LIVE·NEON · 1,211 CONTRACTS
Forecast

Calendar Year 2027

Contract-level enrollment forecast for plan year 2027, gated by a 2024 backtest. Prophet where history is long enough; seasonal-naive where it is not; a flat extrapolation for the small remainder of contracts with sparse history.

Run
2026-06-02
Contracts
990
Month rows
11,880
Horizon
12 months · CY 2027
01 · Outlook

National figure · twenty-four months

forecast_enrollment_2027 ⋈ fact_enrollment_cpsc
loading fan chart
02 · Method

How the contracts were forecast

forecast_enrollment_2027
loading method breakdown
03 · Growers

Largest expected gainers

May 2026 → Jan 2027 (12-month forecast horizon)
loading top growers
04 · Decliners

Largest expected losses

May 2026 → Jan 2027
loading top decliners
05 · Quality

Backtest gate

2024 holdout · pre-publish quality check

Median Prophet MAPE was 6.20% on the 2024 holdout backtest; the seasonal-naive baseline came in at 18.34%. Forecasts publish only when the chosen method beats the baseline on its own contract history — otherwise the contract falls back to seasonal-naive or, where even that is unsupported, to a flat extrapolation at the last observation.