⚠ This document is an illustrative end-state sample — read this first
Every other artifact in this suite sits at the 16 October 2026 status date, with the program mid-flight in Stage 2. This one is dated 07 September 2028 and reports actuals that cannot exist yet. It is included because the suite is deliberately documented mid-flight, and that framing carries one risk: a reader concluding the program never delivered. This shows how the program ends.
It is deliberately NOT reconciled against the Program Dashboard, the Gate 2 Readiness Assessment or the Gate Conditions Register. Those documents describe October 2026; this one describes September 2028. Where they disagree, they are supposed to.
Lighthouse Financial Services Company — Gate 5 convened 07 September 2028 to review the closed post-launch measurement window, 06 Mar 2028 to 01 Sep 2028 (179 days). Gate 5 is not an authorization gate: it uses no score bands and cannot stop a product already in the market. Its mandate is to hold the business case to account and transfer benefit ownership.
- What Gate 5 Can and Cannot Read
- Actual Against What Gate 1 Was Told
- Where the Business Case Was Wrong, and in Which Direction
- The Measure That Beat Plan, and Why That Is Not Good News
- Did the Gate 1 Conditions Achieve What They Were Issued to Achieve?
- Did the Board Depart From Its Own Score Bands?
- Benefit Ownership Transfer
- Findings Carried Forward
1. What Gate 5 Can and Cannot Read
Gate 5 sits 185 days after launch and reviews a 179-day window. That is not long enough to read most of what the business case turns on, and the review states the limitation before it states any result.
| Benefit | Readable? | Basis |
|---|---|---|
| BEN-01 Premium volume | Partial | 179 days of sales against a pro-rated plan |
| BEN-02 Rider election | Partial | Election rate on policies issued in the window |
| BEN-04 Average case size | Partial | Mean premium per policy issued |
| BEN-05 Crediting mix | Partial | Allocation at issue; reallocation behavior unobserved |
| BEN-03 Persistency | No | Requires a full policy year. No policy has reached its first anniversary. |
| BEN-06 Capital efficiency | No | Requires 12 months in force. Statutory experience is not yet meaningful. |
2. Actual Against What Gate 1 Was Told
| Measure | Gate 1 plan | Actual | Variance |
|---|---|---|---|
| Premium written in the window | $110,448,000 | $97,119,000 | -12.1% |
| Policies issued | 936 | 891 | -4.8% |
| Average case size | $118,000 | $109,000 | -7.6% |
| GLWB rider election | 62% | 68% | +6% pts |
| Share into the capped indexed strategy | — | 61% | Concentrated |
3. Where the Business Case Was Wrong, and in Which Direction
Gate 5's mandate requires the error to be recorded specifically, not as an aggregate variance.
| The case assumed | Direction of error |
|---|---|
| Year 1 volume achievable through the existing IMO channel | Optimistic. The channel delivered -12.1% against a pro-rated plan — inside the tolerance the Benefits Realization Plan modeled, but in the direction it warned about. |
| Average case size of $118,000 | Optimistic. Actual $109,000. Not previously flagged as a risk by any register. |
| Rider election of 62% | Understated. Actual 68% — see §4. |
| Crediting strategy demand would spread across the three launch strategies | Wrong. 61% concentrated in one strategy — risk R-12 realizing, exactly as raised in dissent at Gate 1. |
4. The Measure That Beat Plan, and Why That Is Not Good News
Rider election came in at 68% against a 62% pricing assumption — +6% percentage points. It is the only measure in §2 that exceeded plan, and it would be easy to report it as the review's one bright line.
5. Did the Gate 1 Conditions Achieve What They Were Issued to Achieve?
The standing efficacy question, asked one final time with outcomes available.
| ID | Verdict at Gate 5 |
|---|---|
| GC-01 | Effective, and vindicated. Cutting to three strategies removed a dependency the platform could not carry. The concentration it created (R-12) did materialize — but it was disclosed as a consequence at the time, not discovered here. |
| GC-02 | Effective. The New York exclusion held; no launch-scope pressure arose from it. |
| GC-03 | Effective as a control. The external review confirmed the pricing basis. Whether the basis was right cannot be answered here — that is BEN-03 and BEN-06, and Gate 5 cannot read either. |
| GC-04 | The efficacy finding recorded in 2026 was correct. The condition produced written evidence of a shortfall and could not close it; the shortfall then materialized at -12.1%. A condition that surfaces a problem it cannot solve is still worth issuing — but only if somebody acts on what it surfaces, and the record shows the gap was disclosed and carried rather than closed. |
| GC-05 | Closed late, and the risk it addressed outlived it. Hedging readiness was achieved before launch. The rider-election variance at §4 means the hedging program is now operating against a larger book of guarantees than it was designed around. |
6. Did the Board Depart From Its Own Score Bands?
Framework §5 requires departures from the score bands to be reviewed here. There were none. Gate 0 scored 3.85 and carried GO; Gate 1 scored 3.20, inside the conditions band, and carried GO WITH CONDITIONS. The Board followed its own framework at every convening.
7. Benefit Ownership Transfer
Gate 5 dissolves the Gate Review Board. Measurement obligations transfer as specified in the Benefits Realization Plan §8: monthly volume, case size and crediting mix to the product line; the annual experience study covering persistency, rider utilization and capital efficiency to Corporate Actuarial; annual reporting of return against the Gate 1 case to the CFO.
8. Findings Carried Forward
- Volume tolerance was consumed before launch and nobody re-planned. Written commitments sat 9.2% below the Year 1 target at Gate 2 — two fifths of the available tolerance — and the program carried the gap rather than closing it. The shortfall that arrived was the one that had been visible for eighteen months.
- Average case size was never on any register. It moved -7.6% and contributed more to the premium miss than policy count did. No risk, assumption or condition addressed it.
- A benefit beating plan can be bad news, and the framework had no way to say so. Rider election exceeded assumption and increased exposure. Nothing in the gate criteria distinguishes a favorable variance from an unfavorable one of the same sign.
- The final gate cannot see the decisive measures. Persistency and capital efficiency are unreadable at Gate 5 by construction. A successor program should either move Gate 5 later or accept explicitly that its last governance event is provisional.
- The U rating earned its place. It is the only mechanism that altered a gate outcome, and it did so by refusing to let an absence of evidence be scored as mediocre evidence.
Prepared by C. Tyrrell, NPD Program Manager and Chair of the Gate Review Board. Approving authority: G. Marchetti, Chief Product Officer. Related: Benefits Realization Plan · Gate Conditions Register · Gate Decision Framework §12 · Cost-Benefit Analysis.