Scenario
| Insured | Greenfield Medical Center |
| Line | Professional Liability — Medical Facility |
| Term | January 1, 2025 – December 31, 2025 |
Delta Actions Reference
| Action | Description | Matching |
|---|---|---|
| Modify | Replace a scalar or entire object | N/A — overwrites |
| Add | Append to a collection (set semantics) | Objects by id, primitives by equality |
| Remove | Remove from a collection | Objects by id, primitives by equality |
Premiums and Billing
Policy financial data lives at two levels: per-segment rating and full-term billing.Per-Segment Rating
Each segment carries its own rating data — aPolicyRatingResponse on the policy and an ExposureRatingResponse on each exposure. These typically include:
annualPremium— the premium as if that segment’s state applied for the full policy yearproratedPremium— the actual premium contribution for that time slice
Full-Term Billing
fullTermPolicyBilling is a cross-segment invariant — it’s identical across every segment in a version. It contains the billing summary for the entire policy term:
fullTermPolicyBilling delta with startDate/endDate spanning the full policy term (not just the endorsement effective date). This keeps the billing summary current.
fullTermPolicyBilling is not necessarily the sum of prorated slices. Full-term billing can include flat premium minimums, surplus lines taxes, policy fees, or other adjustments that are independent of per-segment proration. The FullTermPolicyRatingResponse (a separate cross-segment field) captures these aggregate rating details.Minimal vs Full Tracking
In This Walkthrough
Each transaction below includes structural per-segment deltas, per-segmentPolicyRatingResponse and ExposureRatingResponse rating, and the fullTermPolicyBilling update. The segment tables include a Policy Annualized Premium column — notice that this per-segment rate typically does not match the fullTermPolicyBilling.policyPremium shown below each table.
Creation
Transaction 1 — NEW_BUSINESS
Transaction 1 — NEW_BUSINESS
Effective: Jan 1 · Version: 1Initial policy creation. One exposure (Main Campus), three named physicians.Segments (1):
fullTermPolicyBilling: Premium 85,000 · Taxes 4,250 · Fees 500 · Total 89,750With a single full-year segment, the annualized premium matches the billing premium. This won’t last.
Request
| # | Date Range | Status | Exposures | Policy Annualized Premium |
|---|---|---|---|---|
| 1 | Jan 1 – Dec 31 | Active | Main Campus (120 beds, 3 physicians) | 85,000 |
Endorsements
Transaction 2 — ENDORSE: Add satellite clinic
Transaction 2 — ENDORSE: Add satellite clinic
Effective: Apr 1 · Version: 2A satellite clinic opens April 1. The new exposure is Added to the collection. Rating and billing updated for both segments — the existing Jan–Mar segment needs its prorated premium updated too, since it now covers only 3 months instead of 12.Segments (2):
fullTermPolicyBilling: Premium 98,000 · Taxes 4,900 · Fees 500 · Total 103,400The annualized premiums are 85,000 and 104,000 — neither matches the billing premium of 98,000. The billing total reflects the blended cost across both periods plus a flat surplus lines assessment.
Deltas
+1| # | Date Range | Exposures | Policy Annualized Premium |
|---|---|---|---|
| 1 | Jan 1 – Mar 31 | Main Campus | 85,000 |
| 2 | Apr 1 – Dec 31 | Main Campus + West Clinic | 104,000 |
Transaction 3 — ENDORSE: Add physician + specialty
Transaction 3 — ENDORSE: Add physician + specialty
Effective: Jun 1 · Version: 3A new surgeon (Dr. Okafor) joins the main campus June 1. Neurology added as a covered specialty. Both deltas use Add on nested exposure fields. Rating updated for the affected periods.Segments (3):
fullTermPolicyBilling: Premium 106,000 · Taxes 5,300 · Fees 500 · Total 111,800The Apr–Dec segment from version 2 split at the June boundary. Three different annualized rates (85k, 104k, 118k) but a single billing premium of 106k.
