Expected annual loss explained

What is EAL?

Expected Annual Loss, or EAL, expresses modeled natural-hazard loss on an annualized basis. You may also encounter average annual loss, AAL, or annualized loss in screening discussions. Use those terms to explain the annualized concept, but preserve the source's EAL field name when copying FEMA values. Terminology alone does not establish that another model has the same inputs or scope.

In the sample, EAL_VALT holds total county expected annual loss and EAL_VALB holds its building component. Hazard-specific fields retain the hazard prefix, such as IFLD_EALT for riverine flooding. Keep total and component fields distinguishable in the workbook. Adding a total to its components counts the same modeled loss twice.

The unit is dollars per year. It describes a modeled annualized measure rather than an invoice, insurance premium or recent observed claim payment. EAL is an annualized modeled measure, not a forecast for the next calendar year. Preserve the original numeric value for calculations and apply display rounding only when presenting the result.

How is it different from a rating?

A rating is a category label. EAL is a dollar measure. The labels Very Low, Relatively Low, Relatively Moderate, Relatively High and Very High have an order, but the words do not define equal intervals of money. Avoid assigning arithmetic scores to those words and then treating a mean score as a source finding.

The comparison below retains hazard-specific labels next to hazard-specific dollar loss. Total county EAL sits in another column. Read across the same hazard before comparing counties. A riverine label beside an earthquake loss would create a misleading relationship, even if the row happened to describe the correct county.

Shared categories can contain different dollar totals. Conversely, an area with a less elevated category can still have a dollar value worth investigating. Relative standing and financial scale are complementary prompts for research. Keep both in the shortlist rather than selecting whichever measure produces the strongest sounding headline.

Why can a Relatively Moderate rating accompany a large dollar loss?

Harris County has an earthquake rating of Relatively Moderate and earthquake expected annual loss of $5,909,099 per year; Chittenden County has a rating of Relatively Low and earthquake expected annual loss of $2,147,955 per year. The Harris dollar loss is greater than the Chittenden dollar loss. Recorded building exposure is $838,838,945,279 in Harris and $40,414,880,347 in Chittenden.

Use the exact source wording “Relatively Moderate” when referring to the category. The contrast shows why a category alone cannot describe the financial scale. Compare the dollar fields for the same hazard and retain the source units.

Exposure helps explain why dollar scale deserves its own column. A county with more building value can present substantial modeled financial loss even when its category does not sound extreme. Exposure alone does not explain every difference: do not infer a causal decomposition from this small table. Keep the source model documentation with the finding before attributing a difference to any particular input.

BUILDVALUE is recorded building exposure in dollars, whereas building EAL is annualized modeled loss. They have different units and purposes. A large exposure value is not itself a loss estimate, and dividing county figures does not produce a validated damage rate for an individual facility. If your team creates a derived ratio, label it as your calculation and define its intended use separately.

How do I use annualized loss in a portfolio screen?

Use county EAL to compare area context and decide where more research is useful. Match facilities to counties, then summarize distinct county records beside the number of matched facilities. Keep the facility counts and county dollar totals in separate measures. Repeating a county total on every location row can inflate a portfolio sum without adding any new evidence.

If several facilities share a county, retain the shared geography in the output so the concentration is visible. Do not allocate county loss among facilities by dividing it by your location count. Your register does not necessarily include all exposed assets in the county, and equal allocation would introduce a model absent from the source.

Before presenting a comparison, confirm that the geography, release and field definitions match. Preserve full precision in the downloaded evidence and explain any displayed rounding. A change in source release should prompt a review of definitions and coverage before it is described as a real change in underlying risk.

What happens when a loss field is missing?

Keep missing numeric values distinct from measured zero values. The sample can contain a non-rating label alongside a null loss value or a numeric value. Read those fields independently and inspect the publisher's documentation before interpreting the combination. No Rating and Not Applicable should remain visible as source labels in a comparison.

Do not fill a missing EAL with a neighboring county's value or a portfolio average to make a chart complete. If the chart cannot display the missing value, explain the omission beside the chart and retain the original row in the evidence table. A visibly incomplete comparison is more useful than a complete-looking result with invented loss estimates.

Check the replacement source before using a refreshed value. Document the source revision, affected output and reviewer so others can trace the change. This lets a later reader distinguish a formatting correction, a source revision and a genuinely different analytical scope.

Which decisions require additional evidence?

Annualized loss does not identify when an event will occur. Compare the same hazard, geographic unit and source release before interpreting a difference in dollars. A missing component prevents a complete breakdown; it does not justify estimating the component from the total. Document which loss fields were used and whether any components overlap.

How should I cite the evidence and review date?

Last reviewed: 2026-10-23. Source: FEMA, FEMA National Risk Index. The county sample supplies the worked values. Keep the release and field notes from the methodology with exported figures. The review date records when this guide was reviewed; the underlying observations have their own periods.