril from the list of perils supported by the policy data.

The pod will appear and present TIV, Location Count, and Exposed Limit for each policy and Exposed Limit in aggregate. Users can also aggregate by any imported policy attribute using the Group By feature, which aggregates Exposed Limit and Policy Counts. If users want to work with this data in Excel, simply export as a CSV file, and all underlying data will be exported.

As with all analytic pods, any filters applied in the map (e.g. defined by map extent, polygons, or event footprints) or other pods will refresh the content of the Policy Exposed Limit pod. In its initial release, Policy Exposed Limit is only computed for one risk zone at a time, which includes subject locations and associated policies defined within the entire dataset or any filter listed above.

SpatialKey will continue to make active investment in this feature by extending support for more policy terms and conditions, use cases, and streamlined workflows.

Scatter Plot Pod

SpatialKey has added a scatter plot chart to its library of analytic pods. This will enable you to efficiently identify patterns and evaluate relationships within your data. Beyond simply plotting points on a chart, we have enabled the ability to leverage analytics to identify concentrations of values easily, akin to the heat map feature you are used to.

Let’s evaluate premium adequacy for a portfolio, using premium and AAL, derived from a catastrophe model. In the figure below, red squares represent areas of the chart with a high concentration of points, where blue represents a very low number of points. The strong concentration of points from the lower left corner to the upper right shows a correlation between AAL and premium. However, this chart isn’t just for determining correlations, it also reveals outliers that may have otherwise gone unnoticed.

As with all other pods, this one is also linked to the map and other active pods. This means that you can filter views by selecting a point or collection of points within the chart. Simply, draw a selection box around several points to filter the contents of the dashboard. This data reveals an unfortunate negative disparity between premium and risk assumed. Insurers can quickly bring focus to these residential policies to understand their risk composition and publish this dashboard to line managers to review and possibly adjust underwriting guidelines.

Insurance companies use many different data models and formulas to assist in managing their portfolios and decision making. Models are one such source of analytic intelligence, but views of risk among them differ. Using two different models for hurricane risk, you want to evaluate where the opinions between these modelers materially differ. Once you understand this, you can calibrate your choice of model for specific lines and geographies that correlate well with your actual experience and management objectives. Scatter Plot charts can help!

In the figure below, you quickly note that Model Q consistently presents a higher view of risk for this portfolio and area of interest than Model Z. You can also see where their opinions significantly diverge. To understand more about this dissonance, you can filter your views by selecting specific groupings of points and leverage other analytic pods and the map to build intuition as to where and why the models might differ. Selecting a collection of points where Model Z is significantly higher than Model Q, we can discern that this is principally located in southwest Florida.

The Scatter Plot chart is just one of the many new ways Spatial Key can help your organization better understand its data. Better understanding leads to better decisions!

Annotation Pod

To assist you in collaborating with your colleagues, we are introducing an annotation pod. This enables you to easily draw attention to areas of import that you wish consumers of your dashboard to note.

It features formatting tools, including the ability to adjust font sizes, background color, transparency, and size. You can include as many annotations within your dashboard as you would like.

Now, you can more efficiently communicate value-added intelligence to your colleagues and reduce cycle time by leveraging this feature as a collaboration tool.

Data Mart

SpatialKey has always made it easy to import third party data; but you had to find, format, and import that data yourself. We have added the Data Mart to mitigate the need for you to invest significant amounts of time in gathering hazard and event data yourself. Within the Data Mart, you can choose from an expanding selection of data and with one click have the data imported into your organization to begin analysis against your exposures.

The Data Mart currently contains both real-time and historical data from industry weather, event, and census data sources. We provide easy access to NOAA SPC storm reports, NOAA tropical storm/hurricane forecast, historical hurricanes, historical tornados, USGS earthquakes, and MODIS Wildfire perimeters. The Data Mart also enables you to easily add visual map overlays to your dashboards, like FEMA flood zones, precipitation, and cloud cover.

The Data Mart can be accessed from two areas within SpatialKey. You can browse and load data within the home interface, or you can access the Data Mart within a dashboard (shown below).

You can add the same dataset more than once. For example, you may want to add the last 7 days of storm reports from NOAA and save that to a dashboard. You could come back later and pull in a new version, but preserve the previous one so the report doesn’t change. The screenshot below shows that example.

From within the Manage Data tab, you will see datasets that have new updates available. When you click “Update”, it will pull the latest version from the Data Mart.

SpatialKey is committed to streamlining your workflows, including your ability to leverage data sources that help illuminate and deliver insight into your risk.

Severe Storms Application

The Severe Storms application is now part of the SpatialKey Natural Perils Suite, which enables insurers to understand loss potential associated with live and historical events. It’s been designed from the ground up to effectively analyze NOAA SPC reports. Insurers can now manage exposure to U.S. tornado, hail, and straight-line wind events, as well as adapt the event footprint as more intelligence is available from NOAA and insurance claims data. Severe Storms provides you with the following purpose-built capabilities:

  • NOAA SPC data feed – always have access to the latest reports through direct data integration with SpatialKey.
  • Calendaring tool – efficiently find your desired SPC event reports based on storm type and date range.
  • Timeline histogram – review the frequency event reports by day and apply filters to target your areas of concern.
  • Storm drawing tool – flexibly develop paths that best describe the observed impacts of events.
  • Analytic pod – manage filters, risk zones, and damage assumptions.

Given the volume of SPC reports available each day, you need an advanced, yet simple, calendaring tool to help you target the desired set of reports.

With a focus on simplicity, Severe Storms introduces a new storm drawing tool, easily allowing you to visualize patterns in the NOAA SPC reported observations and develop potential storm paths in just a few clicks. It also includes easy-to-move nodes and scaffolding to draw the perfect storm and to modify the footprint as new SPC reports and claims data is available.

Once you have created your storm track, you can add several risk zones (i.e. bands or buffers), view TIV and apply damage factors by zone, and view aggregates for the footprint. As with all other SpatialKey apps, you can also leverage analytics available in other pods to understand risk composition. These pods are linked with all new pods associated with the Severe Storms app, so filters are propagated in all visualizations.

The Severe Storms application is yet another example of the ongoing commitment SpatialKey has made to quantify and evaluate risk, enabling insurers to efficiently make informed decisions about the impact of catastrophes on their portfolio.

We hope you are as excited to use these features as we are to bring them to you. Our best innovation comes from our great users, so please keep the great feedback coming. Please drop us a line or send us a tweet with your feedback.