Deltas
+1| # | Date Range | Physicians | Specialties | Policy Annualized Premium |
|---|---|---|---|---|
| 1 | Jan 1 – Mar 31 | Patel, Nguyen, Hoffman | Cardiology, Orthopedics, Surgery | 85,000 |
| 2 | Apr 1 – May 31 | Patel, Nguyen, Hoffman | Cardiology, Orthopedics, Surgery | 104,000 |
| 3 | Jun 1 – Dec 31 | Patel, Nguyen, Hoffman, Okafor | Cardiology, Orthopedics, Surgery, Neurology | 118,000 |
Transaction 4 — ENDORSE: Remove physician, reduce beds
Transaction 4 — ENDORSE: Remove physician, reduce beds
Effective: Jun 1 · Version: 4Dr. Nguyen leaves the practice effective June 1. 10 beds decommissioned. Uses Remove on a primitive collection and Modify on a scalar.Segments (3):
fullTermPolicyBilling: Premium 101,000 · Taxes 5,050 · Fees 500 · Total 106,550Same segment count — the boundaries didn’t move, only the Jun–Dec data and rate changed. The annualized rate dropped from 118k to 112k reflecting fewer beds and one fewer physician.
Deltas
unchanged| # | Date Range | Beds | Physicians | Policy Annualized Premium |
|---|---|---|---|---|
| 1 | Jan 1 – Mar 31 | 120 | Patel, Nguyen, Hoffman | 85,000 |
| 2 | Apr 1 – May 31 | 120 | Patel, Nguyen, Hoffman | 104,000 |
| 3 | Jun 1 – Dec 31 | 110 | Patel, Hoffman, Okafor | 112,000 |
Transaction 5 — ENDORSE: Backdate corrections to Apr 1
Transaction 5 — ENDORSE: Backdate corrections to Apr 1
Effective: Apr 1 · Version: 5Internal audit reveals the bed decommissioning actually happened in April, not June. Dr. Nguyen’s departure was also effective April 1. Dr. Okafor started in April too. Retroactive corrections applied.Segments (3):
fullTermPolicyBilling: Premium 100,000 · Taxes 5,000 · Fees 500 · Total 105,500The structural deltas span Apr 1 – Dec 31, but the Jun–Dec segment already had beds=110, Nguyen removed, and Okafor present — no-ops there. Only Apr–May changed. Segment 2’s annualized rate is 108,000 (vs segment 3’s 112,000) because it lacks Neurology.
Deltas
unchanged| # | Date Range | Beds | Physicians | Specialties | Policy Annualized Premium |
|---|---|---|---|---|---|
| 1 | Jan 1 – Mar 31 | 120 | Patel, Nguyen, Hoffman | Cardiology, Orthopedics, Surgery | 85,000 |
| 2 | Apr 1 – May 31 | 110 | Patel, Hoffman, Okafor | Cardiology, Orthopedics, Surgery | 108,000 |
| 3 | Jun 1 – Dec 31 | 110 | Patel, Hoffman, Okafor | Cardiology, Orthopedics, Surgery, Neurology | 112,000 |
Setup for what’s next: Segments 2 and 3 now share the same beds, physicians, and exposures. They differ only in whether Neurology is a covered specialty — which also produces different annualized rates (108k vs 112k). One more endorsement that adds Neurology to segment 2 will converge both the structural data and the rating.
Transaction 6 — ENDORSE: Retroactive Neurology → MERGE
Transaction 6 — ENDORSE: Retroactive Neurology → MERGE
Effective: Apr 1 · Version: 6Neurology coverage was contractually effective from the clinic opening date (April 1), not June 1 as originally recorded. Correction applied retroactively.The Neurology delta adds it to Apr–May. Jun–Dec already has Neurology — no-op. The rating delta updates Apr–May’s annualized rate from 108,000 to 112,000 to match Jun–Dec.After applying, segments 2 and 3 have identical per-segment state — same structural data, same rating. The system merges them.Segments (2):
fullTermPolicyBilling: Premium 101,000 · Taxes 5,050 · Fees 500 · Total 106,550
Deltas
-1| # | Date Range | Exposures | Specialties | Policy Annualized Premium |
|---|---|---|---|---|
| 1 | Jan 1 – Mar 31 | Main Campus | Cardiology, Orthopedics, Surgery | 85,000 |
| 2 | Apr 1 – Dec 31 | Main Campus + West Clinic | Cardiology, Orthopedics, Surgery, Neurology | 112,000 |
Cancellation and Reinstatement
Transaction 7 — CANCEL
Transaction 7 — CANCEL
Effective: Sep 1 · Version: 7Greenfield is acquired by a hospital network with its own coverage. Policy cancelled effective September 1. The system generates the Segments (3):
fullTermPolicyBilling: Premium 72,000 · Taxes 3,600 · Fees 500 · Total 76,100The Apr–Dec segment split at September. The annualized rate stays 112,000 in both segments — the underlying risk profile didn’t change, only the status. But the billing dropped to 72,000 reflecting the shortened coverage period.
