The three-layer structure, shown in federal regulatory text rather than asserted. The Standard Flood Insurance Policy is printed in full in the Code of Federal Regulations, which makes it a rare chance to read a real policy at primary source. Its Article V opens with the grant and the subtraction in a single sentence: "We only pay for direct physical loss by or from flood, which means that we do not pay you for: 1. Loss of revenue or profits; 2. Loss of access to the insured property or described location; 3. Loss of use of the insured property or described location; 4. Loss from interruption of business or production; 5. Any additional living expenses incurred while the insured building is being repaired or is unable to be occupied for any reason; 6. The cost of complying with any ordinance or law requiring or regulating the construction, demolition, remodeling, renovation, or repair of property, including removal of any resulting debris ... or 7. Any other economic loss you suffer."
Item six then carries the third layer inside itself: "This exclusion does not apply to any eligible activities we describe in Coverage D — Increased Cost of Compliance." A few lines further down the same pattern appears again. Article V excludes loss "caused directly by earth movement even if the earth movement is caused by flood," listing earthquake, landslide, land subsidence, sinkholes and gradual erosion, and then states: "We do, however, pay for losses from mudflow and land subsidence as a result of erosion that are specifically insured under our definition of flood." Two exclusions, two exceptions, in one Article. Anyone who stops reading at the word "excluded" has the wrong answer twice.
Why exclusions exist, which makes them predictable rather than arbitrary. Four reasons account for most of them. Some risks are uninsurable in an ordinary policy because a single event would hit every policyholder at once, which is the usual explanation for war and nuclear exclusions and part of the reason flood became a separate federal program. Some are excluded because they belong on a different policy, so an auto policy excludes the house and a homeowners policy excludes the car, and the household is expected to hold both. Some are excluded because they are within the insured's control or are not accidental at all, which covers intentional acts, criminal acts, and wear, tear and gradual deterioration. And some are excluded because the loss is a business risk rather than a fortuity, which is why professional liability generally sits outside a general liability policy.
An exclusion is not the only way a claim fails, and the difference matters when you are arguing about it. A loss can fall outside the grant of coverage in the first place, in which case the exclusions never come into play. Purely economic loss from bad professional advice, for example, generally fails a general liability policy because that policy insures damages because of bodily injury or property damage and the claim is neither, rather than because an exclusion removed it. Many such policies also carry a professional-services exclusion added by endorsement, but the claim would fail without one. The practical consequence is that "it's excluded" and "it was never covered" call for different responses: the first invites you to look for an exception or an endorsement, and the second means the coverage has to come from a different policy altogether.
The gap between two policies is where households actually get hurt. Because each policy is drawn to its own boundary, the exclusions in one are not guaranteed to line up with the grants in another. A homeowners policy excludes flood; a flood policy excludes the additional living expenses a homeowners policy would have covered. Both contracts are doing exactly what they say, and the household falls between them. The way to find those gaps is to read the exclusions of each policy against the grants of the others, which is tedious and is also the only reliable method. Where a gap is found, the fix is usually an endorsement adding coverage back or a separate policy, and NAIC's life insurance disclosure model shows the same idea at work even on a life contract, requiring a policy summary to state the amount payable at death "regardless of the cause of death, other than suicide or other specifically enumerated exclusions."