policyStatus delta internally — the caller only provides the cancellation date and updated billing.Request
Cancel and reinstate use a simpler delta format — just
path and value, no startDate/endDate/action. The system handles the date ranges and status change internally.+1| # | Date Range | Status | Exposures | Policy Annualized Premium |
|---|---|---|---|---|
| 1 | Jan 1 – Mar 31 | Active | Main Campus | 85,000 |
| 2 | Apr 1 – Aug 31 | Active | Main Campus + West Clinic | 112,000 |
| 3 | Sep 1 – Dec 31 | Cancelled | Main Campus + West Clinic | 112,000 |
Transaction 8 — REINSTATE → MERGE
Transaction 8 — REINSTATE → MERGE
Effective: Sep 1 · Version: 8The acquisition falls through. Policy reinstated at the same effective date as the cancellation. A 250 reinstatement fee is added.Segment 3 returns to
fullTermPolicyBilling: Premium 101,000 · Taxes 5,050 · Fees 750 (was 500) · Total 106,800
Request
"Active" — now identical per-segment state to segment 2. They merge.Segments (2): -1| # | Date Range | Status | Exposures | Policy Annualized Premium |
|---|---|---|---|---|
| 1 | Jan 1 – Mar 31 | Active | Main Campus | 85,000 |
| 2 | Apr 1 – Dec 31 | Active | Main Campus + West Clinic | 112,000 |
Cancel + Reinstate are invisible in the segments. Both transactions are preserved in the audit trail, but the per-segment state is identical to version 6 (pre-cancellation). Only
fullTermPolicyBilling changed — the reinstatement fee increased fees from 500 to 750. This is the second type of segment reduction: reversal, where an action directly undoes a prior change’s per-segment effects.Transaction Deletion
Transaction 9 — DELETE (undo reinstatement)
Transaction 9 — DELETE (undo reinstatement)
Version: 9The acquisition is back on. The reinstatement was premature — delete it.Transaction deletion removes the most recent transaction (the REINSTATE) and recomputes segments from the prior state. The deleted transaction is archived for audit purposes.The result is identical to version 7 (the CANCEL version):Segments (3):
fullTermPolicyBilling: Premium 72,000 · Taxes 3,600 · Fees 500 · Total 76,100The deleted transaction still exists in the audit trail — it’s archived, not erased. But the live policy state has reverted to the cancelled state, including the pre-reinstatement billing.
+1| # | Date Range | Status | Exposures | Policy Annualized Premium |
|---|---|---|---|---|
| 1 | Jan 1 – Mar 31 | Active | Main Campus | 85,000 |
| 2 | Apr 1 – Aug 31 | Active | Main Campus + West Clinic | 112,000 |
| 3 | Sep 1 – Dec 31 | Cancelled | Main Campus + West Clinic | 112,000 |
Summary
Segment Count Over Time
| Version | Transaction | Description | Segments | Change |
|---|---|---|---|---|
| 1 | NEW_BUSINESS | Initial policy | 1 | — |
| 2 | ENDORSE | Add clinic | 2 | +1 |
| 3 | ENDORSE | Add physician + specialty from Jun | 3 | +1 |
| 4 | ENDORSE | Remove physician, reduce beds from Jun | 3 | — |
| 5 | ENDORSE | Backdate corrections to Apr | 3 | — |
| 6 | ENDORSE | Retroactive Neurology to Apr | 2 | -1 |
| 7 | CANCEL | Cancellation from Sep | 3 | +1 |
| 8 | REINSTATE | Reinstatement from Sep | 2 | -1 |
| 9 | DELETE | Undo reinstatement | 3 | +1 |
Two Types of Segment Reduction
Convergence (Version 6)
Retroactive corrections made the Apr–May segment identical to the Jun–Dec segment — both structurally and in rating. Adjacent identical segments merged.Segments can reduce through normal business operations — corrections, reconciliations, and backdated adjustments that happen to converge per-segment state across a boundary.
Reversal (Version 8)
Reinstatement undid the cancellation, making Sep–Dec identical to Apr–Aug. Adjacent identical segments merged.An action that inverts a prior change can collapse the segments that action created — but the audit trail preserves both transactions.
Key Takeaways
Transactions are authored. Policy state is derived. Segments are compressed representations of per-segment state.Segment count may increase, decrease, or stay the same with each transaction — independently of whether a change was added, corrected, or reversed.
fullTermPolicyBilling and other cross-segment invariants do not affect segment boundaries at all — only per-segment data (including per-segment rating) drives merging.